DEPARTMENT OF HEALTH, EDUCATION, AND WELFARE
CONSUMER PROTECTION AND ENVIRONMENTAL HEALTH SERVICE
NATIONAL AIR POLLUTION CONTROL ADMINISTRATION
DIVISION OF PROCESS CONTROL ENGINEERING
DIVISION OF ECONOMIC EFFECTS RESEARCH
CONTROL OF ATMOSPHERIC EMISSIONS
IN THE WOOD PULPING INDUSTRY
FINAL REPORT
CONTRACT NO. CPA 22-69-18
MARCH 15, 1970
VOLUME 2
*#
ENVIRONMENTAL ENGINEERING, INC.. GAINESVILLE, FLORIDA • J. E. SIRRINE COMPANY, GREENVILLE, SOUTH CAROLINA
-------
CONTRACTORS:
SUB-CONTRACTORS:
Environmental Engineering, Inc.
2324 S. W. 34th Street
Gainesville, Florida 32601
J. E. Sirrine Company
P. 0. Box 5456
Ereenville, South Carolina 29606
Reynolds, Smith and Hills
P. 0. Box 4850
Jacksonville, Florida 32201
PolyCon Corporation
185 Arch Street
Ramsey, New Jersey 07441
CONSULTANT:
Professor Donald F. Adams
Washington State University
Pullman, Washington 99163
-------
DEPARTMENT OF HEALTH, EDUCATION, AND WELFARE
CONSUMER PROTECTION AND ENVIRONMENTAL HEALTH SERVICE
NATIONAL AIR POLLUTION CONTROL ADMINISTRATION
DIVISION OF PROCESS CONTROL ENGINEERING
DIVISION OF ECONOMIC EFFECTS RESEARCH
CONTROL OF ATMOSPHERIC EMISSIONS
IN THE WOOD PULPING INDUSTRY
FINAL REPORT by
CONTRACT NO. CPA 22-69-18 E R Hendrickson, Ph. D., P. E.,
MARCH 15, 1970 . Principal Investigator
VOLUME 2
J. E. Roberson, M. S., P. E.,
Sirrine Project Manager
J. B. Koogler, Ph. 0., P. E.,
EEI Project Manager
ENVIRONMENTAL ENGINEERING, INC., GAINESVILLE, FLORIDA
J. E. SIRRINE COMPANY, GREENVILLE, SOUTH CAROLINA
-------
GENERAL TABLE OF CONTENTS
A Detailed Table of Contents for Each Chapter
Will Be Found on the Separator Sheet
Preceding Each Chapter
VOLUME I
Page No.
Letter of Transmittal iii
Abstract v
Acknowledgements vii
Preface ix
Chapter 1 - INTRODUCTION
Air Quality Act of 1967 1-1
General Description of Industry Studies 1-1
Objectives of This Study 1-2
Procedures for the Study 1-2
Chapter 2 - THE CHEMICAL WOOD PULPING INDUSTRY
Summary 2-1
Introduction 2-2
Economic Position 2-4
Present Geographic Distribution 2-6
Forecasts 2-9
References 2-14
-------
Page No.
Chapter 3 - PRESENT PULPING PRACTICES
Summary 3_1
Introduction 3_2
Kraft Pulping 3-12
NSSC Pulping 3-54
Sulfite Pulping 3-62
Chapter 4 - QUANTITY AND NATURE OF EMISSIONS
Summary 4_1
Introduction 4_2
Kraft Gaseous Emissions 4_4
Kraft Particulate Emissions 4-44
NSSC Emissions 4-49
Sulfite Emissions 4-53
Auxiliary Furnace Emissions 4-59
References 4-66
Appendix A - Summary Data for Chapter 2
VOLUME II
Chapter 5 - CONTROL METHODS PRESENTLY IN USE
Summary 5-1
Introduction 5-3
General Description of Control Equipment 5-4
Application, Cost, and Effectiveness of Present
Control Methods 5-25
-------
Page No.
Kraft Sources - 5-33
Sulfite Sources 5-151
NSSC Sources 5-156
References 5-157
Chapter 6 - NEW DEVELOPMENTS IN CONTROL TECHNOLOGY
Summary 6-1
Introduction 5_2
General Description of Control Methods 5-2
Application, Cost, and Effectiveness of New
Control Methods 6-10
Kraft Sources 6-10
Sulfite Sources 6-40
NSSC Sources 6-42
References 6-45
Chapter 7 - CRITICAL REVIEW OF CONTROL TECHNOLOGY
Summary 7_1
Introduction 7-2
Kraft Process 7-3
Sulfite Process 7-18
NSSC Process 7-21
Chapter 8 - POWER BOILER SULFUR RECOVERY
Summary 8-1
Introduction . 8-2
Flue Gas Desulfurization Technology 8-19
-------
Page No.
Process Feasibility Considerations 8-32
R S D Efforts 8-38
References 8-39
Appendix B - Summary Data for Chapter 8
VOLUME III
Chapter 9 - SAMPLING AND ANALYTICAL TECHNIQUES
Summary 9-1
Introduction 9-2
Kraft Sources 9-4
Sulfite Sources 9-65
NSSC Sources 9-76
References 9-77
Chapter 10 - ON-GOING RESEARCH RELATED TO REDUCTION
OF EMISSIONS
Summary 10-1
Introduction 10-2
Emissions Control Technology 10-2
Cost and Effectiveness of Emission Control 10-39
Sampling and Analytical Techniques 10-40
Control Equipment Development 10-50
Process Changes Affecting Emissions 10-54
Chemistry of Pollutant Formation or Interactions 10-57
New Pulping Processes 10-68
Control Systems Development 10-72
-------
Page No.
Chapter 11 - RESEARCH AND DEVELOPMENT RECOMMENDATIONS
Summary 11-1
Areas of Needed Research 11-2
Specific R & D Projects 11-6
Emission Control Technology 11-6
Cost and Effectiveness of Emission Control 11-8
Sampling and Analytical Techniques 11-9
Control Equipment Development 11-10
Process Changes 11-10
Chemistry of Pollutant Formation or Interaction 11-11
New Pulping Processes 11-12
Control System Development 11-12
Other 11-12
Chapter 12 - CURRENT INDUSTRY INVESTMENT AND OPERATING
COSTS
Summary 12-1
Introduction 12-2
Incremental Cost Categories 12-7
Chapter 13 - FUTURE INDUSTRY INVESTMENT AND OPERATING
COSTS
Summary 13-1
Introduction 13-2
Concepts for a Management Model 13-2
-------
Page No.
Analysis of Emission Sources and Controls 13-9
Assignment of Costs 13-33
Trends in Future Capital Expenditures 13-40
References 13-49
-------
CHAPTERS .
, COHTROL METHODS PRESENTLY IN USE
- . TABLE OF CONTENTS , -
Page No.
Summary , 5—1
Introduction ,5-3
General Description of Control Equipment: 5-4
Electrostatic Precipitators , ,5-4
Venturi Scrubbers -, • 5-6
Cyclonic Scrubbers ,, - . 5-9
Impingement Baffle Scrubbers 5-12
Packed Tower Scrubbers 5-14
Mechanical Collectors 5-16
Black Liquor Oxidation 5-19
Orifice Scrubber 5-22
Mesh Pads . 5-22
Tall Stacks , 5-23
Application, Cost, .and Effectiveness of Present-
Control Methods 5-25
Purpose of This Section 5-25
Basis of Selection of Methods 5-25
Definition of Effectiveness 5-25
Basis for Engineering Cost Calculations 5-27
Limitations to Applying Cost Calculations . 5-33
Kraft Sources: "I, , . 5-33
Recovery Systems • - „ 5-33
, Smelt Dissolving Tank , . 5-101
Digester Relief and Blow, Multiple Sffect
Evaporators , 5-109,
Lime Kiln 5-121
Lime Slaker , ..... '. _ ', 5-125
Power and Combination Boilers , ' 5-128
Suifite Sourcesi- . . 5 5-151
Acid Tower , , ... - 5-151
Blow Pit : ! - , '-,,. ' *.-••• \' ' '•'--* -5-153
NSSC Sources ... "' " * .- - .- " " , . - ^,- 5-156
References ' , . . • : •'. , . • ., ' •• " 5-157-
.5-1
-------
CHAPTER 5
CONTROL METHODS PRESENTLY IN USE
SUMMARY
Control methods presently in use in the wood pulping industry
consist of add-on hardware or process^modifications. The methods
considered in this chapter are those which have been in reasonably
successful operation for at least one year at one or more locations.
The one exception is the recently developed recovery systems
which eliminate direct contact between the flue gases and black
liquor. Control methods are briefly described in general terms
and are evaluated under conditions of specific applications.
The evaluations include a cost and effectiveness study as well
as a discussion of engineering factors which are unique to
the application. Cost calculations have been prepared for
the total capital cost and net annual cost. It must be recognized
that installations vary widely from mill to mill, thus the costs
presented can serve only as a general guide. Cost estimates for
specific mills must be based on the situation at hand.
A variety of methods is found to be useful in controlling particu-
late emissions from pulp mill sources. Efficiencies of 99 + percent
are possible. The listing which follows identified those methods
most commonly used and which are described and evaluated in the
chapter. It should be noted that the direct contact evaporator
following the kraft recovery furnace is an important particulate
control device. It should also be noted that where primary or
secondary wet scrubbers are used as particulate collectors on
combustion sources, the emission of gaseous sulfur compounds may
be increased or decreased depending on the nature of the scrubbing
medium.
Kraft Recovery Furnace - Electrostatic Precipitators
- Venturi Evaporator/Scrubbers
- Electrostatic Precipitators
plus Secondary Scrubber
Kraft Lime Kiln - Venturi Scrubber
- Cyclonic Scrubber
- Impingement Baffle Scrubber
5-1
-------
Kraft Smelt Dissolving Tank
Kraft Lime Slaker
Power Boilers
Combination Boilers
- Mesh Pads
- Packed Tower Scrubbers
- Orifice Scrubbers
- Mesh Pads
- Cyclonic Scrubbers
- Mechanical Collectors
- Electrostatic Precipitators
- Mechanical Collectors
- Cyclonic Scrubbers
Fewer numbers of methods are in use for the control of gaseous
emissions, particularly the odorous reduced sulfur compounds.
The majority of odorous emissions can be grouped into three categories:
(a) recovery furnace offgases; (b) low volume high concentration sources
such as the noncondensible gases from the multiple effect evaporators,
and digester relief and blow gases; (c) high volume low concentration
sources such as brown stock washers and smelt dissolving tank. Wet
scrubbers have received limited application on combustion sources in
the kraft process because of the difficulty of absorbing effectively
all of the odorous compounds with a single scrubbing medium. Also as
indicated previously, the emission of gaseous sulfur compounds may
be increased or decreased depending on the nature of the scrubbing
medium. The following list identifies those methods most commonly
used and which are described and evaluated in the chapter:
Kraft Recovery Furnace
Kraft Smelt Dissolving Tank
Kraft Digester Relief and
Blow plus M. E. Evap-
orator
Sulfite Acid Tower
Sulfite Blow Pit
Weak Black Liquor Oxidation
Strong Black Liquor Oxidation
Proper Operation
Venturi Evaporator/Scrubber
Cyclone Evaporator/Scrubber
New Recovery System Design
Packed Tower Scrubber
Orifice Scrubber
Chlorination
Incineration
Packed Tower Scrubbers
Weak Black Liquor Oxidation
(M.E. Evaporators)
Additional Absorption Tower
Packed Tower
5-2
-------
5.1 INTRODUCTION
Control methods presently in use in the wood pulping industry
consist of add-on hardware or process modifications.
The methods considered in this report are those which have been
in reasonably succesful operation for at least one year at one
or more locations. The one exception to this is the inclusion
of recently developed recovery systems which .eliminate direct
contact between the flue gases and black, liquor .wt-£everal«of::
these systems'are being installed and are described because of
the potential they offer for reducing odorous emissions and the
interest expressed in these units by the industry. For a sum-
mary of the estimated number of control methods presently in
use in the industry, the reader is referred to Chapter 12,
Table 12-1.
In this chapter, control methods are briefly described in general
terms and are evaluated under conditions of specific applications.
The evaluations include a cost and effectiveness study and a dis-
cussion of engineering factors which are unique to the application.
Cost calculations are prepared for the total capital cost and the
net annual cost. These calculations and the assumptions and defi-
nitions used are included in Section 5.3. It should be emphasized
that the costs presented in this report are based on generalized
unit cost,data and certain specified assumptions. The costs can
be very useful for comparing several control methods as applied to
a specific source, but they should not be used as an absolute cost
estimate for the application of a specific control method to a
specific source. The application of control methods from mill to
mill varies considerably depending upon such factors as the regional
cost index, the physical arrangement of the equipment, and the
capacity of equipment. For this reason cost estimates for specific
installations must be based upon costs and other considerations
unique to the situation at hand.
The capital costs included in this report are based on January, 1969,
prices. Since it would be desirable to be able to escalate these
capital costs to meet future conditions an attempt was made to pre-
pare a curve for escalation of these costs. Due to the diverse
nature of the control equipment and process changes, it was found
that there was not one set of curves that could be used to escalate
5-3
-------
the capital post of all control methods on an individual basis.
Generalizing then, the contractors believe that the cost for a
number of control methods applied to a number of different mills
can be escalated by the use of equipment indices prepared by
Marshall s Stevens Publication Company, 1717 Beverly Boulevard,
Los Angeles, California 90026. These indices are entitled
''Equipment-National Average" and are listed under the subheading
entitled "Paper Manufacturing." For a more precise capital cost,
the cost of individual control methods must be individually recal-
culated in the future.
5.2 GENERAL DESCRIPTION OF CONTROL EQUIPMENT
5.2.1 ELECTROSTATIC PRECIPITATQRS
An electrostatic precipitator consists in principle of a
number of discharge (emitting) electrodes, collecting
electrode plates and a high-voltage power unit. This
power unit comprises high-voltage transformers and recti-
fiers to convert the available AC power to high-voltage
DC power. Lately, silicon diode rectifiers have almost
exclusively replaced other types of rectifiers.
The dust laden gas enters the electrostatic precipitator
and flows in the passages created by the collecting
electrodes. When high-voltage power is applied to the
discharge electrodes located in the passages, ionization
or corona discharge occurs near the surface of the dis-
charge electrodes. The negative ions attach themselves
to dust particles near the electrodes giving the particles
a negative electric charge. The charged dust particles are
repelled by the discharge electrodes and attracted to the
collecting surfaces connected to ground. Here the dust
particles lose their electrical charge and are deposited on
the collecting electrodes.
The collecting electrodes are provided with a rapping
mechanism for dislodging the dust precipitated on the
electrodes. This rapping system must be designed very care-
fully to avoid creation of a dust cloud in the space between
the electrodes. Incorrect rapping operation or programming
may cause the dust to re-entrain which, in turn, would greatly
decrease the dust collecting efficiency of the precipitator.
Incorrect rapping is the most common cause for so-called
"snow-outs." The rapping mechanism, which is generally
operated on an automatic time cycle basis, is either of
5-4
-------
' vibrator;.-or- -motorrdriveri fall-hammer: • type. \Vibrators
are preferred for pulp mill applications rather than the
hammer rappers which are used more frequently, for example,
on fly ash precipitators.
The high voltage power level is controlled by an automatic
control unit. The discharge electrodes are connected to
the negative rather than to the positive side of the power
supply. A higher voltage can thereby be maintained without
excessive arc-overs which, in turn, would cause waste of
electric power and eventually lead to operating difficulties.
The automatic voltage rate of sparking provides a feedback
to the control unit making it possible automatically to
operate the electric system of the precipitator at optimum
conditions. Little or no sparking will increase the voltage
and too much sparking will reduce the voltage, thereby pre-
venting arc-overs detrimental to the precipitator.
Gas conditions are never uniform throughout the precipitator.
Aside from unintended irregularities in gas distribution and
other conditions, the dust loading varies from inlet to outlet
of the precipitator. To reach optimum dust collecting efficien-
cies , the electrical control units must accommodate these varia-
tions. A high dust loading will increase the sparking and
reduce the voltage. If the precipitator would have only one
electric system with one automatic voltage control, the voltage
throughout the precipitator would be limited to the lowest vol-
tage permissible at any point in the precipitator. The modern
precipitator is, therefore, divided into independent electrical
units, each controlled for maximum voltage depending on the gas
conditions in that particular section of the precipitator. The
discharge system close to the outlet of the precipitator will
consequently operate at a higher voltage level than the inlet.
This need for sectionalizing has led to increased use of the
horizontal flow, plate-type precipitator. This type has super-
seded the tube-type, vertical flow precipitator, once very
common in the pulp and paper industry. The tube-type precipi-
tator does not easily lend itself to sectionalizing.
The electric power consumption of a precipitator depends on
several factors. The input of current is necessary to sustain
the voltage level. There is a constant power drain due to
ionization in the corona, due to the controlled sparking and
other voltage leakages because of poor maintenance, dust build-
up, et cetera. The collection efficiency depends on the voltage
level; but, on the other hand, the higher the voltage level, the
5-5
-------
greater the power drain. At 90 percent collection
efficiency, the power consumption is approximately 0.2
kw/1000 CFM of gas and at 99.9 percent efficiency, the
power consumption has risen to approximately 0.8 kw/1000
CFM.
Most precipitators used in the wood pulping industry on
recovery boilers are designed for collection efficiencies of
90 - 99.9-oercent. The pressure drop of the gas passing
through the precipitator is usually below 0.5 inch W.G. A
typical wet-bottom precipitator with tile shell is -shown .in:. •
Figure 5-1. -
5.2.2 VENTURI SCRUBBERS
The Venturi scrubber consists in principle of a convergent
section (throat) and a divergent section. Dust laden gas
enters the convergent section and is accelerated to high
velocity as it approaches the throat. Gas velocities in the
throat section vary from 100 to 500 FPS.
Water or other scrubbing liquid is injected either directly
into the throat section or the top of the Venturi. In the
latter case, the scrubbing liquid cascades down the walls
of the convergent section. The high velocity gas stream
atomized the liquid into a fine mist—the greater the velocity,
the finer the droplets. Collision between the dust particles
and the water droplets takes place and causes the dust to be
entrapped in the water. Further collision between the water
droplets occurs. This will create aggregate droplets of rela-
tively large size. These droplets are easily separated from
the gas stream in a subsequent separator. The collision or
impaction phenomenon is rather complex, but is mainly due to
mass forces created by the great velocity differential between
the dust particles and the water droplets. For submicron
particles, the Brownian molecular movement, diffusion, and
electrostatic forces also play an important role.
In the divergent section, the gas and the dust particles are
decelerated thereby creating a new velocity differential with
additional agglomeration. Finally, in the elbow connecting
the Venturi and the separator, as well as in the inlet to the
separator (if radial), changes of direction of the gas flow
cause additional impaction and agglomeration.
The gas and liquid enter the separator, usually of the cyclonic
type, where the liquid is thrown to the walls by centrifugal
forces and drains to the bottom by gravity. The clean gas
exits through the upper portion of the separator.
5-6
-------
GAS
OUTLET
BLACK LIQUOUR F E
GAS
IHLET
FIGURE 5-}
PRECIPITATOR FOR RECOVERY BOILER.
5-7
-------
The elbow connecting the Venturi with the separator
is either of the regular type or of so-called flooded
type. The flooded elbow is developed to prevent erosion.
Here the liquid surface absorbs the brunt of the water
and dust stream. For pulp mill applications, such as
lime kiln scrubbers, the flooded elbow is recommended.
Normally the separator is of the cyclonic type with
conical bottom and either top outlet or side outlet.
The type of outlet depends on whether the fan is located
before or after the scrubber. The top outlet lends itself
to a direct mounted stack. In order to facilitate the
installation, the separator is sometimes provided with a
flat bottom. This may sometimes create drainage problems.
For special purposes a so-called false bottom is also
installed. The false bottom creates an intermediate re-
serve for the scrubbing liquid where de-aeration can take
place making the scrubbing liquid (e.g., black liquor)
more suitable to pump. Scrubbers for recovery boilers are
almost always provided with facilities for wall wash of the
separator. This prevents build-up of black liquor solids
which can be a fire hazard, in addition to being a main-
tenance problem.
The Venturi scrubber is capable of high efficiency collec-
tion of dust and fumes even in the submicron range. The
efficiency of collection is a function of the pressure drop
across the scrubber which in turn is a function of the gas
velocity in the throat and the liquid flow rate (liquid to
gas ratio). The higher the gas velocity or the liquid flow
rate, the greater will be the pressure drop and consequently
the greater the efficiency. There is, however, a cut-off point
where increased liquid flow rate will have an adverse effect
on the collecting efficiency. The throat becomes "flooded."
This occurs at about 20 gallons per 1000 CFM. In order to
attain high collecting efficiency, the water droplet size
distribution has to be in a certain relation to the dust
particle size distribution. Too large droplets would reduce
the probability of collection drastically, and even if col-
lection would take place the chance for the dust particles to
be entrapped in water would also be reduced.
Increased gas velocity (pressure drop) increases the atomization.
A fine dust will, therefore, require a higher pressure drop
than a coarse dust for the same collecting efficiency.
5-8
-------
Scrubbing liquid is injected into the Venturi at low
pressure through relatively large jets or open weirs.
In order to be able to recirculate the scrubbing liquid,
in other words maintain a high concentration of solids,
the manufacturers try to completely do away with narrow
constrictions and weirs. The liquid rate is normally
3-15 gallons per 1000 CFM of gas. A typical Venturi
scrubber is shown in Figure 5-2.
5.2.3 CYCLONIC SCRUBBERS
The cyclonic scrubber is an efficient device for removing
dust particles two microns and larger and is also a rela-
tively good gas absorber. In this unit, the dust laden
gas enters tangentially at the bottom of a cylindrical
tower and spirals upward through the scrubber in a con-
tinuously rotating path. A spray manifold is located
axially in the center of the scrubber with banks of spray
nozzles directed radially toward the walls. The spray
sweeps across the path of the gas stream intercepting and
entrapping the dust particles. The centrifugal motion of
the spray imparted by the rotating gas causes the droplets
to impinge against the walls of the scrubber and drain to
the bottom due to gravitational forces.
In another type of cyclonic scrubber, the spray nozzles
are mounted on the wall rather than in a central spray mani-
fold. The advantage with this type is that the nozzles may
easily be serviced or replaced while the scrubber is in
operation.
The mechanism of collecting particles in the cyclonic scrubber
is in essence the same as in a Venturi scrubber; namely,
impaction and agglomeration of liquid droplets and dust parti-
cles with subsequent centrifugal and gravitational separation.
The liquid pressure ranges from 50 to 400 PSIG. The flow rate
is normally 3-8 gallons per 1000 CFM.
The cyclonic scrubber is most efficient on relatively coarse
dust and the efficiency drops off markedly for particles under
two microns. The pressure drop is considerably less than for
a Venturi scrubber and ranges normally from 0.5 to 3 inches WG.
A typical cyclonic scrubber is shown in Figure 5-3.
5-9
-------
SEPARATOR OUTLET
LIQUID INLETS
THROAT-
SEPARATOR INLET NOZZLE
FIGURE 5-2
VENTURI SCRUBBER WITH CYCLONIC
SEPARATOR
5-1 n
-------
SCRUBBER OUTLET
GAS
INLET
LIQUID INLET
WALL MOUNTED SPRAYS/
MAY BE CENTER PIPE,
FIGURE 5-3
CYCLONIC SCRUBBER
5-11
-------
5.2.4 IMPINGEMENT BAFFLE SCRUBBERS
The impingement baffle scrubber is a vertical tower
equipped with one or more impingement baffle stages.
The impingement baffle consists of a perforated plate
having a multitude of small holes so arranged that a
baffle is located directly above each perforation. A
weir on each plate maintains a level of scrubbing liquid.
The contaminated gas enters radially at the bottom of
the scrubber and is subjected to a water spray that will
precipitate out the coarser dust particles. The gas then
passes through the perforations and impinges on the baffles.
The gas velocity through the perforations, ranging from
75 to 100 FPS, creates an atomization of the scrubbing
liquid and the velocity differential between gas and liquid
results in the impaction and agglomeration of liquid and
particulate matter. The scrubbing liquid is introduced into
the scrubber through low pressure spray nozzles located below
the plates and spraying upward. These sprays not only deliver
liquid to the plates, but also serve to cool, humidify, and
condition the gases; remove coarse particles; and keep the
bottom of the plates clean. For scrubbers with more than one
stage, the scrubbing liquid is drained off the plates down
ward from stage to stage in the scrubber. In addition to the
multiple impaction, the change of solids concentration in the
liquid also contributes to increasing the collecting efficiency.
The pressure drop for impingement baffle scrubbers ranges from
2 inches WG to 8 inches WG depending on the number of stages,
size, and numbers of perforations and baffles. As would be
expected, increased number of stages and smaller perforations
result in higher pressure drop and subsequent higher efficiency.
An impingement baffle scrubber is shown in Figure 5-4.
The impingement baffle scrubbers are normally used in lime kiln
applications. The efficiency is, however, not high enough to
comply with today's air quality requirements. For upgrading
of existing impingement baffle scrubbers, it is possible to
install a Venturi and an elbow ahead of the scrubber and to
use the scrubber shell as a cyclonic separator. The inlet to
the scrubber must, in such a case, be changed to enter the shell
tangentially, and all the internal parts have to be removed.
When estimating the cost of altering an impingement baffle
scrubber to a Venturi, the cost figures for Venturi scrubbers
may be used. The actual cost will vary largely, depending on
local conditions.
5-12
-------
TARGET PLATE
CLEAN GAS OUTLET
ENTRAPMENT
ELIMINATOR
IMPINGEMENT
BAFFLE PLATE
(SECOND STAGE)
WATER SEAL
IMPINGEMENT
BAFFLE PLATE
(FIRST STAGE)
WATER LEVEL
ORIFICE PLATE-
FIGURE 5-4
IMPINGEMENT SCRUBBER
5-13
IMPINGEMENT SCRUBBER
MECHANISM
LIQUID INLET
WATER SEAL
\CONTAMINATED
GAS INLET
-------
5.2.5 PACKED TOWER SCRUBBERS
The packed tower scrubber is used primarily as a gas
absorber. Its use for collection of solid particulates
is limited because it is rather inefficient for particles
under five microns and is subject to becoming plugged
because of dust build-up. It is, however, an excellent
device for absorption of such gases as HC1, SO , Cl ,
H S, and NH . Other advantages of the scrubber are simple
design and low manufacturing cost.
The scrubber consists of a vertical cylindrical shell with
the gas inlet at the bottom and the outlet at the top.
Above the gas entrance is a packed section consisting of
four or more feet of packing material. This packing may
consist of Raschig rings, Pall rings, saddles, et cetera,
made from stoneware, ceramic, or polypropylene. Water is
distributed uniformly over the packing by means of low
pressure spray nozzles or weirs located above the packing.
Normally, there is a mist eliminator above the spray nozzles
to prevent entrainment of liquid in the clean gas leaving
the tower. Normally a reservoir is located at the bottom
of the scrubber for direct recirculation. A typical packed
tower scrubber and packing are shown in Figure 5-5.
The contaminated gas enters at the bottom of the tower and
moves upward through the packing counter-current to the
scrubbing liquid. The packing forces the gas to follow a
tortuous path over the contact surfaces and interstices
creating intimate .contact with the descending liquid. .
The pressure drop over the tower depends on the height of
packing but is normally in the order of 1-2 inches VJG.
As mentioned before, the packed tower is an excellent gas
absorber, but less suitable for dust* removal due to dust
build-up. When dust is present in the gas, special attention
has to be given to the selection of packing material. A
larger size packing,- i.e., 4-inch Pall rings instead of
2-inch should be chosen in order to prevent plugging. The
contact surface will thereby be reduced and the gas absorption
efficiency will drop. This can be compensated for by increas-
ing the height of the packing but the plugging tendencies will
increase simultaneously.
5-14
-------
CLEAN GAS OUTLET
DEMISTER PAD
CONTACT BED
CONTAMINATED
GAS INLET
LIQUID INLETS
RECYCLE SECTION
COUNTER-FLOW PACKED TOWER SCRUBBER
SPIRAL
RING
BERL
SADDLE
INTALOX
SADDLE
CERAMIC PACKINGS
FIGURE 5-5
RASCHIG LESSING CROSS-PARTITION
RING RING RING PALL RING
TELLERETTES*
MASPAC
I NTALOX
SADDLE
PLASTIC PACKINGS
5-15
-------
5.2.6 MECHANICAL COLLECTORS
There are a great number of different mechanical col-
lectors in use throughout industry. The most common
ones within the wood pulping industry are the large
diameter cyclones and the multi-tube collectors. This
description will be limited to these two types.
Cyclone dust collectors are of cylindrical or conical
type, utilizing centrifugal and gravitational forces
for separation of dust particles from a gas stream.
The dust laden gas enters the collector tangentially
either directly or via an expanded involute section
where the dust particles are subjected to the separa-
ting forces. The centrifugal force drives the dust
particles to the collector wall; the gravitation drives
the concentrated dust downward to the cone outlet; and
the dust is discharged into a collection hopper while
the cleaned gas flows upward in an inner vortex to the
gas outlet tube.
Two basic types of cyclone collectors are available—
the tangential inlet type shown in Figures 5-6 and 5~7
and the axial vane type. The former is sometimes referred
to as large diameter cyclone, or cyclone collector, and
the latter one as a tubular collector. In the tangential
inlet type, the gas enters the cyclone through a straight
tangential, helical, or involute inlet section. Axial
vane units employ inlet vanes to provide the spiraling
motion to the dust laden gas stream.
Many sizes and designs of cyclone collectors can be pro-
vided to meet specific dust collection problems. The units
may be installed in single or multiple arrangements, in
parallel or in series. Cyclone collectors are generally
suitable for separating solid particles in size ranging
from about 3 microns to 200 microns. They can, of course,
be used for larger particle sizes also.
For a given cyclone collector, the overall collecting
efficiency can be -determined if certain parameters are
known. These parameters include dust particle size distri~
bution, density, concentration, and other properties of the
dust. Gas temperature, pressure, moisture content and gas
composition, as well as pressure drop limitations and local
5-16
-------
GAS
INLET
GAS
OUTLET
DUST OUTLET
FIGURE 5-6
LARGE DIAMETER CYCLONE COLLECTOR
STANDARD ARRANGEMENT
5-17
-------
GAS
INLET
OUTLET TUBE
GAS
OUTLET
MULTI-TUBE COLLECTOR
SPIN VANES
INLET TUBE
FIGURE 5-7
COLLECTOR ELEMENT
5-18
-------
conditions, are also factors of consideration. The
efficiency increases with increased dust particle
size, density, and concentration. Collection efficiency
increases with increased pressure drop across the cyclone.
The collector can be designed for pressure drops of less
than 2 inches WG for large diameter cyclones and for 2-7
inches WG for the smaller diameter (high efficiency) units.
Increasing temperature will decrease the efficiency of the
cyclone. However, up to 700°F, the influence of the gas
temperature is not great. For the same pressure drop,
increased gas viscosity or density will lower the collecting
efficiency.
As particle size , dust density, and inlet gas velocity
decreases the efficiency tapers off. Overall efficiency
of the collection system can be increased by arranging
the cyclones in series . By reducing the diameter of the
cyclone the centrifugal forces increase, thus increasing the
efficiency. The gas volume capacity will, however, simul-
taneously be reduced and the number of cyclones has to be
increased. This principle is utilized in the multi-tube
collectors where the tube diameter is from 12 inches or less
up to 24 inches . Batteries of tubes are mounted in the same
casing and high efficiencies are attainable.
Since high efficiencies require high radial gas velocities,
abrasion is often a problem for cyclone collectors. Proper
selection of material of construction is imperative. Ma-
terials in use include carbon steel , low alloy steels , alumi-
num and special materials . Abrasion resistant linings in
cas table form or brick linings can also be used.
5 . 2 . 7 'BLACK LIQUORv OXIDATION
Blacktliquor oxidation is practiced for the purpose of oxi-
dizing the sodium sulfide in the liquor. This is accomplished
by reacting the sodium sulfide with oxygen in the air to form
sodium thiosulfate. This compound is relatively stable and
will not break down in passing through the direct contact
evaporator. Sodium sulfide reacting with the carbon dioxide
and sulfur dioxide in the flue gases is the cause of hydrogen
sulfide emissions . The chemical reactions of sodium sulfide
(unoxidized liquor) and CO and SO in the fMue gases are as
follows :
Na0S. .+ C00 + H_0 ^ Na_ ,pO, t^JI^S;:~
f, £ ' £• ,&-.T«J f- J . ; ". ^
5-19
-------
Black liquor oxidation can be applied either to weak black
liquor or to concentrated black liquor. The choice is
dependent to a considerable extent on the characteristics
of the liquor. Usually weak liquor oxidation is used on
non-foaming liquor and concentrated liquor oxidation on
foaming type liquor. Regardless of the type of oxidation,
it is essential that complete oxidation of the sodium sulfide
takes place, if hydrogen sulfide emissions from the direct
contact evaporator are to be prevented.
Lately, investigations have been conducted into sodium
sulfide reversion taking place after weak liquor oxidation.
Instances are known to exist where, although there is 99 plus
percent oxidation during weak liquor oxidation, the concentrated
liquor shows the presence of sodium sulfide. One solution is to
add a second system of concentrated liquor oxidation. Among
those who have installed weak liquor oxidation and who are having
sodium sulfide reversion, there are suggestions that concentrated
liquor oxidation might be the correct installation (14).
There are three types of oxidation systems:
1. Packed Towers
2. Bubble Tray Towers
3. Air Sparged Reactors.
Packed Towers
Packed towers have been applied primarily to weak black liquor
oxidation. The principal advantage of a packed tower lies in
reduced power costs. The pressure drop through a packed unit is
usually about 1 to 2 inches WG, and a tower for a 500 ton-per-day
mill will require about 20 hp to operate the required air blowers.
In regions of high power costs, this is an important factor in
selecting an oxidation unit. The disadvantage of this unit is that
it has plugging tendencies and has lower oxidation efficiency.
Bubble Tray Towers
The bubble tray oxidation tower is used for weak black liquor
oxidation only. The liquor is pumped out on a perforated steel
plate, under which air is blown from a fan. The air passes through
the perforations and bubbles through the liquor. The liquor height
on the plate is normally four to six inches and the liquor makes
several passes over the plate. In'order to accommodate larger flows
of liquor, a number of these aeration chambers (bubble trays) are
connected in parallel and stacked on top of each other.
The liquor, air, and foam are discharged into a foam tank, where
mechanical foam breakers convert the foam to liquor. Certain
liquors with low foam characteristics may not require foam breakers
if the tank is large enough. The air leaving the system through
5-20
-------
foam breakers carries some entrained liquor. This liquor
is separated from the air in a cyclone and returned to the
foam tank, while the air is exhausted to the atmosphere.
This air stream, although relatively small, contains some
hydrogen sulfide, methyl mercaptan, dimethyl sulfide, and
dimethyl disulfide.
The system operates at 25 - 35 inches WG pressure drop.
The foam breakers draw approximately 10 BHP each. Includ-
ing necessary pump capacity, the total power consumption
runs about 0.25 BHP/GPM. The air to liquor ratio; is 15 -
20 CFM/GPM and a retention time of 4 - 6 minutes is required
on the perforated plates.
Air Sparger Reactors
The air sparged reactors are of two types—air sparging with
agitation and without agitation.
Air Sparger With Agitation. Air sparging with agitation is used
for both weak and concentrated liquor. This system requires
more electric power, it is more complicated, and the equipment
cost is higher than for the non-agitation system. The liquor is
pumped into a tank where an air header comes in centrally at the
top. The header conveys the air into the sparger located approxi-
mately six feet from the bottom of the tank. The sparger has a
number of arms extending radially from the header and each arm
has a number of branches with aeration nozzles. The sparger is
submerged ten to fifteen feet in the liquor depending on the
desired retention time. The incoming liquor is distributed above
the surface in a system of pipes and nozzles. In addition to
being evenly distributed, the liquor helps to beat down the foam
floating on the surface.
Air Sparger Without Agitation. Air sparging without agitation is
used exclusively for concentrated black liquor oxidation. This
system operates at about 10 psig air pressure and draws approxi-
mately 1.5 - 2.0 BHP/GPM of liquor. The air to liquor ratio is
20 - 30 CFM/GPM and the retention time 2-3 hours. Systems with
agitation use more horsepower, but should otherwise be comparable.
Effect of Batch and Continuous Digesters
Mills with continuous digesters usually experience higher sodium
sulfide loadings than mills with batch digesters. This is at-
tributed to the fact that batch digester systems expose the liquor
5-21
-------
to more contact with air (primarily during washing) than
continuous digester systems, thus resulting in more oxi-
dation of the liquor. For batch digester systems the sodium
sulfide content in the weak black liquor storage tanks may
be 50 percent of that in the white liquor storage tanks. This
condition must be considered when selecting oxidation systems.
Summary
As stated previously, foaming is a major consideration in the
selection of an oxidation system. Foaming depends on the wood
species, which will contain varying amounts of resin and fatty
acid salts. These compounds will create a foam, that in some
cases will preclude the oxidation of weak liquor. The foaming
is less pronounced in strong liquor. Especially severe is the
foaming of liquors from southern pine. The addition of fuel
oil or kerosene to the liquor to reduce the foam helps in some
cases. Foaming characteristics of the liquor influences the
geographical locations of the various oxidation systems. In
general, the air sparger type is located in the South due to
the resinous content of the southern liquors. Weak liquor
oxidation has found extensive application in western United
States. The mid-west and northeastern United States have
installed primarily weak liquor oxidation systems; however, (1)
only one installation is reportedly operating at a high oxi-
dation efficiency. Currently, the general trend in oxidation
systems appears to be directed towards more acceptance of the
air sparger type for concentrated liquor oxidation.
5.2.8 ORIFICE SCRUBBER
The orifice scrubber is a collection device consisting of a
restricted air passage partially filled with water. The re-
sulting dispersion of the water causes wetting of the particles
and their collection. Pressure drop is comparable to cyclonic
scrubbers.
5.2.9 MESH PADS
Mesh pads are collection devices composed of material such as
knitted wire mesh. Dust and liquid droplets are collected on
the pads.
A spray washing system is provided for back washing the mesh pads
to remove accumulated solid particulate. In normal operation, a
maximum pressure drop of approximately 0.2 inches WG is maintained
by periodic operation of the spray system.
5-22
-------
5.2.10 TALL STACKS
Tall stacks have been recognized by many authorities as
a method which may be used to reduce the concentration of
pollutants at ground level. Of course, the use of the
tall stack will not in itself reduce the amount of pollutant
being discharged into the atmosphere (17).
In discussing tall chimneys, Carlton-Jones and Schneider (13)
state that "A chimney's air pollution control function is to
discharge gases at a point high enough so that the maximum
concentration of odors and toxic components at ground level
is within acceptable limits. It is seldom practical to rely
on chimneys alone for solving particulate pollution problems.
For most odors and gases, however, chimneys may be a practical
answer."
While there is no such thing as a standard chimney, the cost
curves in Figure 5-8 are a rough estimate of chimney cost.
These costs are generally based upon the following considerations:
Concrete Outer Shell
Free Standing Inner Liner Constructed of Acid Resistant
Brick and Mortar
Inner Liner for the Entire Chimney Height
Foundation Allowances Included
Platforms and Ladders
Lightning Protection System
Obstruction Lighting
Wind Velocity of 100 MPH
Indirect Cost of 30 Percent
These chimney costs are based upon data from Carlton-Jones and
Schneider (13) and have been updated to reflect January 1969
costs. Of course, these rough estimates will be influenced by
many factors including geographic area, soil conditions, design
wind velocity, labor costs, and existing physical limitations at
a specific mill.
5-23
-------
1600
o
o
o
r 1400
X
-*/»•
CO
o
o
o.
-------
5.3 APPLICATION, COST, AND EFFECTIVENESS OF PRESENT CONTROL METHODS
.3^1 PURPOSE OF THIS SECTION
A number of control techniques for gases and particulates
has been developed and applied in the wood pulping industry
in recent years. In order to provide a basis for comparing
the cost-effectiveness of alternative solutions to controlling
emissions from a single source and for evaluating the most
effective complete system, engineering estimates are essential.
.3.2 BASIS OF SELECTION OF METHODS
This chapter of the report is concerned with control methods
presently in use in the wood pulping industry. Therefore,
only control methods which have been applied successfully in
at least one United States mill for at least one year are in-
cluded in this chapter. The only exception to this selection
basis is the inclusion of capital costs for new 'recovery.
system designs which .have a great potent ial^ fgr^^educ ing
' "" "'
odorous 'emissions and were. 4nfiT*^.|>ecause "of- the"' interest"
expressed by the industry. -:*rr-"ns- . —
Because these methods represent United States applied practice,
techniques are not described for some sources and for some
emissions . The reader is referred to Chapter 6 for a review of
promising potential control methods .
DEFINITION OF EFFECTIVENESS
A concise definition of effectiveness of an emission control
method is difficult to obtain because of the many factors to
be considered. In addition to the physical efficiency of a
device, its reliability, application, operating conditions,
degree of maintenance, and energy reliability must be con-
sidered. For the purpose of comparing control methods, this
report discusses the effectiveness of the methods in terms of
the following categories:
1. Particulate removal
2. Sulfur oxide removal
3. Total reduced sulfur (TRS) removal
4. Operation
Particulate Removal. The particulate removal of control methods
has been evaluated in terms of "Mean Annual Operating Efficiency"
5-25
-------
(A.O.E.) which is a function of a number of variables
including the "Guaranteed Efficiency." The "Guaranteed
Efficiency" is defined as the manufacturer's guaranteed
efficiency which is calculated in accordance with
Industrial Gas Cleaning Institute (IGCI) procedures.
IGCI specifies a test procedure (15, 16) for determining
the collection efficiency of particulate collectors, similar
to ASME Power Test Code 21, which is used by most equipment
manufacturers for guaranteeing collector efficiencies.
The method is based on a determination of the average
particulate concentration at the inlet and outlet of the
collector. Isokinetic sampling at specified locations
in the ducts for specified times is required at steady
operating conditions. The particle collector is specified
as a filter of any material and form which has a collection
efficiency in excess of 99.0 percent for particulates of
the approximate size distribution to be encountered during
the test. Collectors other than filters are acceptable if
it can be demonstrated that they have an efficiency exceed-
ing 99.0 percent at test conditions.
The "Mean Annual Operating Efficiency" (A.O.E.) is defined
as:
Total Annual Particulate to Control Method - Total Annual
Particulate Leaving Control Method
Total Annual Particulate to Control Method
x 100
The A.O.E. is a value that is difficult to obtain in actual
practice. The contractor has used past experience, operating
data which has been obtained from a number of mills, and
theoretical calculations to derive A.O.E.'s which are
theoretical, but are considered typical of the particular
control method.
Sulfur Oxide and Total Reduced Sulfur (TRS) Removal. The con-
trol methods are also evaluated in terms of their effectiveness
in removing sulfur oxides and TRS. Since data are limited in
this area, the discussion is generally limited to whether the
control methods have any potential for removing the sulfur oxides
and TRS.
Operation. The effectiveness evaluation also includes a compari-
son of the operation of control methods. Included in this com-
parison are such things as ease of maintenance, corrosion
potential, and plume development.
For each source, the most effective method is chosen based on
the effectiveness evaluations and the control method cost.
5-26
-------
5.3.4 BASIS FOR ENGINEERING COST CALCULATIONS
Capital costs and annual costs have been calculated, based on
assumptions and values that may be considered reasonably
typical for the industry in general. All of these costs can
be expected to vary, depending on various factors such as the
particular company's accounting policies and financial situation
at that time, market conditions, space limitations, geographical
location, and particular mill operating considerations. All
costs are based on January 1, 1969, prices.
Utility costs and chemical costs which have been used are
tabulated on page 5-32.
Costs for control methods are calculated for significant
sources of emissions. Where older equipment exists, con-
trol methods are considered to reduce that particular
source to the minimum emission based on current technology.
If a further reduction in emissions may be necessary,
it is possible that new equipment designs would have to be
installed to replace the existing equipment. However, the
consideration of replacing existing process equipment with
new equipment is beyond the scope of this study; with the
exception of New Recovery Systems included in Section
5.3.6.1.6.
Capital Cost
Capital cost has been calculated for each control method based
on a typical arrangement as shown in this section. The capital
cost has been calculated on an arrangement designed to mini-
mize installation downtime. A description of the "Capital Cost"
is included on page 5-29.
"Loss of Production" cost has not been included since this will
vary from mill to mill. However, "Loss of Production" cost may
be very significant and must be included for individual mill
applications. For example, a two-week shutdown at a 500-ton
mill might result in the following loss to the mill:
Loss of Profit:
$20/Ton (profit after taxes)
x 500 T/D x 14 days = $140,000
Loss of income to pay for con-
tinuing expense of general
overhead and administration—
approximately the same as
Loss of Profit = 140,000
Approximate Loss to the Mill $280,000
5-27
-------
The above calculation does not include 'production personnel
wages, which may also be a "Loss of Production" cost.
Annual Cost
Annual cost has also been calculated in accordance with the
description on pages 5-30 and 5-31.
As described under "Depreciation and Interest (Capital Recovery) "
page 5-31,'an interest rate of 10 percent has been used. After the
total annual cost is calculated, applicable credits are
deducted, resulting in a net annual cost or savings. This
net annual cost or savings is then plotted on a curve along
with the capital cost. In calculating the net annual cost,
income tax credits, and investment tax credits have not
been included.
Control Methods
Control methods to be considered for evaluation are discussed
under each source. Where preliminary engineering evaluations
and cost calculations indicate that a particular control
method is not applicable, this particular control method
has not been investigated further. However, the results of
these evaluations and cost calculations will be reported for
the particular control method.
Sketches and Flow Schematics
Where appropriate, sketches and flow schematics are included
for a control method to better describe the basis for the
calculations of "Capital Cost" and "Net Annual Cost."
Cost Curves
Cost curves are included for control methods which have been
evaluated as being reasonably effective in reducing air
emissions for an individual source. The cost curves—"Capital
Cost" and "Net Annual Cost"—have been plotted versus a function
of size, tonnage, or CFM.
Detailed cost back-up sheets were prepared. However, due to the
voluminous quantity of this data, these detailed sheets are not
included with this report.
5-28
-------
DESCRIPTION OF CAPITAL COST ITEMS
Purchased Equipment
Cost of the emission control device and all accessories and
auxiliaries required for its full operation.
'Equipment Erection
The erection of purchased equipment only. The erection
cost of building and foundations, piping and wiring is
included elsewhere.
Equipment Foundations and Building
The installed cost of all equipment foundations, building
and building foundations, floors, roof, stairs, and walkways,
that are required for support, access, and enclosure of the
control device.
Process and Instrument Piping
Cost of all pipe and pipe supports, erected, and including
insulation and protective coating where required.
Power Wiring and Lighting
The installed cost of all power wiring and lighting is
included. Wherever required, any substation, transformers,
or switchgear costs are included under "Purchased Equipment."
Indirect Capital Cost
The summation of the above capital cost composes the total
direct capital cost to which is added indirect cost as a
percentage of direct cost. The following is a breakdown of
this indirect cost:
15% Contingency—Unforeseen conditions and
items not practical to estimate
7% Engineering—Preparation of specifications
and working drawings, selection,
and evaluation of equipment
1% General Construction Overhead—Includes tempo-
rary facilities, contractual
supervision, timekeeping, et cetera
3% <. Start-up Cost—Loss of production not included
2% Spare Parts
2% Sales Tax
30% Total
5-29
-------
DESCRIPTION OF ANNUAL COST ITEMS
General
Any estimated annual cost may vary considerably from
mill to mill. The following values may be considered
typical and should provide reasonable comparisons of
emission control device annual cost:
Direct Operating Cost
Direct operating cost consists of charges for:
1. Operating Labor, including Overhead. A
national average hourly rate of $4.25 has
been assumed. To this rate, an additional
29 percent is added for vacations, sick pay,
holidays, payroll taxes, insurance, and
fringe benefits. (Total $5.50 per hour)
2. Power, Electric, and/or Steam. An electric
cost of $0.01 per KWH, and a steam cost of
$0.65 per 1000 Ibs. of steam has been used.
3. Water. A water cost of $0.13 per 1000 gallons
has been used.
4. Maintenance, including maintenance labor, re-
placement-,'parts, .and maintenance materials.
Taxes and Insurance
Local property taxes vary widely over the country. Generally,
this is applied as a millage on the assessed valuation.
Assuming the property is assessed at 50 percent and 30 mills
tax is applied, then the property tax would be one and one-half
percent of the construction cost.
1.5% of Capital Cost
An average insurance cost has been
used for the entire pulp and paper
mill.
0.5% of Capital Cost
Total Taxes and Insurance 2.0% of Capital Cost
5-30
-------
Administrative Costs
This is an average cost applied to the overall plant and
includes all salaried personnel (officers and supervisors),
fringe benefits for salaried personnel, legal and other
professional services, public relations, contributions, and
office supplies and expenses. This does not include any mar-
keting costs.
5.0% of Capital Cost
Depreciation and Interest (Capital Recovery)
Depreciation plus interest charges were calculated from the
following formula which is one of several commonly used for
this purpose. An interest rate of 10 percent was assumed.*
i * IT, 4- i(l + i)n x 100%
Equal Annual Payment =
i = interest rate
n = life in years
SUMMARY OF CAPITAL CHARGE ITEMS
; (As % of Capital Cost)
Estimated Life of Equipment in Years
Item 8^ 10 16
Taxes & Insurance 2% 2% 2%
Administration 5% 5% 5%
Depreciation &
Interest 18.7% 16.3% 12.8%
(Capital Recovery)
TOTAL 25.7% 23.3% 19.8%
*Much of the industry uses a rate of nearly 20 percent
before taxes which reflects the fact that an investment
in air pollution control equipment is a lost opportunity
for profit compared with the alternative course of
investing the funds in productive equipment which would
produce income equal to the average current rate of return
or unamortized investment for the industry (18).
5-31
-------
UTILITY COST
Operating Costs: Total
Item
Steam $/1000 Ibs.
Based on 1000 BTU/lb.
Electricity
$/KWH
Water
$/1000 gal.
Kraft Waste Treatment
$/1000 gal.
(Primary and Secondary
Treatment)
Fixed
Charges
$0.116
0.0023
0.104
0.13
Fuel, Labor, Maint.
Etc.
$0.535
0.0077
0.046
0.03
(Rounded to
Nearest Cent)
$0.650
0.010
0.150
0.160
These are values that may be considered typical in a general sense. All of these
values vary considerably from mill to mill—any proper evaluation of control costs
should be specifically calculated for the individual mill. However, the above
costs should provide reasonable comparisons of emission control method for this
study.
Salt Cake
Lime (CaO)
Sulfur
Soda Ash
Caustic Soda
Magnesium Hydroxide
Chlorine (Papermakers)
CHEMICAL COSTS
Per Ton Except as Noted
$34.00 East
$24.50 West
$15.00
Average: $30.00
Average: $15.00
$39.00 to
$42.00 per Long Ton Average: $40.00
$31.00 to
$32.00
Average: $31.00
$57.00 per ton of NaOH
in a 50% Solution
$37.68 per ton 100% Solids—
Tank Car Lots, FOB, Michigan
$ 3.35 per 100 Ibs.
5-32
-------
5.3.5 LIMITATIONS TO APPLYING COST CALCULATIONS
The design, arrangement, and operation of pulp mills demon-
strate as many variations as there are mills. The applica-
tion of control methods to mills is equally as varied as the
mills themsslves. In order to determine a source "arrangement"
which might be considered typical, a survey of all configura-
tions at all mills would have to be completed. Rather than
attempt this magnitude of survey and data compilation, the
experience and judgment of the contractor was employed to
assume arrangements which might be considered reasonably typical.
These arrangements are the basis for computing cost calculations
for control methods applied to the various sources.
The arrangements and cost calculations are considered to be
reasonably typical of pulp mills in the United States. Appli-
cation of these costs to an area smaller than the United States
is not intended. The costs are considered to be accurate within
+ 20 percent over the entire United States. However, if these
costs were to be used for one mill only, the variation might be
from 70 to 200 percent of the U. S. average. For individual
applications, specific and detailed costs must be calculated.
For instance, the application of a stainless steel precipitator
in a relatively inaccessible location would more than double the
precipitator cost which has been used to calculate the average
U. S. cost involving a precipitator. Therefore, these cost
calculations must be used judiciously.
5.3.6 KRAFT SOURCES
5!.3.6.1 Recovery Systems
The design and operation of recovery systems and auxiliary
equipment has been described in detail in numerous publi-
cations (2_, 3_, 4_, 5_) . This discussion is primarily limited
to factors influencing air emissions from recovery systems.
Design Considerations
Kraft pulping operations yield approximately 2,000 to
3,500 pounds of dry black liquor solids per ton of pulp
5-33
-------
produced. Heating values range from approximately
5,000 to 7,200 BTU's per pound of dry solids. Three
thousand pounds of solids at a heating value of 6,600
BTU's per pound is generally regarded as an average.
The actual size of a recovery unit is based on the
heat input (solids flow times heating value). Re-
covery unit rating is defined as the manufacturer's
guaranteed heat input to the recovery furnace. There-
fore, percent overload or percent of rated capacity is
normally a percentage of the manufacturer's guaranteed
heat input. For ease of reference and discussion,
recovery unit operators often use the manufacturer's
guaranteed solids flow to the recovery system as the
rated capacity. This type of reference may be satis-
factory for general discussions, but the guaranteed
heat input to the unit is the precise definition of
rated capacity.
Steam pressure and steam temperature have increased
over the years as required by economics relating to
the steam's power balance. Present pressures vary
from 300 to 1500 psi. Temperatures have also followed
turbine practice with present units as high as 925°F.
Furnace
The furnace or combustion chamber is one of the most
important components in the recovery system from an
air emission standpoint. The furnace performs the
following functions:
1. Dehydrating the black liquor and burning the
organic constituents with maximum combustion
efficiency.
2. Reducing that portion of the chemicals present
as sodium sulfate and other sodium-sulfur-oxygen
compounds to sodium sulfide.
3. Melting the inorganic chemicals for continuous
removal in molten form.
4. Conditioning the products of combustion to reduce
chemical carryover and flue gas temperature.
Black liquor is introduced to the furnace by one or more
sprays located in the furnace walls. The fundamental
5-34
-------
difference between the furnaces of the two major
United States manufacturers is the method of de-
hydrating the black liquor.
In one design of furnace, heavy black liquor
is sprayed and deposited in a partially dehydrated
form in a band on the furnace walls where it remains
until dry before falling to the hearth to form the
char bed. The spray consists of one or two nozzles
oscillated in a "paintbrush" motion to obtain uni-
form distribution of liquor on the walls. Usually,
the spray is located in the front wall, although
larger units also have additional sprays in the
other furnace walls.
The other furnace design introduces liquor
through a multiplicity of sprays located in two or
more walls so that the droplets dry in suspension.
The sprays are oscillated in a vertical plane moving
from horizontal to an adjustable downward angle. A
very coarse spray is maintained so that, in falling
from the nozzle to the hearth, the particles of
liquor are flash dried. Most units use nozzles in
at least two walls, with larger ones having sprays
in all four walls.
Combustion of the black liquor char begins on the
hearth of the furnace. Air for combustion is sup-
plied by a forced draft system to the reducing and
oxidizing zone of the furnace. Since a reducing
atmosphere is required to convert sodium sulfate and
other sodium-base sulfur compounds to sodium sulfide,
only a portion of the air required for complete com-
bustipn is supplied to the char bed through the lower or
primary air ports. The heat released by the combustion
in this zone is sufficient to liquefy the chemicals in
the char and to sustain the endothermic reduction. The
liquefied chemical, or molten smelt, is continuously
drained from the hearth.
Air is admitted above the primary zone to complete the
combustion of the volatile gases from the char on the
hearth. The final stage of combustion on the "paintbrush"
spray system is supported by a series of ports, at
secondary and tertiary levels, which direct the air across
the furnace, whereas this is accomplished in the other
furnace design by air admitted through a vertical bank of
ports tangentially directed into the gases.
5-35
-------
The control of total air in relation to the
black liquor flow and the proper distribution
of air is essential for efficient combustion,
chemical reduction, and minimum emissions. A
number of tests has been conducted and published
indicating that approximately 10 to 15 percent
excess air is essential to obtain complete,
efficient combustion and minimum air emission.
Tests (6_) have been conducted indicating the
importance of proper distribution of primary
and secondary air to the furnace.
When the furnace is operated to provide for mini-
mum emissions (primarily minimum H S) to the atmos-
phere, the most complete combustion conditions also
exist. While not proven completely at the present
time, it appears that SO in the flue gas is also
low when the H_S is a minimum.
Apparently, best furnace operation from the stand-
point of efficient, complete combustion and chemical
reduction results in minimum emission of pollutants
from the furnace.
Direct-Contact Evaporator
Black liquor from the multiple-effect evaporators
at 45 - 55 percent solids is not sufficiently con-
centrated to permit firing directly into the furnace.
To raise the concentration to a level which will en-
sure stable combustion, present U. S. practice requires
the removal of additional water by means of a cascade
evaporator, a cyclone evaporator, or a Venturi evapo-
rator-scrubber. In practice, the liquor is concentrated
to the range of 60 - 70 percent solids before firing.
The direct contact evaporator has been identified as a
major source of H S emission in the recovery system.
The amount of H S emission from the evaporator varies
depending primarily upon the sodium sulfide concentra-
tion in the liquor, the pH of the liquor, and the
degree of gas-liquor contact, and the amount of H S
emissions coming from the recovery furnace.
5-36
-------
Corrective action for the prevention of H S emission
from the direct contact evaporator is either oxidation
of the liquor or conversion to one of the new recovery
systems which eliminate direct contact between the flue
gases and black liquor (See Chapter 6, New Developments.)
The new systems (See Section 6.2.1.1) concentrate the
liquor to the firing range of 60 - 70 percent by the use
of additional evaporation in the multiple effect evapo-
rators (up to approximately 55 percent), and finishing
in an air cascade evaporator (from 55 percent to 65 - 70
percent), or by the use of an additional multiple evapo-
rator effect (commonly called a concentrator) for a
resultant liquor concentration of approximately 62 percent.
Electrostatic Precipitators
The electrostatic precipitator is a very important part of
the recovery system. The cost of installation and operation
of a precipitator is offset by the recovery of chemicals for
all but the highest collection efficiencies. With present
emphasis on pollution control, however, the collecting
efficiencies are much higher than would be justified from a
recovery economics standpoint.
The gases emanating from a recovery boiler are essentially
flue gases with1 a small amount of sulfur compounds and
organic sulfur compounds. The particulate matter consists
primarily of sodium sulfate, or in case of a soda mill, of
mostly sodium carbonate. Originating from the high tempera-
ture combustion zone, this material is sublimed fumes and of
extremely small particle size. It is very light and has an
average bulk weight of only nine pounds per cubic foot when
precipitated in dry form.
The gaseous constituents, the high water vapor content, and
the characteristics of the chemicals recovered in the precipi-
tator present serious corrosion conditions unless the equip-
ment is properly designed. The temperature of the gas leaving
the cascade or cyclone evaporator is normally 300 - 350°F. At
temperatures below 300°F, local cold spots may cause condensa-
tion and create severe corrosion problems.
5-37
-------
To provide greater flexibility in operation, the
precipitator is often designed with two identical
units in parallel. This arrangement makes it
possible to shut down one chamber for maintenance
or inspection and still maintain reasonable cleaning
efficiency while routing the entire gas flow through
the other chamber.
There are many materials of construction used for
the shell of the precipitator such as mild steel,
reinforced concrete, hollow or filled tile, or combi-
nations of these materials. As indicated above, it
is of great importance that local condensation be
prevented by proper insulation or correct gas tempera-
ture. The material of the shell should, therefore, be
selected with this in mind. Mild steel will have a
short life under other than ideal conditions. This
material is, of course, attractive because of its
relatively low cost for manufacturing as well as
erection. It is a well-known fact skilled welders
and boiler makers are more readily available than tile
setters. A steel construction precipitator is also
relatively, light weight, which is an advantage from
the standpoint of foundation, structural steel, and
erection. In order to extend the life of the steel
shell, some corrosion resistant coatings have been
used with reportedly good success. It is, however,
of utmost importance that the steel plate be covered
completely. A very minute hole in the coating may be
the starting point for corrosion under the coating.
If a reasonable life is to be expected from a steel
shell precipitator, it must be provided with good
insulation. This insulation may, for example consist
of a blanket insulation outside the shell followed by
an air space and another blanket insulation. The
exterior is then usually covered by aluminum panels.
Hoppers or wet bottoms are usually insulated with
blankets without intermediate air space.
Concrete has somewhat better corrosion resistance
than steel and is often used for roof sections. For
outer walls of the shell where the temperature tends
to be lowest, concrete is normally used only in loca-
tions where suitable tile or skilled tile setters are
not available. Some years ago, great claims were
5-38
-------
made for so-called hollow tile due to its superior thermal
insulating capacity. Many cases of seepage of salt
cake into the hollow spaces of the tile have been
reported and filled tile is now generally considered
a more practical choice.
The bottom of the precipitator can be designed in many
different ways to meet various operating conditions,
plant requirements, and mill preferences. There are
two basic designs of precipitator bottoms—dry and wet.
The dry bottom can be constructed as a flat bottom
collecting chamber underneath the precipitator zone.
From there the collected material is removed by means
of a drag scraper and screw conveyer and finally dis-
charged into a rotary valve. Trough hoppers with
screw conveyers and rotary discharge valves are pre-
ferred by some mills. The dry bottom may be of some
advantage when used in conjunction with recovery
boilers where the flue gas must not get in contact
with the black liquor.
The wet bottom design allows collected material to
be brought back immediately and continuously into the
black liquor flow to the direct contact evaporator.
The liquid surface is located below the precipitator
electrodes at sufficient distance to prevent arcing.
Dust falling down into the liquor is kept in suspension
by motor driven agitators or pump mixers. Contact
between unoxidized black liquor and the flue gases may
create a source of H S emission.
Sneak-by of gas in wet or dry bottom precipitators is
minimized by means of fixed baffles extending from the
precipitation zone down to the bottom.
Possible uneven distribution of the gas flow in the
inlet end of the precipitator is corrected by perforated
distribution plates or adjustable baffles. Sometimes
the conditions call for vibrators or rapping mechanisms
to keep the baffles or the distribution plates clean.
There are different designs of supports for the emitting
electrode system. Some designs are not exposed to dust
build-up in the support insulators. Build-up on the
inside of an insulator can lead to arc-over and reduction
of collecting efficiency and eventually to cracking of the
5-39
-------
insulator which in turn may require a precipitator shut-
down. Support insulators fabricated from alumina have
proven themselves to withstand substantial amounts of
arc-over without cracking. Arcing is easily detectable
from instruments in the control cabinet. At such
occurrences, it is necessary to shut down the power in
the precipitator long enough to allow cleaning of the
insulators before the precipitator can be put into
operation again. Under certain difficult conditions,
it is sometimes necessary to blow hot air into the
support insulators to prevent the excessive dust build-up.
The high voltage power units should be located as close
as practical to the high voltage connections of the pre-
cipitator in order to shorten the length of the high voltage
conductor. It is, therefore, common to locate the trans-
former and the rectifier unit directly on the precipitator
roof. However, it is recommended that the control cabinet
be located in the main control room, thus enabling the
operators to keep close control of the operation of the
precipitator. Another advantage is, of course, that the
control cabinet is more protected in the control room than
out in the mill.
An estimated 75 percent of all new installations of electro-
static precipitators are atop the recovery boiler. This
preserves valuable ground space for future expansion of the
boiler. The location of the precipitator on top of the
boiler also reduces the required stack height and may also
possibly reduce the draft loss. An important advantage is
the reduction of horizontal duct runs which are subjected to
dust fall-out with subsequent maintenance problems.
Fiberglass reinforced plastic is becoming a popular material
of construction for ductwork and stacks because of its ability
to withstand corrosion. This material is also lightweight and
easy to erect.
The I.D. fan is usually located ahead of the precipitator. The
reason for this is that for roof-mounted precipitators, which
is the most modern installation, the duct work will be less
complicated and less expensive. The fan is, however, more
exposed to erosion and dust build-up in this location. In
order to overcome the build-up, the fan is normally equipped
with soot blowers, which automatically provide for 10 - 15
minute cleaning periods at certain intervals. By having
5-40
-------
the fan before the precipitator, the dust generated during
the steam cleaning can be captured. The disadvantage of
this fan location is, however, that the precipitator operates
under positive pressure and leakage can easily occur around
inspection doors and flange connections. On the other hand,
when the precipitator operates under negative pressure, cold
air can leak in and cause cold spots and local corrosion.
With the high efficiencies of modern precipitators, there is
normally no problem with build-up on fans located subsequent
to precipitators and steam blowers are normally not necessary.
The steam may, under certain conditions, cause corrosion.
Rebuilding Existing Precipitator
Sometimes the question is brought up of how much the efficiency
can be upgraded of an existing precipitator, if the internals
and the electric system be completely replaced in the existing
shell„ This approach has to be evaluated for each individual
case, but is seldom feasible due to the following reasons:
The collection efficiency attainable will be slightly higher
than the original guarantee efficiency, which would not be in
compliance with today's air pollution codes. Addition of a
new precipitator in series with the existing one would bring
the overall collection efficiency to an acceptable level with
higher equipment cost,, The down time, which in the first
case would be considerable, would be kept at a minimum with
the add-on concept. The overall cost would most certainly be
lower for the add-on alternative.
Scrubber Installed Subsequent to Precipitator
In order to improve the collecting efficiency of an electro-
static precipitator or to prevent "snow-outs," a scrubber is
sometimes installed subsequent to the precipitator. The snow-
outs sometimes occur when the gas velocity through the precipi-
tator is too high (overload conditions). Under such conditions,
the dust re-entrains into the high gas velocity. Too frequent
or vigorous rapping can also cause snow-outs. Despite very
high efficiencies, some dust will always escape'the precipitator.
This dust has a tendency to accumulate on the inside walls' of the
stack and will eventually be torn off-in-large flakes and thrown
out the stack.
There are two reasons for installing a scrubber subsequent to the
precipitator:
5-41
-------
1. The combination of precipitator/scrubber may, in
some cases, give a more economic recovery unit
than a precipitator alone for the same overall
efficiency.
2. The scrubber (normally of low energy type) will
essentially eliminate the snow-out problems at
an overloaded or deteriorated precipitator.'
A bypass duct is sometimes installed around the precipi-
tator so that repair and maintenance on the precipitator
can be done without boiler shut-down. The scrubber will
give the same or better collecting efficiency when the
precipitator is bypassed. The overall efficiency will,
of course, decrease.
Often corrosion is a great problem at these scrubbers.
The gas will, in most cases, reach saturation and, due
to the presence of sulfur dioxide, carbon steel construc-
tion is normally precluded. In some cases where the
timber has been floated in salt water, the sodium chloride
content in the black liquor is rather high. This has to
be given special consideration when selecting materials of
construction.
Other potential disadvantages with wet scrubbers are
associated with the steam plume created at the stack dis-
charge. Such wet plumes are quite visible and may result
in complaints from neighbors. The wet plume may also
result in a lower plume rise which may adversely affect
dispersion of the stack gases.
Low Energy Scrubbers. The low energy scrubbers are
normally installed directly behind the electrostatic
precipitator. The cyclonic-scrubber is the most common
type. Very often the scrubber is furnished with a flat
bottom to facilitate the installation. .
The scrubbing liquid is recirculated, but the solids content
is kept below 10 percent by maintaining a proper bleed-off.
At higher concentrations, the spray nozzles have a tendency
to plug.
Water is most commonly used for the scrubbing medium. Addition
of caustic will increase the absorption of SO , while weak wash
may increase the emission of H S and other sulfur compounds.
5-42
-------
The vessel must be corrosion resistant. In :
addition to manufacturing it from stainless steel and
fiberglass, lined carbon steel is frequently used. The
lining is normally a 1 to 2 inch concrete lining type
Gunite or similar material.
The stack also has to be protected against corrosion and
since a concrete lining would make the stack too heavy
and necessitate reinforcing of the scrubber vessel, other
means for corrosion protection have to be found. Thus, the
stack is often made from carbon steel with a stainless
steel sheet lining of 1/16 inch thickness. Sometimes
the stack is manufactured-from polyester"fiberglass, which
has proven to give excellent life and low maintenance. For
mills with high sodium chloride content in the black liquor,
the whole scrubber should be made in fiberglass.
The pressure drop over the scrubber is normally 1 to 3
inches WG and very often due to the large gas volumes, two
scrubbers in parallel have to be installed. The collecting
efficiency of a cyclonic scrubber for this application is
80 - 90 percent on particulate matter.
The fan is normally located ahead of the scrubber making it
possible to have the stack mounted directly on top of the
scrubber.
The recirculation pump is normally made from corrosion
resistant material.
High Energy Scrubbers. When a higher collecting efficiency
is required than is possible to obtain in a cyclonic scrubber,
a high energy scrubber must be installed. The scrubber is
almost exclusively of the Venturi type.
The scrubbing liquid should be water for the same reasons
previously cited. Addition of caustic or alkaline chemicals
will, of course, increase the absorption of SO . The use of
weak wash may increase the emission of H S and other sulfur
compounds. Fully oxidized black liquor can be used providing
the liquor has low foaming characteristics.
The solids content in the recirculation liquid is normally
kept below 30 percent. At higher solids content, the scrub-
bing efficiency will decrease at the same pressure drop.
In other words, more energy is needed to atomize the scrubbing
liquid.
5-43
-------
The scrubber has to be corrosion resistant and
frequently is manufactured in stainless steel.
The pressure drop over the scrubber is between
15 and 30 inches WG, depending on the desired
collecting efficiency. Efficiencies in excess of
99 percent are attainable in this pressure drop
range.
The fan is normally located between the scrubber
and the stack. Due to the corrosive atmosphere,
the fan has to be constructed of stainless steel,
be provided with split housing for easy wheel
removal, and have inspection doors, water sprays,
and drain connections.
The recirculation pump is normally made from
corrosion resistant material.
Venturi Evaporator-Scrubber. The Venturi evaporator-
scrubber is a device which both collects dust and fume
from the flue gas and provides evaporation of the black
liquor. By nature, the Venturi creates intimate contact
between the flue gas and the scrubbing medium thereby
causing almost instantaneous saturation of the
gas and at the same time evaporation of the
scrubbing medium. If the scrubbing medium is
black liquor, the direct contact evaporator and
the fume collector can both be replaced by this
evaporator-scrubber. There are two types of
this device—the single stage and the two^stage
Venturi evaporator-scrubber.
The first type has been used on approximately thirty
recovery boilers over the years, while the
latter type is a rather new development, and
experience is rather limited (only one or two instal-
lations) .
Both types offer some advantages. Space require-
ment and equipment cost are lower than for any
other combination of evaporator and fume collector.
The Venturi evaporator-scrubber provides higher
thermal efficiency than a conventional evaporator.
The temperature of the flue gas leaving a one-stage
5-44
-------
Venturi is ISO - I90CF and from a two-stage, 160 -
170°F. In Scandinavian countries, a separate cooling
loop is sometimes installed to reduce the steam plume
and produce warm water. The flue gas leaving such
cooling loops can be brought down to 100 - 110°F.
There are, however, some serious disadvantages with
these systems. The power consumption is much higher
than for a conventional system. The corrosion problem
is often severe but can, of course, be overcome by
proper selection of materials of construction. The
advantage with low equipment cost will, however, be
reduced if too sophisticated a material is chosen.
The single-stage system evaporates the liquor and
scrubs the flue gas in one Venturi. The entire black
liquor flow received from the multiple effect evaporators
is recycled to attain 60 - 65 percent solids. The
bleed-off from the system is fed to the recovery boiler
to be burned. Due to the high concentration of solids
in the liquor, the nozzles in the Venturi throat have a
tendency to plug and the liquor is difficult to atomize.
To overcome these problems, steam is fed with the liquor
into the throat„ This steam helps atomize the liquor and
keep the nozzles operu
The Venturi which normally is used has a rec-
tangular throat with maximum width of about
twelve inches. At greater widths, the sprays cannot
penetrate into the center of the gas stream and the gas
will escape agglomeration with loss of efficiency as a
result. Due to this limitation of width, these Venturi
throats tend to be extremely long at large gas volumes.
The inlet transition from the boiler to the Venturi is
space consuming and difficult to erect.
The pressure drop required to atomize the liquor
is increased at higher solids concentration.
The scrubbing efficiency drops off drastically
as the liquor concentration goes up. For a given
pressure drop, water as scrubbing medium will give
99 percent efficiency. At the same pressure drop,
45 percent black liquor will give 94 percent and
60 percent black liquor can be brought up to 85
percent efficiency.
The pressure drop necessary to attain good evapora-
tion is approximately 5-6 inches WG, but in order
to get 95 percent scrubbing efficiency, a pressure
drop of about 30 inches WG is necessary.
5-45
-------
The schematic arrangement of this system is shown
in Figure 5-9.
In the two-stage approach, (See Figure 5-10), some
of the disadvantages with the one-stage system are
overcome. This system uses a so-called SF Venturi,
as described in Section 5.2. For large gas volumes,
a modified version with an annular throat is used.
The first stage has a low pressure drop, 5-6 inches WG,
and is used exclusively as an evaporator. The total
flow of black liquor from the multiple effect evapora-
tors is introduced into this stage and is recycled until
the liquor has attained 60 - 65 percent solids. The
bleed-off is pumped to the recovery furnace to be burned.
Some dust collection is achieved in this stage, but
hardly more than in a conventional cascade or cyclone
evaporator. In order to prevent build-up of black
liquor solids, which would eventually impede the operation,
the walls of the first stage separator are flushed with
black liquor from the recycle pump. The elbow connecting
the Venturi and the separator is of regular and not of
the so-called flooded type. The buildup of solids is, of
course, a fire hazard too.
When the gas enters the second stage, it is close to
saturation and has a temperature of approximately 190°F.
The pressure drop in this Venturi is about 30 - 35 inches
WG depending on the desired collection efficiency.
(approximately 90 - 95%) . This pressure drop necessitates
a high velocity. The elbow connecting the Venturi and the
separator is of the flooded type to prevent excessive
erosion. The scrubbing liquid is recycled to
between 5 and 20 percent solids concentration.
The bleed-off is fed into the first stage and
collected dust is thereby automatically returned
to the process. In this stage, little evapora-
tion takes place, but the gas leaving the scrubber
has a temperature of approximately 160 - 170°F and
is completely saturated. The make-up can be black
liquor, but excessive foaming normally precludes the
use of liquor. Addition of caustic will, of course,
increase the absorption of SO , while white liquor
or weak wash will increase the emission of H S and
other sulfur compounds. Unoxidized black liquor will
also increase emission of sulfur compounds. Water is
normally the best make-up and since the amount added
5-46
-------
FLUE GAS
FROM BOILER
CLEAN GAS OUTLET TO FAN
WALL
WASH
CYCLONIC SEPARATOR
iRECYCLE LIQUOR
60-70% SOLIDS
190°F
LIQUOR TO
BOILER
60-70% SOLIDS
LIQUOR FROM
MULTIPLE EFFECT
EVAPORATOR
45-50% SOLIDS
FIGURE 5-9
TYPICAL FLOW SHEET FOR SINGLE STAGE
VENTURI EVAPORATOR/SCRUBBER SYSTEM
FOR RECOVERY BOILER
5-47
-------
--.CLEAN GAS TO FAN
GAS FROM
BOILER
en
i
co
WALL WASH
AND/OR BLEED
RECYCLE
5-15% SOLID
165°F
FALSE
BOTTOM
ADDITIONAL
BLEED
LIQUOR FROM
RECYCLE
i60-70% SOLIE
;190°F
LIQUOR TO
BOILER
60-70% SOLID
MAKE-UP
TER
FIGURE 5-10
TYPICAL FLOW SHEET FOR TWO (2) STAGE
VENTURI EVAPORATOR SCRUBBER SYSTEM
-------
is rather small, it does not affect the overall heat
balance, but helps to reduce the steam plume. No wall
wash is necessary in the separator of the second stage.
This type of Venturi, used in both the first and the
second stage, does not have any nozzles, weirs or
narrow restrictions, which can plug up or cause dust
build-ups. No steam atomization is, therefore,
necessary. Since the solids concentration in the
recycle liquid of the second stage is low, a higher
collection efficiency can be attained for the same
total pressure drop as compared with the one-stage
evaporator-scrubber. The two-stage system also has
a higher thermal efficiency than the single stage
system.
For both types of systems, the scrubbers have to be
corrosion resistant.
The fan is normally located between the scrubber
and the stack. Due to the corrosive atmosphere,
the fan has to be constructed in stainless steel,
be provided with split housing for easy wheel
removal, and have inspection doors, water sprays
and drain connections. Ductwork is normally made
from stainless steel.
The recirculation pump is normally made from
corrosion resistant material.
Recovery Furnace Scrubbing Liquors. The liquid scrub-
bing systems which have been described depend to a
considerable extent for their effectiveness on selection
of a suitable liquid phase. In some instances, the scrubber
may remove both particulates and gaseous compounds from the
gas stream. The use of some liquids may actually add to
the concentration of odorous compounds in the flue gases.
Selection of a scrubbing liquor for removing
particulate matter and odoriferous gas components
from recovery boiler flue gases is a difficult task
because the gas volume is large and the CO content
is very high resulting in the formation of H S in
the presence of Na S as follows:
5-49
-------
H2° + C02 t H2C03
H2C°3 + Na2S t Na2C°3
Therefore, if either weak wash from the causticizing
operation or weak black liquor is used as the scrub-
bing liquor, it must first be oxidized to change the
Na S content to Na S O . However, Harding and
Galeano {(7_) have shown that above an oxidation efficiency
of 80 percent, the absorption of odoriferous gases de-
creases because the NaOH originally formed by the oxi-
dation reaction is destroyed by further reaction with
O and Na S O as shown below:
2Na S + 2O + HO -> Na S 0 + 2NaOH
Na S O + 2NaOH + 2O -»• 2Na SO + HO
After studying the above reactions, the successful
use of Na S containing liquors is doubtful since
the presence of Na S in partially oxidized liquor
results in formation of additional H S, while the
use of 99 percent plus oxidized liquor results in
greatly reduced H S absorption efficiency.
Water has been moderately successful in removing
relatively large particulate matter in low pressure
drop scrubbers following electrostatic precipitators.
More efficient particulate removal can be obtained
with water if high pressure drop Venturis are used.
However, the efficiency of water in scrubbing odor-
iferous gases is very low regardless of the equip-
ment used.
In summary, technology has yet to produce a tried
and proven system for simultaneously eliminating
both gases and particulates from recovery furnace
flue gases with any of the scrubbing liquors available
for use. More research pilot plant data and opera-
ting experience with systems using different types
of scrubbing liquors must be done before any valid
conclusions can be reached in the selection of the
most feasible scrubbing liquor.
5-50
-------
Case Descriptions
The following cases have been considered for application
of control methods to recovery systems.
Case 1. Replacement of an existing 90 percent AOE
precipitator with a higher efficiency precipitator.
Case 2. Installation of additional control equipment
in series with an existing 90 percent AOE precipitator.
Case 3. Installation of a cyclonic scrubber in series
with an existing 95 percent precipitator located on the
roof.
Case 4. Application of additional control methods to
an existing recovery system with an 80 percent AOE
black liquor Venturi.
Case 5. Installation of a black liquor oxidation system
to an existing recovery system.
Case 6. Capital Costs of new recovery systems.
5-51
-------
5,3.6.1.1 Recovery System, Case 1
Application
This case is based on the replacement of an existing
90 percent AOE precipitator with a higher efficiency
precipitator. Most of the existing older recovery
precipitator installations are arranged with the
precipitator on the ground. Usually there is no
additional space on the ground in the vicinity of the
recovery unit; therefore, an additional precipitator
would have to be installed above the existing precipi-
tator. Further, the existing precipitator could not
be removed and replaced with a new one since this would
require a lengthy mill shutdown resulting in a high
dollar cost due to the loss of production.
Costs
The capital costs are based on a couble chamber, common
wall, tile construction precipitator arranged as shown
in Figure 5-11. Auxiliaries included are: agitators,
dampers, circulation pumps, instruments and controls
for proper operation of the above, and revised ductwork
to connect the addition of replacement precipitator in
the system. The cost also reflects the structure re-
quired to support this addition, and the demolition
cost for removal of the existing precipitator. Capital
costs and net annual costs are presented in Figures 5-12,
5-13, and 5-14.
Effectiveness
Particulate removal. Methods are considered for re-
placing a 90 percent AOE precipitator on an existing
recovery boiler. These methods and their particulate
efficiencies are as follows:
Annual
Guaranteed Operating
Control Method Efficiency Efficiency
Precipitator 99.9 99.5
Precipitator 99.5 99.0
Precipitator 99.0 98.5
5-52
-------
U1
I
m -yo t— i
o m en
o o •
< •— < in
m no ^
30 t-i .
m pa -<
30 m -o
-a i— <
i— o
CD 73
m m
X 2
i— i m
CO 3
z -n
CD O
79
-a
•jo m
m i—
<-> m
•-i r>
T3 -H
I O
co
I
m
>
m
CO
EXISTING
PRECIPITATOR
REMOVED
-------
c? 210
8
o
-' 180
x
o
o
150
120
90
60
50
o"
CD
°. 40
o
X
** 30
oo
S 20
-------
g 175
o
•k
o
150
125
100
so
50
§ 40
30
s 20
10
10 20 30 40
GAS VOLUME (CFM X 10,000)
456 7 8 9 10 11
ADT/DAY X 100
BASED ON 350 CFM/ADT/DAY
FIG. 5 - 13 CONTROL METHOD COSTS FOR 99.5% EFF.
ELECTROSTATIC PRECIPITATOR REPLACING AN EXISTING
PRECIPITATOR - RECOVERY BOILER (99.0% A.O.E.)
5-55
-------
o
8
2 150
X
V*
~ 125
CO
o
" 100
£
h— 1
0.
5 75
£ rn
X
x
X
X
X
V
X
X
JO
o"
0
5 30
X
** 25
c/5
o
« 20
1
^ 15
i—
LU
T f\
/
X
x
^*
.A
X
X
X
F^
10 20 30 40
GAS VOLUME (CFM X 10,000)
345 6 7 8 9 10 11
ADT/DAY X 100
BASED ON 350 CFM/ADT/DAY
FIG. 5 - 14 CONTROL METHOD COSTS FOR 99.0% EFF.
ELECTROSTATIC PRECIPITATOR REPLACING AN EXISTING
PRECIPITATOR - RECOVERY BOILER (98.53! A.O.E.)
5-56
-------
The attached cost effectiveness curves (Figure 5-15) indicate
the change in costs as efficiency and size of equipment are
varied.
Reduced Sulfur and Sulfur Dioxide Removal. Since precipitators
do not remove either reduced sulfur or sulfur oxides, all three
methods would be equally ineffective on such emissions.
Operation. All three control methods should be equal in terms
of operation.
5-57
-------
G
O
•fc*
I 20
S 15
10
A
99,0
% A.O.E.
99.5
200
S 175
c
150
oo
S 125
99.5
A. 105,000 CPU (300 ADT/P.AY)
B. 175,000 CFM, (500 ACT/DAY)
C. 35C,"00 CFM, (1000 APT/DAY)
- AHT/DAY ?ASED OK 3BG CF
FIG.5-15PAPJICULATE COST EFFECTIVENESS
FOR ELECTROSTATIC PRECIPITATOR REPLACING
AN EXISTING PRECIPITATOR - RECOVERY BOILER
5-58
-------
5.3.6.1.2 Recovery System, Case 2
Application
This case is based on the installation of additional
control equipment in series with an existing 90 percent
AOE precipitator located on the ground. The control
methods considered are electrostatic precipitators,
Venturi scrubbers, and cyclone scrubbers.
Costs
Electrostatic Precipitator. This control method is
based on the installation of an electrostatic pre-
cipitator in series with and located above an exist-
ing precipitator as indicated in Figure 5-16. This
location was chosen as the likelihood of having room
for adding a second precipitator on the ground would
be practically nil in an average mill. The capital
cost and net annual cost shown in Figures 5-17, 5-18,
and 5-19 are based on a double chamber, common wall,
tile construction precipitator with a wet bottom and
include the auxiliaries such as agitators, dampers,
circulation pumps, instruments and controls for proper
operation of the above, and revised ductwork to con-
nect the additional precipitator to the existing
precipitator and to the existing stack. The costs
also reflect the structure required to support this
addition.
Venturi scrubber. This control method is based on the
installation of a 304 stainless steel Venturi and a
concrete lined mild steel separator in series with an
existing electrostatic precipitator as indicated in
Figure 5-20. The outlet duct and additional I.D. fan
are 304 stainless steel construction. Due to the ad-
dition of the Venturi, the pressure drop through the
exhaust gas system will be greatly increased requiring
additional fan pressure. This case is based on an
additional variable speed I.D. fan being installed
between the Venturi and the existing stack. Capital
costs and net annual costs are presented in Figures
5-21, 5-22, and 5-23.
5-59
-------
o
m
o
Oi
o
ui
I
m o\
5-,
rn -<
oo -o
s
1-1 3»
Co -Z.:
—I CD
O m
•X3 -H
7D
S3
S*
HI m
— \ o
O — I
•yo PO
o
I CO
70 >
rn —i
O i-t
o o
' JO
. m
o -o
73 —1
O
70
-------
*-* IHU
O
8
£120
X
*•»•
~100
oo
O
^ 80
i— i
5 60
_l
P Jin
>
X
X
/
x
^
.4
/
/
35
o
O
25
8 20
UJ
10
X
10 20 30 40
GAS VOLUME (CFM X 10,000)
345 6 789 10 11
ADT/DAY X 100
BASED ON 350 CFM/ADT/DAY
FIG. 5 - 17 CONTROL METHOD COSTS FOR 99% EFF.
ELECTROSTATIC PRECIPITATOR ADDED IN SERIES WITH
AN EXISTING PKECIPITATOR-RECOVERY BOILER (99.8X A.O.E.)
5-61
-------
g
o
X
JS
I—I
a.
i
120
100
80
60
20
-— * £.J
o
o
c^
2 20
X
•*»"»•
~ 15
t—
o
0
i in
_J IU
«c
<• s;
o
UJ
n
^
^
^
^
^
^
10 20 30
GAS VOLUME (CFM X 10,000)
40
_L
I
15 6 7 8 9 10 11
ADT/DAY X 100
CASED ON 350 CFM/ADT/DAY
FIG. 5 - '8 CONTROL METHOD COST FOR 95X EFFICIENCY
ELECTROSTATIC PRECIPITATOR ADDED IN SERIES WITH AN
EXISTING PRECIPITATOR-RECOVERY BOILER (99.4X A.O.E.)
5-62
-------
*
~ 100
X
TT so
h- u"
s
O
^ 60
OL
o 4Q
_j
^ 20
30
o"
o
o
o 25
X
^ 20
V)
8 ^
<
3 10
UJ
5
-------
Ul
I
3»-n
a •-"
o o
rn •
0 ui
3 •
X -<
1-1 -u
l/> I-H
•O 50
» 53
m i>
o z
i— i £7>
-o m
m m
o z
O -4
*z cr
m jo
50 >-•
C^
CO O
O 50
•-t CZ
moo
50 m
•73
NOTE:
LOCATION OF VEMTURI
SCRUBBERS AS SHOWN IS
FOR CLARIFICATION. COSTS
WERE FIGURED WITH THE
LOCATION BEING TO THE
SIDE OF PRECIPITATOK.
ADDED EQUIPMENT SHOWN WITH HEAVY LINES
Jl
v
-------
0
o
o
o 30
X
~ 25
h-
1/5
0
" 20
<
I— t
O.
5 15
o in
/
s
/
/
/
s
/
/
-, 24
o
o
a
o 20
X
^ 16
12
10 20 30
GAS VOLUME (CFM X 10,000)
40
I
I
I
I
I
^ 5 6 7 8 9 10 11
ADT/DAY X 100
BASED ON 350 CFM/ADT/DAY
FIG. 5 - 21 CONTROL METHOD COSTS FOR A 99% EFF.
VENTURI SCRUBBER ADDED TO AN EXISTING 90X EFF.
PRECIPITATOR - RECOVERY BOILER (99.8* A.O.E.)
5-65
-------
s
o
00
o
35
30
20
0.
2 15
10
o
o
o
*
o
X
00
o
o
-------
35
o
- 30
X
25
^ 20
co
o
o
o
0
o
S
0
15
10
14
10
10
II
20 30 40
GAS VOLUME (CFM X 10,000)
L
I
I
_L
8
10 11
ADT/DAY * 100
BASED ON 350 CFM/ADT/DAY
FIG, 5 - 23 CONTROL METHOD COSTS FOR A 9Q% EFF.
VENTURI SCRUBBER ADDED TO AN EXISTING 90* EFF.
PRECIPITATOR-RECOVER* BOILER (9Q.B% A.O.E.)
5-67
-------
In addition to the Venturi and fan, the costs include
a recirculattion pump and motor, instruments and controls
for proper operation, required piping for make-up, recircu-
lation pump and motor, instruments and controls for proper
operation, required piping for make-up, recirculation and
effluent, and revised ductwork to connect the Venturi with
the existing precipitator and stack.
Gas volumes are actual duct conditions at the inlet to the
Venturi. Particulate removal efficiencies are based on a
loading of 1.8 grains per actual cubic foot.
Cyclonic scrubber. This control method is based on the
installation of two or more cyclonic scrubbers in series
with an existing electrostatic precipitator as shown in
Figure 5-24. In addition to the scrubbers, a recirculation
pump and motor for each scrubber, instruments and controls
for proper operation of the above, required piping for make-
up, recirculation, and effluent to the precipitator, and
revised ductwork to connect the scrubbers with the existing
system are included. One 20-foot stub stack is included
for each scrubber located on the roof, while additional duct
is included to connect each scrubber located on the ground
to the existing stack.
There are several widely used materials of construction for
cyclonic scrubbers, 304 and 316 stainless steel, fiberglass,
and mild steel with concrete lining. The capital cost is
based on 304 stainless steel cyclone and outlet duct or stack
as this would be the most representative material for this
application; inlet duct is carbon steel.
Gas volumes are actual duct conditions at the inlet to the
scrubber. Particulate removal efficiencies are based on a
loading of 1.8 grains per actual cubic foot.
Capital costs and annual operating costs are presented in
Figure 5-25.
Effectiveness
Particulate removal. Considerations for this case are based
on an existing recovery furnace operating with a 90 percent
AOE precipitator. Various control methods are considered for
increasing the particulate efficiency to levels of 98.0, 99.4,
and 99.8 percent AOE. These control methods are as follows:
5-68
-------
NOTE:
LOCATION OF SCRUBBERS AS SHOWN
IS FOR CLARIFICATION. COSTS
WERE FIGURED WITH THEIR LOCATION
BEING TO EACH SIDE OF PRECIPITATOR.
ADDED EQUIPMENT SHOWN WITH HEAVY LINES
Ul
I
-------
§-
o
30
0
o
5 20
a.
£
o
15
10
-— iu
o
o
o
2 3
X
6
o
o
g 4
t 2
^
rt
^
*^
f*
^
^
s^
^
^
^
10 20 30 40
GAS VOLUME (CFM X 10,000)
5
8
10 11
ADT/DAY X 100
BASED ON 350 CFM/ADT/DAY
FIG. 5-25 CONTROL METHOD COSTS FOR CYCLONIC
SCRUBBER ADDED IN SERIES WITH EXISTING PRECIPITATOR
ON GROUND - RECOVERY BOILER
5-70
-------
Guaranteed AOE of AOE of
Control Method Efficiency New Equip. Exist. Equip.
Precipitator 99.0 98.5 90.0
Venturi Scrubber 99.0 98.0 90.0
Precipitator 95.0 94.5 90.0
Venturi Scrubber 95.0 94.0 90.0
Precipitator 90.0 89.5 90.0
Venturi Scrubber 90.0 89.0 90.0
Cyclone Scrubber 80.0 90.0
Curves for particulate cost effectiveness are presented
in Figures 5-26 and 5-27.
Reduced Sulfur and Sulfur Dioxide Removal. Precipitators
remove particulate only and cannot absorb either reduced
sulfur or sulfur oxides. Depending on the scrubbing liquid,
both the Venturi and cyclonic scrubbers have the potential
for absorbing reduced sulfur and sulfur oxides. However,
present technology is not well enough developed to assign
values for absorption efficiencies.
Scrubbers may also increase the sulfur compounds in the
flue gas with certain scrubbing liquids (e.g., unoxidized
black liquor).
Operation. As compared to precipitators, Venturi or cyclonic
scrubbers produce a lower temperature, higher moisture con-
tent gas. These conditions can result in more corrosion and
a lower plume rise.
Summary. The cyclonic scrubber has the lowest cost of the
methods considered; however, the efficiency is limited to
98.0 percent. Where higher efficiencies are required, the
Venturi should be the most effective method.
5-71
-------
160
i—
t/O
o
0 120
i
£0"
•-H 0
Sc° 80
0 0
1— 1
_l X
«£*0-
£*-' 40
1—
C
A
B —
— <*•
MM^^M
^— •"
-•
160
98.0 98.5 99.0 99.5 100
% A.O.E.
105,000 CFM,(300 ADT/DAY)
° 120
•-H O
£°. 80
o o
40
98.0 98.5 99.0 99.5 100
% A.O.E.
175,000 CFM,(500 ADT/DAY)
CO
o
160
120
i-. O
^°- 80
C_> C
98.0 98.5 99.0 99.5 100
% A.O.E.
350,000 CFM,(1000 ADT/DAY)
A. PRECIPITATOR ADDED IN SERIES
WITH AN EXISTING PRECIPITATOR
B. VENTURI SCRUBBER ADDED IN
SERIES WITH AN EXISTING PRE-
CIPITATOR
C. CYCLONIC SCRUBBER ADDED IN
SERIES WITH AN EXISTING PRE-
CIPITATOR
NOTE - ADT/DAY BASED ON 350
CFM/ADT/DAY
FIG.5-26 PANICULATE COST EFFECTIVENESS (TOTAL
CAPITAL COST COMPARISON) FOR ADD-ON CONTROL
EQUIPMENT - RECOVERY BOILER
5-72
-------
«/5
O
O-
tu
30
20
10
98
1
C
A
«
B -
-m •*
— •—
r-±L
-^^-
.0 98.5 99.0 99.5 100
% A.O.E.
05,000 CFM,(300 ADT/DAY)
40
So 20
< x
««•
H--^*
z 10
98.0 98.5 99.0 99.5 100
X A.O.E.
175,000 CFM,(500 ADT/DAY)
oo
40
30
O
_l O
•to
z o 9n
^ »-i £(J
< x
10
98.0 98.5 99.0 99.5 100
X A.O.E.
350,000 CFM,(1000 ADT/DAY)
A. PRECIPITATOR ADDED IN
SERIES WITH AN EXISTING
PRECIPITATOR
B. VENTURI SCRUBBER ADDED
IN SERIES WITH AN EXISTING
PRECIPITATOR
C. CYCLONIC SCRUBBER ADDED IN
SERIES WITH AN EXISTING
PRECIPITATOR
NOTE - ADT/DAY BASED ON 350
CFM/ADT/DAY
FIG.5-27PARTICULATE COST EFFECTIVENESS (NET
ANNUAL COST COMPARISON) FOR ADD-ON CONTROL
EQUIPMENT - RECOVERY BOILER
5-73
-------
5.3.6.1.3 Recovery System, Case 3
Application
This case is based on the installation of cyclonic
scrubbers on the roof of an existing recovery build-
ing. The scrubbers would be in series with and
following an existing 95 percent AOE precipitator
also located on the roof.
Costs
Estimated costs are based on the equipment arrangement
shown in Figure 5-28. These costs are shown in Figure
5-29 and include the scrubbers, a recirculation pump
and motor for each scrubber, instruments and controls,
piping for make-up, recirculation, and effluent to the
precipitator, revised ductwork to connect the scrubbers
with the existing system, and one 20-foot stack for
each scrubber. It was assumed that allowances were
made in the original building design for future scrub-
bers on the roof, and no costs for building structure
are included.
There are several widely used materials of construction
for cyclonic scrubbers, 304 and 316 stainless steel,
fiberglass, and mild steel with concrete lining. The
capital cost is based on 304 stainless steel cyclone and
outlet duct or stack as this would be the most repre-
sentative material for this application; inlet duct is
carbon steel.
Gas volumes are actual duct conditions at the inlet to
the scrubber. Particulate removal efficiencies are
based on a loading of 1.8 grains per actual cubic foot.
Effectiveness
Particulate Removal. The addition of cyclonic scrubbers
to an existing 95 percent AOE precipitator would result in
an expected overall AOE of 99 percent.
Reduced Sulfur and Sulfur Dioxide Removal. Cyclonic
scrubbers have the potential for either removing or
releasing sulfur compounds depending on the scrubbing
liquid utilized.
5-74
-------
SHOWN WITH HEAVY LINES
FI6.5-28TYPICAL ARRANGEMENT FOR CYCLONIC SCRUBBER ADPED IK SERIES
WITH EXISTING PP..1CIPITATOR Of' ROOF - RECOVERY BOILER
5-75
-------
CO
o
O
30
25
< 20
o
15
10
*
8
X
u* 3
S 4
i i
10 20 30 40
6AS VOLUME (CFM X 10,000)
345
7 8 9 10 11
ADT/DAY X 100 '
BASED ON 350 CFM/ADT/DAY
FIG.5 - 29 CONTROL METHOD COSTS FOR CYCLONIC
SCRUBBER ADDED IN SERIES WITH EXISTING PRECIPITATOR
ON ROOF - RECOVERY BOILER
5-76
-------
Operation. Possible operating disadvantages of the
cyclonic scrubbers would be corrosion potential
and lower plume rise resulting from high moisture.
Summary. The addition of cyclonic scrubbers appears
to be the only practical method of increasing the
efficiency of an existing precipitator located on
the roof. The possibility of installing Venturi
scrubbers was not considered in detail because
space is: not usually available for the required addition
of an I. D. fan and ductwork.
5.3.6.1.4 Recovery System, Case 4
Application
This case is based on applying additional control
methods to an existing recovery system with an 80
percent AOE black liquor venturi. The methods con-
sidered are as follows:
1. Convert Venturi to cyclone evaporator and
add an electrostatic precipitator.
2. Convert existing Venturi to low energy type
and add a second stage Venturi scrubber.
Costs
Freeipitator. Control method costs are based on the
equipment arrangement shown in Figure 5-30. These costs
as shown in Figures 5-31 and 5-32 include the addition
of a double chamber, common wall, tile construction
precipitator and the removal of the Venturi section and
the addition of inlet spray nozzles to the existing
cyclone. The capital cost for this method is based
on 50 percent solids being fed to the existing Venturi.
Any concentration less than 50 percent would require
additional multiple effect evaporator capacity to
increase the feed liquor concentration.
The existing tubular air heater will be replaced with a
steam coil air heater. This replacement cost, as well
as the cost for piping and duct changes, is included in
5-77
-------
ADDED EQUIPMENT SHOWN WITH HEAVY LINES
FIG. 5-30 TYPICAL ARRANGEMENT FOR CONVERTING EXISTING VENTURI
TO DIRECT CONTACT EVAPORATOR AND ADDING PRECIPITATOR - RECOVERY
BOILER
5-78
-------
3" 280
o
o
0
" 240
X
H- 200
CO
O
o
Sc 160
i—
D-
" 120
o
• — on
sS
.s
S*
/
S
/
s
^^
70
cr-
0
o
X
50
o 40
20
10 20 30
SAS VOLUME (CFM X 10,000)
40
156 789
ADT/DAY X 100
:;ASEH ON 350 CFM/ADT/DAY
10 n
FIG. 5-31 CONTROL METHOD COSTS FOR CONVERTING
EXISTING VENTURI TO CYCLOME EVAPORATOR AND ADD
9S.9X EFFICIENCY PRECIPITATOR - RECOVERY BOILER
5-79
-------
o 210
o
o
«
o
- 180
X
150
to
o
120
(X
JS
o
90
60
o
X
60
50
40
oo
S 30
20
10
10 20 30
GAS VOLUME (CFM X 10,000)
40
45 6 789
ADT/DAY X 100
BASED ON 350 CFM/ADT/DAY
10 11
FIG. 5-32 CONTROL METHOD COSTS FOR CONVERTING
EXISTING VENTURI TO CYCLONE EVAPORATOR AND ADD
99.0% EFFICIENCY PRECIPITATOR-RECOVERY BOILER
5-80
-------
the capital cost estimate. The cost for additional steam
requirement for the steam coil air heater is reflected in
the annual net cost for this method.
A new steel stack is included in the capital cost.
Second Stage Venturi. This method is based on a second
stage Venturi scrubber being installed subsequent to an
existing Venturi as shown in Figure 5-33. The existing
Venturi will be converted to a low energy Venturi and
used as an evaporator. To complete this arrangement, a
new I.D. fan with a greater statis head, a new recircula-
tion pump, a stack, and revised 304 stainless steel duct-
work from the existing Venturi to the Venturi scrubber
and from the Venturi to the stack will be required. The
second stage Venturi will utilize fresh water as the
scrubbing medium. The scrubber effluent will be pumped
to the first stage Venturi as shown in the flow schematic
for this control method.
Costs as shown in Figure 5-34 are based on the above
equipment and the instruments, motor controls, piping and
power wiring required for proper operation. The net
annual cost shown reflects a large credit due to chemical
recovery. This credit is based on the increase of an
operating collection efficiency from 80 to 97 percent
AOE. For an actual case with a more efficient existing
system and an existing higher system pressure drop, the
credit for chemical recovery would decrease, a power cost
savings may be realized, and the existing I.D. fan may be
adequate, thus reducing the capital cost.
Effectiveness
Particulate Removal. Efficiencies for removal of particulate
are as follows (based on 2 grains per ACF):
Annual
Guaranteed Operating
Control Method Efficiency Efficiency
Precipitator 99.9 99.5
Precipitator 99.0 98.5
Two Stage Venturi 98.0* 97.0
Curves for particulate cost effectiveness are shown in Figures
5-35 and 5-36.
*The two stage Venturi has a guaranteed efficiency of 99.0
percent based on the 4 grain per ACF particulate loading leaving
the recovery economizer. This 99.0 percent corresponds to a
guaranteed efficiency of 98.0 percent when adjusted for a 2
grain per ACF particulate loading, leaving a conventional DCE.
5-81
-------
40'
ADDED EQUIPMENT SHOWN WITH HEAVY LINES
FIG. 5-33 TYPICAL ARRANGEMENT FOR ADDING A
SECOND STAGE VENTURL TO Ail EXISTING VENTURI -
RECOVERY BOILER
5-82
-------
8
o
•i
o
« 50
«*
fc 40
o
f 30
»M
Ow
^o
H-
X
X
X
r
>
X
X
X
o
«-l
X
UJ
< 2
5 10 15 20 25 30
GAS VOLUME (CFM X 10,000)
35
II I I I 1 I I I I I I
234 567 8 9 10 11 12
ACT/DAY X 100
BASED ON 280 CFM/ADT/DAY
FIG.5- 34 CONTROL METHOD COSTS FOR ADDING
A SECOND STAGE VENTURI SCRUBBER TO AN EXISTING
VENTURI - RECOVERY BOILER (B7% A.O.E.)
5-83
-------
to
o
o
•-• o
o o
rH
_l X
cuu
Ibu
100
50
A
V
'
CO
o
o
200
150
_1 X
? A.O.E.
A. as.ooo CFM.OOO ADT/DAY)
3. 105,000 CFM,(300 ADT/DAY)
50
A.
B.
97
98 99
% A.O.E.
140,000 CFM,(500 ADT/DAY)
175,000 CFM,(500 ADT/DAY)
zuu
CO
o
0 150
_i
^°. 100
C_> O
t— i
_J X
£^ 50
t—
A
v
x
A.
B.
97 98 99
% A.O.E.
280,000 CFM,(1000 ADT/DAY)
350,000 CFM,(1000 ADT/DAY)
A. ADDITION OF A SECOND STAGE
VENTURI TO AN EXISTING
VENTURI
NOTE - ADT/DAY BASED ON 280
CFM/ADT/DAY
B. CONVERT EXISTING VENTURI
EVAPORATOR TO CYCLONE
EVAPORATOR AND ADD AN
ELECTROSTATIC PRECIPITATOR
NOTE - ADT/DAY BASED ON 350
CFM/ADT/DAY
FIG.5-35PARTICIPATE COST EFFECTIVENESS (TOTAL
CAPITAL COST COMPARISON) FOR INCREASING THE EFFICIENCY
OF AN EXISTING 80% A.O.E. VENTURI EVAPORATOR SYSTEM -
RECOVERY BOILER
5-84
-------
uu
A C.
j- 4b
10
0
-J ^ ^/\
-•— * 15
LU
n
A
i
B
*•"••
.-— •
uu
O
O
4K S *irt
3-30
z o
z •—»
C 15
r\
A
B
A.
B.
97
98 99
X A.O.E.
85,000 CFM,(300 ADT/DAY)
105,000 CFM,(300 ADT/DAY)
97 98
99
A.
B.
X A.O.E.
140,000 CFM,(500 ADT/DAY)
175,000 CFM,(500 ADT/DAY)
ou
^ 45
to
o
^§30
ID «
Z O
Z i— I
^5 15
LU
Z
n
A
1
B
S
^
X
A.
B.
97 98
99
A. ADDITION OF A SECOND
STAGE VENTURI TO AN
EXISTING VENTURI
NOTE - ADT/DAY BASED ON
280 CFM/ADT/DAY
B. CONVERT EXISTING VENTURI
EVAPORATOR TO CYCLONE
EVAPORATOR AND ADD AN
ELECTROSTATIC PRECIPITATOR-
NOTE - ADT/DAY BASED ON 350
CFM/ADT/DAY
X A.O.E.
280,000 CFM,(1000 ADT/DAY)
350,000 CFM,(1000 ADT/DAY)
FIG.5-36 PARTICULATE COST EFFECTIVENESS (NET
ANNUAL COST COMPARISON) FOR INCREASING THE EFFICIENCY
OF AN EXISTING 80% A.O.E. VENTURI EVAPORATOR SYSTEM -
RECOVERY BOILER
5-85
-------
Reduced Sulfur and Sulfur Dioxide Removal. If the recovery
unit is operated at complete combustion conditions, the
removal of the Venturi would not significantly affect the
emissions of reduced sulfur compounds.
However, the removal of the Venturi would probably increase
the SO emission somewhat. Limited in-house test data indi-
cate that the SO emission might increase from an approximate
range of 1/2 - 10 Ibs/ADT to an approximate range of 1 - 15
Ibs/ADT, when the Venturi is removed.
Operation. In comparing operation, the second stage Venturi
method would result in more potential corrosion and a lower
plume rise.
Summary. The second stage Venturi method has the lowest cost;
however, the efficiency is limited to 97.0 AOE. If a higher
efficiency is required, then the 99.5 AOE precipitator method
should be considered.
5.3.6.1.5 Recovery System, Case 5
Application
This case is based on the installation of a black liquor oxi-
dation system to reduce hydrogen sulfide emissions from an
existing recovery system. Both weak and concentrated black
liquor oxidation systems are considered. As discussed in
Section 5.2.7, the selection of an oxidation system for a
specific installation will depend to a large extent on the
characteristics of the liquor. Flow diagrams for the systems
considered are shown in Figures 5-37 and 5-38.
Costs
The total capital costs of weak and also concentrated liquor
oxidation systems appear on the cost curves, Figures 5-39 and
5-40. Air requirements for the weak liquor system are shown
in Figure 5-41. These total capital costs consist of the
total direct capital costs, plus indirect capital costs. The
indirect capital costs are computed as 30 percent of the direct
cost. Included in the total direct capital costs are: Equip-
ment and equipment erection (which includes instrumentation and
motors), equipment foundations and buildings, process and instru-
ment piping, power wiring and lighting.
5-86
-------
VENT
AIR
BLOWER
WEAK BLACK LIQUOR
FROM WASHING SYSTEM
Ol
I
00
STEAM
CYCLONE
OXIDIZED WEAK
" BLACK LIQUOR
TO MULT. EFFECT
EVAPORATORS
EXIST. UNOXIDIZED WEAK
BLACK LIQUOR
STORAGE
(WIN. ia HRS STORAGE ASSUMED
EXIST.
PUMP
FOAM TANK
PUMP
NEW OXIDIZED WEAK
BLACK LIQUOR
STORAGE
PUMP
FIG. 5-37
WEAK LIQUOR OXIDATION SYSTEM
-------
VENT
AIR
BLOWER
CONC. BLACK LIQUOR FROM
MULT. EFFECT EVAPS.
01
I
00
00
UNOXIDIZED CONC.
BLACK LIQUOR
STORAGE
PUMP
OXIDATION TANK PUMP
EXIST. OXIDIZED CONC.
BLACK LIQUOR
STORAGE
PUMP
FIG. 5-38
CONCENTRATED LIQUOR OXIDATION SYSTEM
-------
45
- 40
X
I/O
o
o
O.
35
30
25
20
16
o
o
o
*
o
X
- 14
12
S 10
10 12 14 16 18 20 22
AIR FLOW (ACFM X 1,000)
II I I | I I I I I I
3 4 5 6 7 8 9 10 11 12
ADT/DAY X 100
BASED ON 3,000 LBS. B. L. SOLIDS/ADT,
15% B. L. SOLIDS, 10 GRAMS Na?S/LITER,
SPECIFIC GRAVITY OF 1.1 *
FIG. 5-39 CONTROL METHOD COST FOR WEAK
BLACK LIQUOR OXIDATION
5-89
-------
35
s
o
- 30
X
v»-
r 25
to
o
0
^ 20
h-
i— i
Q.
«t
" 15
<:
£
•- in
X
x
X
X
X
w
s
/
/
o
X
* 10
oo
S 8
LkJ
12 18 24 30 36 42 48
GPM I 10
345 6 789 10 11
ADT/DAY X 100
BASED ON 3,000 IBS. B.L. SOLIDS/ADT,
50% B.L. SOLIDS, SPECIFIC GRAVITY OF 1.25
FIG. 5 - 40 CONTROL METHOD COSTS FOR CONCENTRATED
BLACK LIQUOR OXIDATION
5-90
-------
WEAK BLACK LIQUOR OXIDATION
AIR REQUIREMENTS
LIQUOR
FLOW
1004-GPM
1504-
200-J-
300--
400--
500--
600
700
800
900
1000
1500-h
2000-f
•50
•40
-30
-20
-5
-4
•3
-2
4-1
AIR FLOW
30 x 1000 Acfm
25
-t-20
15
10
9
8
+5
5-91
FIG. 5-41
-------
Effectiveness
Particulate. There are usually no particulate emissions
from the oxidations systems. Oxidation of liquor has no
effect on particulate emissions from the recovery system.
Reduced Sulfur and Sulfur Dioxide Removal. The basic
purpose of an oxidation system is to reduce the malodorous
emission from the direct contact evaporator. This mal-
odorous emission from the direct contact evaporator primarily
occurs in the form of reduced sulfur gases (H S, RSH, RSR,
RSSR). As stated previously, when the Na S is oxidized to
sodium thiosulfate (Na S 0 ), the above reaction does not
occur, since sodium thiosulfate is a relatively stable
compound.
Of course, the oxidation of the liquor will not reduce the
emission of odorous compounds from the recovery furnace.
Odorous furnace emissions are a function of recovery furnace
operation, such as overload, air distribution, and spray
droplet size. Oxidation will only reduce emissions from the
direct contact evaporator.
Experience at several mills has proven that any oxidation
system must operate at the highest possible efficiency. The
term efficiency is rather nebulous since the measure of
Na S in the liquor is the true indicator of oxidation per-
formance. Therefore, any oxidation system should be designed
to provide liquor at a Na S content of 0.10 grams per liter
or less to the direct contact evaporator.
At this level, the H S level in the flue gases should average
5-20 ppm by volume, if the furnace is operated properly.
Upsets in furnace operation may extend this range as high as
100 ppm.
Concentrations of other reduced sulfur compounds are even
more difficult to define than H S.
For high efficiency oxidation systems, the reduced sulfur
compounds—RSR, RSH, RSSR—are present in very low concen-
trations. Most test results on properly operated furnaces
report a range from 0 to 3 ppm leaving the direct contact
evaporator.
Since H S has been the most predominant odorous compound
leaving the direct contact evaporator, the other reduced
sulfur compounds have been viewed in the past as secondary
5-92
-------
in importance. While this is true, the resolution of H S
emissions will require increasing emphasis to be directed
to the remaining reduced sulfur compounds. Perhaps, the
emissions of these compounds will become more fully under-
stood in the future.
The role of sulfur dioxide emission in relation to black
liquor oxidation has not been clearly defined. However,
actual experience indicates that SO emission after oxi-
dation (based upon proper furnace operations) will not
exceed the SO loss before oxidation. Mills which operate
at a low sodium sulfide residual to the direct contact
evaporator have reported SO levels in the order of 100
ppm by volume.
Operation. The major emissions from the oxidation system
stack are dimethyl sulfide and dimethyl disulfide. The
dimethyl sulfide, dissolved in the liquor, escapes to the
atmosphere. On the other hand, the dimethyl disulfide is
created by the oxidation of methyl mercaptan by oxygen in
the presence of alkali, the dimethyl disulfide then being
released to the atmosphere.
Maintenance costs for the two oxidation systems are not
clearly defined due to many design variations and the
unavailability of adequate maintenance cost records. Since
the capital costs of the two systems are relatively equal,
the maintenance costs are also assumed to be approximately
equal.
^=afc_
For properly designed and applied systems, the reliability
of weak or concentrated oxidation systems are approximately
equal.
Summary. The capital costs and net annual costs for this
method do not include any allowance for equipment to treat
the exhaust gas from the oxidation systems. Further, the
net annual cost does not include any allowance for chemical
recovery, since this savings (if any) will vary from mill to
mill and must be individually calculated. Provisions should
be considered for treatment of this exhaust gas, where mini-
mum emission is essential.
In current technology, oxidation syst:ems must compete with
new designs which eliminate direct contact between the black
liquor and the flue gases to determine the overall recovery
5-93
-------
system most suitable for a particular mill. A comparison
of these new systems applied to existing mills is discussed
in Section 6.2.1.1. This case is based only on an appraisal
of weak liquor oxidation versus concentrated liquor oxidation.
Cost curves have been prepared showing net annual cost versus
tons per day production for weak and concentrated liquor oxi-
dation systems in Figure 5-42.
As shown in the curves, the net annual costs of the two systems
are very close, depending upon the sodium sulfide concentration.
For liquor oxidation systems designed for 10 grams/liter of
Na S, a concentrated system appears to be the best choice for
sizes below 800 ADT/Day. Considering the fact that a concentrated
system eliminates the sodium sulfide reversion reaction which takes
place with a weak liquor system, a concentrated system should pro-
bably be selected even for tonnages higher than 800 ADT/Day.
In conclusion, the selection of the oxidation system must be
analyzed for each mill; however, considering all factors, a
concentrated system appears to be the best choice.
5.3.6.1.6 Recovery System, Case 6
Application
For older recovery systems (more than ten years old), odorous
emissions from the recovery furnace may be significant, even
though the most modern current technology (proper operation and
black liquor oxidation) is employed. It is conceivable that for
some older recovery systems, the odorous emissions may never be
reduced to levels which are currently being considered by regula-
tory authorities (this refers to H S levels in the order of 17 ppm)
This may be attributed to the older furnace configurations and the
older systems which introduce air into the recovery furnace.
Therefore, these older recovery systems may have to be replaced
with one of the recovery systems listed below. Accordingly,
approximate capital costs have been prepared for a range of sizes
of the following recovery systems:
1. New Conventional Recovery System (Figure 5-43)
2. New Conventional Recovery System with Concentrated
Black Liquor Oxidation (Figure 5-43)
5-94
-------
o
o
o
X
•tq-
CO
o
o
13
12
11
10
WEAK LIQUOR OXIDATION
CONCENTRATED LIQUOR OXIDATION
X
567
ADT/DAY X 100
8
10
FIG.5-42 COMPARISON OF NET ANNUAL COSTS FOR
BLACK LIQUOR OXIDATION SYSTEMS
5-95
-------
3. New Recovery System Utilizing 62 Percent Solids
Black Liquor from the Multiple Effect Evaporators
(Figure 5-44)
4. New Air Contract Evaporator Recovery System
(Figure 5-44)
Costs
In estimating capital costs, the recovery system tonnages
were based on 3,000 pounds of dry solids per ADT and 6,600
BTU's per pound of dry solids. Capital cost was estimated
for two steam conditions of 600 psig-750°F and 1250 psig-
900°F. Capital costs for 850 psig-825°F would be approxi-
mately midway between these two curves and may be interpolated.
The above capital costs are based on the U.S. average labor
costs (U.S. Avg. = 100), and should be adjusted for the area
of the country in regard to construction labor costs. This
can be accomplished by adjusting 50 percent of the capital
cost (labor and material comprise approximately 50 percent
each of the capital cost), in relation to the Construction
Labor Cost Index, Figure 5-45. Even after this adjustment,
the capital cost may vary ± 10 percent depending upon geo-
graphical area (primarily due to weather considerations),
company preferences, and differences in engineering design.
These capital costs assume that the new recovery system will
replace an existing recovery system which is being operated
at complete combustion conditions (approximately design
capacity). Therefore, it is anticipated that only minor
quantities of additional make-up water, cooling water,
feedwater, compressed air, electricity, et cetera, will be
required for this new system. If any of these utilities may
be required in significant quantity, they must be added
to the capital cost in this section. Further, these
costs are based on ah open area and do not include any
allowance for site clearance. The cost of loss of pro-
duction due to downtime for the installation of the new
recovery system .is not included."
Generally speaking, these capital costs may be described
as follows:
1. A recovery system designed to produce steam at
the turbine-generator operating conditions specified
on the cost curves
5-96
-------
PAGE NOT
AVAILABLE
DIGITALLY
-------
PAGE NOT
AVAILABLE
DIGITALLY
-------
FIGURE 5-45
CONSTRUCTION LABOR COST INDEX
THE LABOR INDEX SHOWN BY STATE IS BASED ON AVERAGE OF SELECTED CITY INDICES
WITHIN THE STATE .
STATES NOT HAVING PULP MILLS NOT SHOWN.
INDEX DATA FROM "BUILDING CONSTRUCTION COST DATA 1969" 27TH EDITION,
ROBERT SNOW MEANS CO., INC. EXCEPT* FROM "DOD COST REVIEW GUIDE, 1970."
INDEX 100 = 1968 U.S. AVERAGE
5-97
-------
2. An electrostatic precipitator at specified efficiency
located on the building roof. The precipitator would
be a filled tile shell unit of two-chamber construction.
3. The recovery building would be enclosed with asbestos
or equivalent siding up to approximately the mud drum
level. The floors in the building would be of concrete
construction.
4. A control room is included for this recovery system.
While it may be desirable to include the controls for
this unit in a central control room, the capital costs
are based on a control room adjacent to this unit.
5. All liquor piping is based on carbon steel material.
6. All equipment which is within the recovery building
is included in the estimate. The following additional
equipment which is usually located outside the building
is also included: (No building enclosures are included
for this equipment located outside the recovery building).
a. Piping tie-ins for green liquor, weak wash, salt
cake make-up, high pressure steam, 50 Ib. steam,
and feedwater and black liquor
b. Black liquor dump tank
c. A black liquor evaporator concentrator is included
for the systems which require this.
d. Concentrated black liquor oxidation unit for the
recovery system utilizing oxidation
e. Black liquor storage tank for 62 percent solids
and 55 percent solids liquor system; 12-hour
storage is included.
Capital cost curves are shown in Figures 5-46 and 5-47.
It should be pointed out that these cost curves do not include any
potential credits for better smelt reduction, increased pulp pro-
duction, increased steam production, or manpower reductions. Depend-
ing on individual circumstances, the credits may be a substantial
amount.
5-98
-------
CONVENTIONAL KRAFT RECOVERY SYSTEM
o" 14 99.5% EFFICIENCY PRECIPITATOR
6 8 10
ADT/DAY X 100
CONVENTIONAL KRAFT RECOVERY SYSTEM WITH
§14
o
o
o
X
P0
oo
o
0
< 8
1—
t— 1
a.
5 6
P ,
CONCENTRATED BLACK LIQUOR OXIDATION 99.5%
EFFICIENCY PRECIPITATOR
/
/
>
//
c
jp
^
S>'^
^^
^
O j
o!^
*r •
| »
^
»»
^^
^-
6 8 10
ADT/DAY X 100
12 14
FIG. 5-46 TOTAL CAPITAL COST BASED ON
3,000 LBS. SOLIDS7ADT AND 6,600 BTU/LB.
DRY SOLIDS
5-99
-------
o
o IA
S.
0
0
o
_ri2
X
~1G
oo
0
0
_j 8
<:
i— i
Q-
S 6
__i
<:
P 4
62* BLACK LIQUOR SOLIDS KRAFT
RECOVERY SYSTEM WITH 99. 5X
EFFICIENCY PRECIPITATOR
/
'/
\f
//
^
^
•
4 6 8 10
ADT/DAY X 100
12
14
I 14
o
o
0
-12
X
•faO-
h- 10
oo
o
t 8
i— »
Q.
o 6
_i
•a:
i 4,
AIR CASCADE EVAPORATOR KRAFT
RECOVERY SYSTEM WITH 99.5%
EFFICIENCY PRECIPITATOR
/
V
^
'/
^
X
X
fe^
^x
X^
(?s^'
^'X
x^
x^
^^
X
x^
•
1 4 6 8 10 12 14
ADT/DAY X 100
FIG. 5-47 TOTAL CAPITAL COST BASED ON
3,000 LBS. SOLIDS/ADT AND 6,600 BTU/LB.
DRY SOLIDS
5-100
-------
3.6-2 Smelt Dissolving Tank
Application
The molten inorganic chemicals leave the furnace through
smelt spouts to an agitated dissolving tank which contains
weak wash liquor obtained from washing lime sludge in the
succeeding causticizing operation. After being cooled
and dissolved in the weak wash, the smelt solution is
referred to as green liquor. Shatter nozzles are used to
inject steam and/or recirculated green liquor into the
solid smelt stream from the furnace to disperse the molten
chemicals and ensure its safe dissolution.
Large quantities of water vapor are released by the green
liquor which cools the molten smelt and by the steam from
the shatter jets. Particles of smelt and droplets of green
liquor become entrained in these vapors and are exhausted
through the smelt dissolving tank vent.
The gas vented from the smelt dissolving tank has a tempera-
ture of approximately 200*F. and contains traces of organic
sulfur compounds and particulate matter with a concentration
of 1 - 1.5 grains/SCFD. The particulate matter is mostly
sodium sulfide and sodium carbonate.
The majority of mills provide control equipment to minimize
the emission of these chemicals and droplets to the atmos-
phere. Control equipment which is currently in use is as
follows:
1. Mesh Pads. Mesh pads are used extensively to
collect chemicals in dissolving tank vents.
Sprays located above the mesh pads operate
periodically to remove the collected chemicals
which are returned to the smelt dissolving tank.
A fan is not normally required due to the low
pressure drop.
2. Packed Towers. Packed towers have been reported at
two mills where they operate successfully. One con-
cern with packed towers is the possibility of plug-
ging; however, periodic use of condensate as scrubbing
liquid is reported to remove any solids buildup.
5-101
-------
Orifice Scrubber. This scrubber has been installed
at several mills and is reported to be successful
where the scrubber is adequately sized. Due to
the pressure drop (approximately 8 inches HO) of
this scrubber, inadequate sizing or improper opera-
tion may have a significant effect on emissions
(For example, higher than design flows may require
by-passing the scrubber.).
Costs
Cyclonic Scrubber. This control method is based on the
installation of a single cyclonic scrubber on an elevated
floor within the confines of an existing recovery boiler
building and adjacent to the existing vent stack. Since the
capital cost and operating cost of the cyclonic scrubber is
very similar to a packed tower, costs have not been calculated
for the cyclonic scrubber. The reader is referred to the costs
on the packed tower.
Packed Tower. This control method is based on the installation
of a single packed tower on an elevated floor within the con-
fines of an existing recovery boiler building and adjacent to
the existing vent stack. In addition, the following auxiliaries
are included: an axial flow fan and motor located in the outlet
of the tower, a circulation pump and motor, instruments and con-
trols for proper operation of the above, piping for make-up,
circulation, and tower effluent to dissolving tank, and revised
ductwork, including a by-pass to connect the tower to the exist-
ing vent stack. Cost curves are presented in Figure 5-48.
Orifice Scrubber. This control method is based on the installa-
tion of a single low energy scrubber on an elevated floor within
the confines of an existing recovery boiler building and adjacent
to the existing vent stack. In addition, the following auxiliaries
are included: an axial flow fan and motor located in the outlet of
the tower, a circulation pump and motor, instruments and controls
for proper operation of the above, piping for make-up circulation
and scrubber effluent to dissolving tank, and revised ductwork
including a by-pass to connect the scrubber to the existing vent
stack. Cost curves are presented in Figure 5-49.
Mesh Pad. This control method is based on the installation of a
mesh pad assembly in an existing recovery boiler vent stack.
5-102
-------
^T 85
o
o
A
3 75
*— fif;
o
o
_i
i* RR
n.
u
AC
— 1 HO
o
H-
oc
JO
30
o
o
0 25
X
«/»• op
OO
O
0 15
_!
^)
P in
«t 1 U
LU
t;
X
/
/
^
/
/
x
x
^x
/"
x^
x
x
10 20 30 40
GAS VOLUME (CFM X 1,000)
345 6 7 8 9 10 11
ADT/DAY X 100
BASED ON 35 CFM/ADT/DAY
FIG. 5 - 48 CONTROL METHOD COSTS FOR PACKED TOWER
ADDED TO SMELT DISSOLVING TANK VENT
5-103
-------
70
o
o
0
A
- 60
X
so
a.
-------
Allowances have been included for the following: alterations
to existing vent stack, normal piping for spray wash, and
instruments and controls for proper operation. Cost curves
are presented in Figure 5-50.
Effectiveness
Particulate Removal. Efficiencies for the methods considered
are as follows:
Annual Operating
Control Method Efficiency
Cyclonic Scrubber 80
Packed Tower 90
Orifice Scrubber 97
Mesh Pad 75
Curves for particulate cost effectiveness are shown in
Figures 5-51 and 5-52.
Reduced Sulfur and Sulfur Dioxide Removal. Depending on the
scrubbing liquid used, both the packed tower and orifice scrubber
would have potential for removing sulfur compounds.
Operation. Of the methods considered, the orifice scrubber should
be the least subject to plugging.
Summary. The orifice scrubber and the packed tower are the most
effective control methods for either particulate or gaseous
pollutant removal. The orifice scrubber is less susceptible to
pluggage and has the highest particulate collection efficiency.
5-105
-------
o
o
0
•V*
oo
o
o
o.
O
20
0
o
0
X
•*/»•
o
0
16
14
12
10
7
6
10
1
20 30
GAS VOLUME (CFM X 1,000)
1 1
1 1 1 I 1
1
8
456'
ADT/DAY x 100
BASED ON 35 CFM/ADT/DAY
9 10 11
FIG. 5 - 50 CONTROL METHOD COSTS OF MESH PAD
ADDED TO SMELT DISSOLVING TANK VENT
5-106
-------
JU
CO
o
_i
H^O
•-i O
f£ ~i 40
0 t-l
X
0 ^L
>-
A
T
B '
C
«
»
dU
V)
0
*-* 50
\-^^
t-i O
r\ f H A
< ^. 40
CJ T-H
X
-JV9-
o 2u
1 —
1
B
C
70
100
X A.O.E.
10,500 CFM,'(300 ADT/DAY)
70 86 100
% A.O.E.
17,500 CFM, (500 ADT/DAY)
i—i O
Q- C
40
20
70
35
100
% A.O.E.
35,000 CFM, (1000 ADT/DAY)
A. MESH PAD, 75X A.O.E.
B. PACKED TOWER, 90* A.O.E.
C. ORIFICE SCRUBBER, 97% A.O.E,
?!OTE - ADT/DAY BASED ON 35
CFM/ADT/DAY
FIG.5- 51PARTICULATE COST EFFECTIVENESS (TOTAL
CAPITAL COST COMPARISON) FOP. CONTROL EQUIPMENT
ADDED TO SMELT DISSOLVING TA!!K VENT - RECOVERY
"OILER
5-107
-------
An
o 30
o
_l O
<: o
~ ° or
• «
Z^ i— i
•a: x
•b^-
i ^ ^
UJ
10
% A.O.E.
10,500 CP1, (300 ADT/2AY)
30
20
1U
A
T
R
t
C
70
35
100
% A.O.E.
17,500 CFM, (500 ADT/DAY)
f.n.
to
o
C_3
< C
r: c
T
B T
70 v; io(
X A.O.E.
3:,000 CFM, (1000 ART/DAY)
A. MESH PAD, 75% A.O.E.
B. PACr.EF TOWER, ?0« A.O.E
C. ORIFICE SCRUBBER, T7X />
f!OTE - ADT/DAY BASED ON 35
CFM/ADT/DAY
FIG.5-52PARTICULATE COST EFFECTIVENESS ('!ET
ANNUAL COST COMPARISON) FOR CONTROL EQUIPMENT
AOHEP TO SMELT DISSOLVING TAflK VEMT - RECOVERY
ROILFP
5-108
-------
ei3 6 3 Digester Relief and Blow, Multiple Effect Evaporators
Application
The gases leaving the batch digester blow tank, turpentine condensers,
and multiple effect evaporators all contain high concentrations of
organic sulfur compounds. Despite the fact that the gas volumes
are rather small, the high concentration of odorous compounds makes
these vents major contributions to the odor level in the mill. Because
a continuous digester blows into a blow tank at a temperature below
the flash point there are no gases venting from a continuous digester
blow tank. Few mills have control equipment on these sources. With
the present emphasis on odor control, much attention lately has
been given to these sources. Depending on local conditions, these
sources can be treated independently or commonly. The odor removal
problem is the same in both cases. Methods which are presently in
use are as follows:
Chlorination. This system consists of the equipment necessary to
collect and chemically oxidize pulp mill noncondensible gases as
shown in Figure 5-53.
The digester relief and blow gases and gases emerging from the
evaporator hot well vent and from the turpentine condenser and
decanter vents are exposed to chlorination stage washer effluent
from the bleach plant, either in the dropleg of the washer or in
a scrubber. The dimethyl sulfide (RSR) is absorbed and oxidized
to sulfone. The dimethyl disulfide (RSSR) present in the stream
of gas is absorbed and oxidized to methyl sulfonyl chloride. This
compound consumes a large proportion of the chlorine and may not
be fully absorbed and oxidized.
A portion of the required chlorine is usually available in the
chlorination stage washer effluent from the bleach plant. Supple-
mental chlorine may be required. In an unbleached pulp mill, chlorine
gas has to be supplied from an outside source. There is a system
available that scrubs the gases with chlorine and caustic. This
system also treats the condensate by air stripping.
Combustion. This system consists of the equipment necessary
to collect and thermally oxidize the sulfur bearing compounds in
pulp mill noncondensibles to sulfur dioxide by incineration in the
lime kiln or a separate incinerator if the distance from the source
to the lime kiln makes the installation cost prohibitive. Sulfur
dioxide is considered less objectionable and more easily removed
than the compounds from which it was formed. Figure 5-54 shows
the arrangement of this system.
5-109
-------
FROM
SEAL TANK
FROM Ci2
EVAPORATOR
NONCONDENSABLE GASES
DIGESTER
NONCONDENSABLE GASES
SEAL TANK 1 CAUST|C
CHLORINATOR^ SCRUBBER'
Ul
VENT
t\
FIG. 5-53
CHLORINATION SYSTEM
-------
NOTE: GAS ACCUMULATOR it> REQUIRED FOR
COLLECTING AND STORING GASES
DURING PEAK FLOWS OF A BATCH
DIGESTER SYSTEM ONLY.
COMBINATION
PRESS. AND VAC. |
RELIEF VALVE
GAS
ACCUMULATOR (SEE NOTE)
FLOW
NTROL
BLOW TANK
CONDENSEK VENT .
(BATCH DIGESTERS)
Ul
i
TURPENTINE
CONDENSER VENT
(BATCH DIGESTERS)
EVAPORATOR
NON-CONOENSIBLE
GASES
CONTINUOUS
DIGESTER
VEiNT
AIR
INLET
Ltl DAMPER
AR
FLAME
"RESTER
COMPENSATE
TRAP (TYR)
TO TURPENTINE DECANTER
OR LIQUOR ROOM
— -O--
FLAME OUT
CONTROL
FAN I
TURPENTIN
COND" ""
STEAMING VESSEL VENT
FIG. 5-54
INCINERATION OF NQN-CONDENSlBLE GASES_J.N__LjME,_,KJLN
FLASH TANK VENT
-------
In order to accomplish this conversion, the gases must be collected
and delivered to the kiln at a constant rate of flow. The non-
condensible portion of the digester blow gase$ discharges periodi-
cally from the blow heat condenser vent. These gases are piped
to a gas accumulator havj.ftg sufficient volume to allow collection
of the gases as they occur while discharging at a constant rate.
This accumulator is required only for use with batch digesters.
The noncondensible relief gases from the evaporator hot well vent
occur at essentially a steady rate and need not be collected in
a storage vessel and can be sent directly to the kiln.
All of the pulp mill noncondensible gases are directed to a packed
condenser-scrubber for the removal of turpentine mist from the gas
stream. Turpentine mist may interfere with the flame out control
through development of a false signal. From the scrubber the ga$es
are directed to the suction of the kiln fan, which is located near
the kiln. At this point the gases are diluted beyond the limits
of inflammability and introduced to the firing end of the kiln.
At the two thousand degree temperature usually encountered in the
lime kiln, the conversion takes place rapidly and completely.
Safety measures in the form of flame arresters are required to
avoid explosive conditions ahead of the lime kiln.
Caustic Scrubber. The scrubber is used to remove H S and mercaptang
by absorbing the gas in white liquor. (See Figures 5-55 and 5-56).
The method of collection of the pulp mill noncondensibles in this
system is similar to the method used for combustion previously dis-
cussed. A notable exception is that the gas holder for use with
batch digesters can be omitted. This is made possible by using a
scrubber sized to take the surges of gas that occur. The volume
supplied to the scrubber is actually constant, with the gas to
air ratio changing when a digester is blown.
Catalytic Oxidation. Catalytic methods of oxidizing noncondensible
gases have been tried. So far their use has not proven applicable
to this source due to severe operating problems and excessively high
maintenance costs (12).
5-112
-------
SEAL VENT
Ul
OJ
ACCUMULATOR
VENT
TURPENTINE
CONDENSER AND
DECANTER VENTS
EVAPORATOR
HOT-WELL VENT
VENT
w
1
BLOWER
VENT
WHITE LIQUOR
PACKED TOWER
TO CAUSTICIZING
FIG.5-55
CAUSTIC SCRUBBER - BATCH DIGESTERS
-------
VENT
VENT
Ol
i
TURPENTINE
CONDENSER AND
DECANTER VENTS
-€?-
1
BLOWER
EVAPORATOR
HOT-WELL VENT
WHITE LIQUOR
PACKED TOWER
TO CAUSTICIZING
FIG. 5-56
CAUSTIC SCRUBBER - CONTINUOUS DIGESTERS
-------
Costs
Following are cost curves for various systems showing the total
installed cost and the annual operating cost. (Figures 5-57
through 5-61).
Effectiveness
Particulates. There are no particulate emissions from this
source.
Reduced Sulfur Removal. Of the methods discussed, combustion
could be expected to have an effectiveness approaching 100
percent removal. Chlorination has almost no effect on H S,
about 98 percent reduction of mercaptan and more than a 50
percent reduction of sulfide. (Phillips said quite a reduction
in Hanible Island report).
Sulfur Dioxide Removal. The thermal oxidation of reduced sulfur
compounds results in the formation of sulfur oxides. However,
limited information developed in field studies by NCASI (19)
engineers has failed to show significant amounts of sulfur oxides
in the kiln gases when incinerating the noncondensibles. The
phenomena which reduce these sulfur oxides is unknown; however,
it is possible that the alkaline lime material in the kiln bed
provides an efficient removal mechanism for the acidic sulfur
oxides formed.
The relative levels of sulfur oxides for these control methods are
unknown. However, there does not appear to be a significant dif-
ference between these methods.
Operation. All of the methods except catalytic oxidation have
proved to be practical from an operational point of view.
Summary. At the present time, combustion of noncondensible
gases appears to be favored by mills in the United States
because of the 99 plus percent removal of total reduced sulfur
compounds.
5-115
-------
o
o
o
oo
o
o
35
30
25
20
15
10
8
o
o
o
X
* 6
o c
o 5
12
ADT/DAY X 100
FIG.5 - 57CONTROL METHOD COSTS FOR CHLORINATION
SYSTEM APPLIED TO NON-CONDENSI3LE GASES
5-116
-------
o /U
o
o
"- 60
X
•**)•
IT 50
o
CJ
Q.
5 30
_i
S
fe
•- 20
s*
s*S
^
^
^
^
r
§ 20
o 15
0 13
= 10
6 9
ADT/DAY X 100
12
FI6.5-58 CONTROL METHOD COSTS FOR INCINERATION
OF NON-CONDENSIBLE GASES IN LIME KILN - CONTINUOUS
DIGESTERS
5-117
-------
o
o
o
-" 120
X
o
o
a.
ff.
fe
100
80
60
40
-jvj
o"
O
°. 44
r—
X
^Z 38
t—
oo
o
0 32
_i
<:
1 26
H-
LU
sz
9n
>
X
X
X
x
s
/
/
/
6 9
ADT/DAY X 100
12
FIG..5- 59CONTROL hCTHOD COSTS FOR
'INCINERATION OF NON-CONDENSIBLE GASES IN
LIME KILN - BATCH DI6ESTERS
5-118
-------
— 40
" 35
x
**
30
25
£
o
I
20
• 15
_ n
o"
o
o
^ 10
X
I/O
o
o
6 9
ADT/DAY X 100
12
FIG. 5-60 CONTROL METHOD COSTS FOR CAUSTIC
SCRUBBER APPLIED TO NON-CONDENSABLE GASES - CONTINUOUS
DIGESTERS
5-119
-------
~ 45
o
o
o
*" 40
X
H- 35
oo
o
0
-------
5.3.6.4 Lime Kiln
Application
This control method is based upon the addition of a high
energy Venturi scrubber to replace existing 80 percent
AOE cyclonic type dust collectors. The system is to be
arranged such that change over to a new system may be
made with a minimum of lost production.
As a matter of convenience and water conservation, most
kraft mills use contaminated condensate as a scrubbing
liquid in the kiln scrubber. The condensate system is
backed up with a fresh water supply. During time of
chemical unbalance or lack of supply of condensate, fresh
water is substituted as a scrubbing medium.
Any mercaptans which are emitted from the kiln stack come
from the contaminated condensate used as a scrubbing liquid
and makeup wash water in this system. Substitution of fresh
water in these systems would eliminate emission of mercaptans
from the kiln stack.
This substitution of fresh water as a scrubbing medium
appears simple and relatively inexpensive. However, it
must be emphasized that the lime kiln and causticizing
areas are the prime users of contaminated condensate
generated in other areas of the mill. If fresh water is
used in place of contaminated condensate, the load to the
waste treatment system will be increased by the amount of
water substituted.
The emissions from incineration of noncondensible sulfur
bearing compounds in the lime kiln are not being considered
in this particular control method.
Costs
The capital costs are based upon purchase of equipment and
installation of the new scrubbing system while the existing
system is in operation. Items included are as follows:
1. New Induced Draft Fan with Drive
2. New Scrubber, Mist Eliminator and Stack
3, New Pumps with Drives
4, Modification to Existing Hot Gas Ductwork
5-121
-------
Also included in the costs are new foundations and structures
required to support the equipment and demolition cost for
removal of the existing scrubbing system, once the new
system is placed in operation. Cost curves are presented
in Figures 5-62 and 5-63.
Effectiveness
A single method is considered in this study of replacing a
low efficiency cyclonic type dust collector with a high
energy Venturi scrubber. The new method is to have a 99.0
or 99.9 percent lime solids collection efficiency. AOE and
design efficiencies are considered to be the same.
1. The new system will be to replace an existing and
operating inefficient system.
2. The efficiency of solids collection of high energy scrub-
ber is a direct factor of the pressure drop taken across
the throat of the scrubber. Because of this item the
capital cost between a 99.0 percent and 99.9 percent
efficiency would be very similar. Equipment sizing would
be essentially duplicate in either case and the only.
variable would be a higher horsepower motor for driving
the induced draft fan. Variables in operating cost are
shown on the cost curves.
3. High efficiency scrubbers are important to the pulping
industry for recovery of soda. Volatile soda compounds
in the kiln exhaust gases are more difficult to recover
than solids but are also more valuable. Equipment
manufacturers presently estimate that a scrubber with
99.9 percent lime solids collection efficiency will
recover 90 to 95+ percent of the soda fume in the kiln
exhaust gases. 99.0 percent lime solids collection
should be equivalent to approximately 70 percent collec-
tion of soda fume.
5-122
-------
8
o
o 12
5 10
o
o
—' o
£ 8
I—I
CL.
5 6
o
*
o
I—I
X
I/)
0
0
4
6
5
4
3
6 9
ADT/DAY X 100
12
FIG. 5-62CONTROL METHOD COSTS FOR FRESH WATER
VENTURI ADDED TO LIME KILN —99.0% LIME SOLIDS
COLLECTION
5-123
-------
o
o
o
to
o
o
o.
5
14
10
8
6
o
o
o
I/)
S 4
6 9
ADT/DAY X 100
12
FIG. 5-63 CONTROL METHOD COSTS FOR FRESH WATER
VENTURI ADDED TO LIME KILN —99.9% LIME SOLIDS
COLLECTION
5-124
-------
5.3.6.5 Lime Slaker
Application
Two control methods are considered for this source of
emissions, addition of a cyclonic scrubber and addition
of mesh pads. Each application is predicated on instal-
lation of a control device on an existing slaker which
has no control equipment. Flows from slaker stacks are
assumed to be saturated air containing calcium oxide
and inert particulates.
Costs
The capital costs are based on installing and operating
new control equipment. Demolition of existing vent stacks
is included. The estimated systems consist of the
following:
1. Figure 5-64
New Cyclonic Scrubber
Induced Draft Fan with Drive
Liquid Supply Pump with Drive
Vent Stack and Ductwork
Structural Steel Supports, Foundations,
Piping, Controls and Wiring
2. Figure 5-65
Mesh Pad Mist Eliminator with Housing
Indusced Draft Fan with Drive
Liquid Supply Pump with Drive
Vent Stack
Structural Steel Supports, Foundations,
Piping, Controls and Wiring
5-125
-------
— x CD
0
O
O
-" 32
X
^ 28
o
— ' 24
t — i
Q.
5 20
— i
£
fe
i- 16
/
/
/
'
/
/
o
0
°
• 12
to
o
10
8
6
6 9
ADT/DAY X TOO
12
FIG. 5-64 CONTROL METHOD COSTS FOR CYCLONIC
SCRUBBER ADDED TO SLAKER
5-126
-------
o
o
o
35
30
o
CJ
20
15
10
«
o1
o
5 7
X
**• c
0
Ul
0
^ 5
•=>
< /I
t—
UJ
z
X
x
x^
X
X
y
X
X
6 9
ADT/DAY X TOO
12
FIG. 5-65 CONTROL METHOD COSTS
FOR MESH PAD ADDED TO SLAKER
5-127
-------
Effectiveness
No measure of the actual effectiveness of either control method
has been developed to date. Further, no accurate measurement
of the quantity or quality of pollutants from lime slaker vent
stacks has been made. The volume and nature of these materials
have been assumed from knowledge of the chemical reaction taking
place and from laboratory data.
5.3.6.6 Power and Combination Boilers, General
The emission control equipment used in the steam power plants
has been almost exclusively limited to removal of particulate
matter. Lately efforts have been made to find ways and means
to remove sulfur from the fuels and alternatively sulfur dioxide
from the flue gases. As yet no commercially economic and
feasible method is in operation, but a number of promising pilot
studies have been made. The recognition of nitrogen oxides as
undesirable compounds in the flue gases is relatively new and
efforts in that direction have only begun.
Power and combination boilers in the pulp and paper industry
incorporate a wide variety of firing equipment. Some of these
types are:
Oil and Gas Burners
Pulverized Coal
Spreader Stokers with Traveling Grates
Spreader Stokers with Vibrating Grates
Spreader Stokers with Water Cooled Grates
Spreader Stokers with Stationary Grates
Spreader Stokers with Dump Grates
Dutch Ovens
Aside from coal the main fuels are oil, gas, and bark. Today
the burning of bark for steam generation is confined to locations
where bark is available as a waste or by-product. In producing
5-128
-------
lumber half of the wood in the log is often discarded as
saw-dust, bark, shavings, slabs and ends, all of which
can be used as fuel. New methods in barking and pulping,
however, are continuously reducing the quantity of waste
wood available.
The devices commonly used for removal of particulate
matter in flue gases of steam boilers are centrifugal
collectors and electrostatic precipitators'. The following
list indicates the type of dust collectors normally
used for various types of fuel.
Mechanical Electrostatic Precipitators
Coal (pulverized) X X
#6 Oil X Has been used but is not common
Gas
Bark X
Batk + Coal X
Bark -i- Oil X
Bark + Gas X
Fly ash may be approximately classified according to type of
firing as shown in Table 5-1. Ply ash from pulverized coal
fired boilers is generally quite fine—about 70 percent
(or more] samller than 325 mesh. It is usually light gray
in color as it contains only a small amount of unburned carbon.
The samll particles, having a large surface area, will impart
a definite color to the stack gases. This may, under certain
light conditions, indicate that a sizable amou'ht of fly ash
is being discharged into the atmosphere. Ply ash from stoker
fired boilers on the other hand is much coarser and has, since
it contains larger quantities of unburnt carbon, a dark gray
to black color, This fly ash does not discolor the flue gas
in the same manner as described above. A clean appearing
stack discharge from a stoker fired boiler does not necessarily
indicate that there is no emission problem. The coarse particles
will fallAout relatively close to the stack and on an equal
weight basis will cause a greater local nuisance than fly ash
from pulverized coal, Thus fine particles make a small plume
appear much denser than coarse particles„ It is for this
reason that removal of particles larger than 20 microns has
little effect on rhe appearance of the plume.
5-129
-------
TOTAL
TABLE 5-1
TYPICAL PARTICLE SIZE DISTRIBUTION OF FLY ASH FROM BOILER GASES OF VARIOUS BOILERS
(Percent by Weight)
Particle
Size
Microns
0-10
10-20
20-30
30-40
40-74
74-149
+ 149
Pulverized
Coal
25%
24
16
14
13
6
2
COAL FIRED BOILER
Cyclone Spreader Traveling
Furnace Stoker Grate
72% 11%
15 12
6 9 11
2 10
12 12
5 17 30
29 47
Underfed
Stoker
7%
8
6
9
8
19
43
100 Percent Bark Fired
Boiler
12%
10
7
6
14
16
35
100
100
100
100
100
100
Data frow Bituminous Coal Research Association, Pittsburgh, Penn. and the Industrial
Gas Cleaning Institute, "Criteria for the Application of Dust Collectors to Coal
Fired Boilers."
-------
Electrostatic Precipitators
The performance of an electrostatic precipitator is directly related
to the level of power input maintained. The amount of electric
charge that the dust particles will acquire depends upon the
conductivity or resistivity of the dust. When the resistivity
exceeds approximately 2 x 1010 Ohm centimeters excessive sparking
is to be expected with subsequent drop in collecting efficiency.
Research has shown that the resistivity of the fly ash is closely
related to the presence of water, soluble sulfates, and free
sulphuric acid formed on the surfaces of the dust particles by
absorption of SO_ from the flue gas. If the sulfur content is
greater than 1 percent in the coal there is likely to be sufficient
SO in the flue gas to create favorable conditions for an acceptable
resistivity. However, SO is formed by oxidation of SO which
occurs between 750° and 1500° F. The amount of SO formed will
depend on the design of the boiler and the rate of combustion. The
lower limit of soluble sulfur compounds in the coal ash, necessary
to create an acceptable resistivity has been found to be between
0.5 and 1.0 percent. The resistivity also varies with the flue
gas temperature and reaches a peak at approximately 300°F for average
fly ash. The peak resistivity of some fly ash is high enough
to adversely effect the precipitator performance. Yet 300°F is a
favorable temperature level for boiler and evaporator operation.
Thus, requirements for the boiler and for the precipitators are
contradictory. The amount of magnesium and aluminum in the coal
ash is also believed to be a factor contributing to high resistivity.
Fly ash from strip mined coal usually has higher resistivity than
fly ash from deep mined coal.
Fly ash containing more than 15 percent carbon will effect pre-
cipitator performances adversely. The reason for this is that
carbon is a relatively good conductor and will acquire, but
also loose, an electrical charge very easily. When a particle
with high carbon content precipitates on the collector plate
it loses its charge immediately. Owing to a phenomenon referred
to as "charge by induction," the particle assumes a weak positive
charge. It is repelled from the collector plate and reentrains
into a gas stream. This phenomenon is repeated over and over
again until the particle is carried out by the gas stream and
escapes from the precipitator. In order to overcome this problem,
the collector plates are given a special form and the so-called
pocket or screen grid collecting electrodes have proven to be
very effective. There is also a certain correlation between the
carbon content and the coarseness of the fly ash. A coarse fly ash
has normally a high carbon content.
5-131
-------
Recent improvements in the design of electrostatic precipitators
has reduced many of the advantages of combinations of electro-
static precipitators and mechanical collection systems, so common
some years ago. The plate electrode precipitators did not have
sufficient collecting efficiency on high carbon fly ash with
problems of "snow outs" from the stack during soot blowing
and collector plate rapping periods. Pocket electrodes and
continuous rapping have, to a large extent, circumvented these
problems. Another advantage of a combined system installation
was that the mechanical collector would still remove a major
portion of the fly ash should the precipitator be out of
order for some reason. Sectionalized design and more reliable
rectifiers have shortened the shut-down time considerably on
modern precipitators.
Mechanical Collectors
Bark char is a difficult material to collect. While the particle
size is relatively large, the specific gravity is very low and
the sliver shape of the particles makes collection difficult.
Bark char has a specific gravity in the 0.2 range; pulverized
coal ash for example is 10-15 times heavier. In addition to
the unique particle shape, the low specific gravity and the
very fragile particles, the bark boiler collector must also be
able to handle abrasive sand accompanying the bark. There are
consequently a number of considerations to make when selecting
a collector for a bark boiler.
Electrostatic precipitators are not generally suitable for use
on bark boilers due to (1) the poor electrical characteristics
of bark char and (2) the possibility of fires.
Some studies report that the bark char disintegrates
into fine particles in a multi-tube collector due to
the high centrifugal forces imposed. These fine
particles coupled with their low specific gravity have
a tendency to float through the outlet tube of the
collector and out the stack. This phenomenon is
also attributed to the short distance from the wall of
the collecting tube to the outlet tube. A survey shows
that there are more large diameter cyclone instal-
lations than multi-tube units operating on bark boilers
in the wood pulping industry. It should be noted, how-
ever, that both collection devices can give sufficient
efficiency providing they are selected properly.
5-132
-------
Many bark char collectors in the southern part of the
United States are operating with reinjection systems
that are not equipped with provisions for separation
of sand from the char before reinfection. Sand sepa-
ration is essential for reducing excessive erosion
of the collector and the boiler tubes. It is not
unusual for a system operating on bark from southern
pine to accumulate a number of truck loads of sand
in an operating day. Some mills are now selling the
collected bark char for manufacturing of charcoal
briquettes.
In many multi-tube collectors recovery vanes are
installed to reduce pressure drop and collector
size. The experience shows that collectors operating
on boilers firing oil in combination with bark must be
designed without recovery vanes. Some oil fractions
having high dewpoints combine with the bark char
particles and plug the recovery vanes. One single
plugged outlet tube can by-pass some 10 times its
normal share of flue gas uncleaned into the stack.
It is consequently imperative to prevent plugging
of the tubes.
Studies show that substantial improvement of the
collecting efficiency of bark char collectors can
be attained by changing the normally conical bottom
outlet of the cyclone (tube). By using straight
cylindrical tubes with a peripheral discharge rather
than conventional conical discharge, the ash is re-
moved from the tube before the gas reverses into the
inner vortex and in sufficient distances from the
turning point. This prevents re-entrainment of
already collected particles.
The volume handled by a given diameter collector
tube depends on the shape of the inlet guide vane.
Shallow pitched vanes give more spin and handle
less volume. Steep pitched vanes reduce the spin
and can handle more volume„ The shape of the
inlet guide vane can make a difference in volume
capacity of the tube at a ratio of 1 to 2. The
vanes are of obvious importance for collector
size as well as operating efficiency.
5-133
-------
Bark char collectors should be easily accessible
for inspection, maintenance and repair because of
the difficult operating condition under which they
work. One always has to bear in mind the abrasive-
ness of the dust and possibility of clogging. Large
diameter cyclone collectors are easily accessible
for maintenance. Some modern multi-tube collectors
also offer design with accent on maintenance. The
place subject to wear in a well designed collector
is limited to the collection tube or cyclone cone.
Hard cast iron is often used in tubes. Hardness
of 420 Brinell has proven to be a maximum, because
hard tubes are subject to thermal shock damage.
Cyclone cones, as well as the entire cyclone, is
sometimes furnished with abrasion resistant linings.
This lining is usually installed over a hexagonal
steel liner.
Both mechanical collectors and electrostatic precipi-
tators are sensitive to uneven gas and dust distribution.
The entrance to the collector is normally provided
with devices for correction of uneven gas distribution.
The space for proper distribution ahead of the collector
is almost always very limited. Sharp bends and short
transitions tend to produce uneven distribution across
the entrance. Various devices such as spreaders, turn-
ing vanes and perforated baskets in the collector en-
trance are used in efforts to overcome these problems.
Hopper fires occur in bark char collectors. Unlike
ash from coal fire boilers, bark char must be con-
tinuously and completely removed. A small air leakage
is enough for accumulated char to catch fire. Gas
tight construction is essential in bark char collectors.
Design must consider the possibility and effect of
fires. This removal from the hopper must be accom-
plished without a back flow of air. This can be done
must successfully with rotary steel valves.
Bark boiler dust collectors should be carefully sized
for correct pressure drop inlet velocity. While ef-
ficiency generally increases with increased pressure
drop (and velocity) there are limits beyond which
fragile material like bark char will be broken into
smaller particles which has a detrimental effect on
the collecting efficiency.
5-134
-------
The low specific gravity of bark char, when the boiler
is operated on 100 percent bark without reinjection, or
the finer particle size distribution when the ash is
reinjected, create operating conditions which limit
the application of large diameter cyclones on these
installations.
Only when other conditions are superimposed over these
conditions can the large diameter cyclone approach 92
percent efficiency in a single stage unit. One of the
conditions which will increase the efficiency is the
amount of sand in the bark char.
When operating on 100 percent bark, the efficiency
will vary greatly depending upon the amount of sand
imbedded in the bark. Without reinjection, much of
the imbedded sand remains in the bark char. This
greatly increases the average specific gravity of
the bark char, making the centrifugal separation
easier with higher efficiencies.
Example: 80% Bark Specific Gravity (as char) 0.3
20% Sand Specific Gravity 2.8
S. G. = (0.8) (0,3) + (0.2) (2.8) = 0.24+0.56 = 0.80
For a large diameter cyclone where the specific gravity
is assumed at 0.3, the most practical unit would give
efficiencies in the 70 to 75 percent range. If, as in
the example above, the char contained 20 percent sand,
the efficiency would increase to 85 to 89 percent.
The performance of mechanical collectors, with 40 percent
bark and 60 percent oil feed to the boiler, is
greatly improved due to the higher specific gravity
of the flyash. The nature of this dust, however, in
multi-tube collec-cors creates a plugging problem.
The higher specific gravity and tendency not to plug
make the large diameter cyclone practical for combina-
tion fired boilers. It is possible to achieve the
92 percent design efficiency in both multi-tube and large
diameter cyclones with single-stage units. The 96
percent design efficiency is possible with two-stage large
diameter or multi-tube units.
5-135
-------
5.3.6.7 Combination Boilers
Application
The following methods are considered for reducing emissions
from an existing combination boiler:
1. Replace existing multi-tube collector
with new design multi-tube collector(s).
2. Add a cyclonic scrubber in series with
an existing 85 percent dust collector.
3. Addition of surge bin to reduce peak
bark flows from the woodyard.
4. Convert 100 percent reinjection system
to 70 percent or less.
The application of equipment for the first two methods has
been discussed in the preceeding section.
Addition of Surge Bin to Reduce Peak Bark Flows. Particulate
emissions from a combination boiler burning bark and oil are
approximately proportional to the rate of bark feed to the boiler,
assuming other variables are approximately constant. In
order to minimize particulate emission and insure good combustion
conditions to the furnace, current practice is to install
a surge bin with a new combination boiler. Many existing mills
feed bark directly from the woodyard to the boiler without a
surge bin. For a number of reasons, the bark flow rate from a
woodyard fluctuates widely, and bark supply does not always
coincide with fuel demands.
The installation of a properly sized and designed surge bin
will provide a relatively constant bark flow to the combination
boiler. The surge bin size may vary considerably for a particular
boiler depending upon the relative woodyard debarking capacity
in relation to the particular boiler's bark burning capacity.
While surge bins are usually installed to obtain more efficient
use of the combination boiler and better combustion conditions, a
definite reduction in particulate emission will also result.
Where particulate emissions must be reduced from a combination
boiler, the installation of a "surge bin" should be
evaluated.
Convert 100 Percent Reinjection System to Approximately 70
Percent or Less. Combination boilers burning bark and oil
may utilize bark ch*ar' reinjection from the boiler hoppers
(located below the mud drum), dust collectors hoppers, and/or
ductwork hoppers. The bark char results from the incomplete
5-136
-------
combustion of the bark fuel. Reinjection is practiced
primarily to eliminate disposal problems with bark char
and to obtain the additional heat from burning the char.
Current practice in the design of combination boilers
is to use reinjection systems which reinject 70 percent
or less of the collected bark char. A few older mills
still practice 100 percent reinjection. An assessment of
the impact.of.reinjection has been investigated in detail
by Mullen — . In essence, 100 percent reinjection of
ash particles results in a continuous cycling of the ash.
This continual cycling degrades the ash particle size
until the ash is emitted from the stack, thereby increasing
particulate emission. For bark which contains extremely
low ash (less than 1.0 percent) the influence of 100
percent reinjection may be minor. This is especially true
where the bark and bark char are burned in refractory furnaces.
Where particulate emission from a combination boiler is
high and 100 percent bark char reinjection is practiced, reduction
of reinjection to 70 percent or less should be evaluated. The
cost of reducing the percentage reinjection will vary considerably
from mill to mill, and should be developed on an individual mill
basis. The cost will be expected to be very much higher where
an ash disposal system must be installed.
Costs
New Design Multi-tube Cyclones. This method is based on the
replacement of an existing relatively low efficiency, multi-
tube cyclone collector with a new design, high efficiency,
multi-tube cyclone system, two cases being considered. The
first case assumes the replacement of the existing system
with a single stage system, and the second case assumes
the replacement of the existing system with a two-stage
system. Both are shown in Figure 5-66.
It was assumed that 40 percent of the heat input to the
boiler was from bark and 60 percent from oil.
Capital costs are based on mild steel construction for the collector
and duct work. Erection costs are based on components being shipped
knocked-down. It was also assumed that existing sand classifiers
could be used on the new single stage unit and that classifiers
would not be required on the secondary stage collector. The ash
collected in the second stage would be discharged directly to
the existing ash handling system.
5-137
-------
INGLE UNIT
TWO UNITS IN SERIES
. ,0 CE REMOVED
ADDED EQUIPMENT SHOWN WITH HEAVY LINES
FIG. 5-66 TYPICAL ARRANGEMENT FOR REPLACING AN EXISTING
SINGLE STAGE MULTI-TUBE COLLECTOR WITH A NEW SINGLE STAGE
OR TWO STAGE MULTI-TUBE COLLECTOR - COMBINATION BOILER
5-138
-------
Capital costs and net annual cost are presented in Figures
5-67 and 5-68. For a specific combination boiler, it should
be recognized that arrangement and location of replacement
collector system would vary widely, but the prepared estimates
are typical of the costs that could be encountered.
Cyclonic Scrubber. This control method is based on the
installation of one or more cyclonic scrubbers depending
on combination boiler exit gas volume, with the scrubbers
in series with and downstream of the existing dust collector
as shown in Figure 5-68.
In addition to the scrubber, a circulation pump and motor
for each scrubber, a strainer for removal of ash and char
from cyclone effluent, required piping for make-up, circu-
lation, and effluents to ash lagoon and waste disposal
system, reivsed ductwork to connect the scrubber with the
existing system, and instruments and controls for
proper operation of the above are included.
The ash and char may be collected in containers or discharged
into the existing ash handling system. For the purposes of
this method, the ash and char is assumed to be discharged to
the existing ash handling system.
The capital cost is based on mild steel construction with
concrete lining for the cyclonic scrubber and ductwork as
this would be the most representative material for this service.
Cost curves are shown in Figure 5-<70.
Effectiveness
Particulate Removal. Efficiencies for particulate removal
are predicted as follows:
Control Method Annual Operating Efficiency
Single-stage Collector 90.0%
Two-stage Collector 94.0%
Cyclonic Scrubber 94.0%
5-139
-------
300
o
o
0
-" 250
X
~ 200
o
CJ
-j 150
H-
Q-
5 100
_i
P en
^
^
^
^
^
^
o
8 65
55
00
S 45
35
25
10 20 30 40
GAS VOLUME (CFM X 10,000)
FIG. 5-57 CONTROL METHOD COSTS
REPLACING AN EXISTING MULTI-TUBE
COLLECTOR WITH A SINGLE-STAGE
COLLECTOR-COMBINATION BOILER
(90.0% A.O.E.)
FOR
5-140
-------
o
o
o
oo
o
0.
«c
o
g
500-
440
380
320
260
200
X— X '
CD"
0
o
- 150
X
•fa"*
~125
i/>
o
«_)
5? TOO
rs
? »
UJ
z
en
X
x
r
s
/
/
/
y
X
10 20 30 40
GAS VOLUME (CFM X 10,000)
FIG. 5-€e CONTROL METHOD COST FOR REPLACING AN
EXISTING MULTI-TUBE COLLECTOR WITH A TWO-STAGE
COLLECTOR - COMBINATION BOILER (94.OX A.O.E.)
5-141
-------
ADDED EQUIPMENT SHOWN WITH HEAVY LINES
FIG. 5-69 TYPICAL ARRANGEMENT FOR CYCLONIC SCRUBBER
ADDED IN SERIES WITH AN EXISTING 85X EFFICIENT DUST
COLLECTOR - COMBINATION BOILER
5-142
-------
o
o
o
I
380
320
260
200
140
175
150
125
100
75
">0
/
X
/
x
'
x
f
X
X
x
/
X
/
10 20 30 40
GAS VOLUME (CFM \ 10,000)
FIG. 5-70 CONTROL METHOD COSTS FOR A CYCLONIC
SCRUBBER ADDED TO AN EXISTING 85X EFFICIENCY
DUST COLLECTOR - COMBINATION BOILER (94.0* A.O.E.)
5-143
-------
Particulate removal efficiencies for the remaining methods
are not predictable.
Curves for particulate cost effectiveness are presented in
Figures 5-71 and 5-72.
Reduced Sulfur and sulfur dioxide Removal. None of the
methods considered would be effective in reducing emission
of sulfur dioxide.
Operation. In comparing operation the cyclonic scrubber
would result in more corrosion potential and might also
be more prone to plugging. In considering surge bins,
arching of the bark can be an operating problem.
Summary. Due to the range in efficiencies and costs, the
selection of the most effective method would have to be
determined for each individual case.
5-144
-------
50
c
o
< "
<_> o
10
I
50
S
40
£0"
o o
•—I
—I X
90 ?2 !
% A.O.E.
100,000 CFM
10
I
or>
% A.O.E.
200,OPO CFM
50
CO
o
<_>
c^ o
r-4
_J X
c.
10
B <
A -
' C
CP; O2 *Vv.
% A.O.E.
400,000 CFM
A. CYCLOHIC SCP.UEBER
ADDITION TO /V? EXISTING
05% DUST COLLECTOR
B. REPLACE EXISTING f'ULTI-TUBE
DUST COLLECTOR WIT! I A
SINGLE-STAGE DUST COLLECTOR
C. REPLACE EXISTIfiG MULTI-
TUBE DUST COLLECTOR WITH A
DOUBLE-STAGE MIJLTI-TUGE
DUST COLLECTOR
FIG.5-71 PARTICULATE COST EFFECTIVENESS (TOTAL
CAPITAL COST COMPARISON) FOR CONTROL EQUIPMENT
ADDED TO COMBINATION BOILER
5-145
-------
20
o
_J O
-------
5.3.6.8 Power Boilers
Application
This case is based on the addition of an electrostatic
precipitator following an existing dust collector for
a coal-fired power boiler. Two precipitator sizes
are considered.
Costs
Estimated costs are based on the equipment arrange-
ment shown in Figure 5-73. Curves for capital costs
and net annual operating costs are presented in Figures
5-74 and 5-75.
The equipment costs are based on a double chamber, steel
shell precipitator. All labor and material for a complete
installation are included in the capital cost and the
net annual operating cost. The precipitator size is
based on firing coal containing 2 percent sulfur.
Effectiveness
Particulate Removal. Guaranteed particulate efficiencies
for the two precipitators considered are 99.0 percent and
90.0 percent. The resulting total annual operating
efficiencies are:
Precipitator Precipitator Dust Collector Total
Guaranteed Efficiency APE APE APE
99.0% 98.0% 80.0% 99.0%
90.0% 89.0% 80.0% 98.0%
Reduced sulfur and Sulfur P.ioxide Removal. Precipitators
are ineffective for removing sulfur compounds.
Pperation. The operating characteristics of the precipitators
should be identical.
5-147
-------
ADDED EQUIPMENT SHOWN WITH HEAVY LINES
FIG. 5-73 TYPICAL ARRANGEMENT FOR PRECIPITATOR
ADDED IN SERIES WITH AN EXISTING DUST COLLECTOR -
COAL FIRED POWER BOILER
5-148
-------
£• 70
o
- 60
X
*••»•
CO
o
50
•
40
5 30
^ 20
18
§16
X
~14
V)
8 12
8
8 10 16 20 24 28 32
GAS VOLUME (CFM \ 10,000)
FIG. 5-74 CONTROL METHOD COSTS FOR 99.0% EFF.
ELECTROSTATIC PRECIPITATOR ADDED TO AN EXISTING
90% EFF. DUST COLLECTOR COAL FIRED POWER BOILER
(99.0% A.O.E.)
5-149
-------
o
o
50
CO
o
o
ex
o
g
44
38
32
26
14
o
o
o
12
o
X
s 10
I-
8 8
3
8 12 16 20 24 28
GAS VOLUHE (CFM X 10,000)
32
FIG. 5-75 CONTROL METHOD COSTS FOR 90.0% EFF.
ELECTROSTATIC PRECIPITATOR ADDED TO AN EXISTING
90% EFF. DUST COLLECTOR COAL FIRED POWER BOILER
(98.0% A.O.E.)
5-150
-------
5.3.7 SULFITE SOURCES
The evaluation of control methods for sulfite sources is
limited to those which are distinct from the kraft sources.
The reader is referred to the previous discussions of kraft
sources for the following sulfite sources: Washer vents,
evaporators, power boilers, and combination boilers. It
should also be noted that control methods have not been
considered for the following sulfite sources: Dump tank
vent, Venturi absorbers, absorption towers, and ammonia
incineration. These sources are either of minor impor-
tance from an emission standpoint, or there are no control
methods presently in use.
5.3.7.1 Acid Tower
Application
The pressure accumulator is vented into the acid storage
tank because the two tanks are nearby. The acid storage
tank is then vented to the absorption tower. This system
effectively prevents emissions from these items of equip-
ment, thereby leaving the acid absorption tower as the
only significant source of emission in the acid system.
The construction of most absorption towers incorporates
a mesh pad distribution system for absorption liquid.
Therefore, additional mesh pads are deemed necessary.
The efficiency of absorption of most sulfiting towers
exceeds 90 percent. Some mills have placed a second ab-
sorption tower in series with the sulfiting unit for
scrubbing exhaust gases. This method has been tried only
where an existing second tower was available and could be
used economically. An example would be the conversion of
calcium base, requiring two towers, to ammonium base
which requires one tower.
A cost estimate has been prepared for installation of a
second packed tower in series with the absorption towers.
Costs
The capital costs are based on purchase and installation
of a new absorption tower in series with an existing acid
absorption tower. Costs are predicated on installation of
the new tower in an existing mill with a minimum of down
time for placing the new tower on stream. Capital and net
annual costs are shown in Figure 5-76.
5-151
-------
CD
O
X
•faO-
C/1
o
o
15,
13
11
9
Q-
5 7
S
6 9
ADT/DAY X 100
12
FIG. 5-76 CONTROL METHOD COSTS FOR
PACKED TOWER ADDED TO ACID TOWER
5-152
-------
Effectiveness
Data are not presently available on economic feasibility or
efficiency of recovery of this system.
5.3.7.2 Blow Pit
Application
Two systems were considered to replace the existing multiple
wooden blow stacks and showers with a high efficiency SO
recovery system as follows:
1. Condenser with Cyclone and Absorption Tower
2. Packed Tower
Flash steam, sulfur dioxide, and inert gases are released in
the blow pit during a digester blow. These gases then exit
from the blow pit and after the steam is condensed, the non-
condensible gases, mainly sulfur dioxide, are absorbed in a
packed tower. The recovered sulfur dioxide is reused in the
process and the condensate creates a source of hot water.
Costs
The costs of these controls are comprised chiefly of the
purchase and installation of equipment for condensing the
flash steam and a packed tower for absorbing the sulfur
dioxide. In addition, the costs include piping for con-
veying the gases from one stage to another and carrying the
water required for gas absorption and heat transfer. The
costs for these control methods are "Savings" not costs
and are shown in Figures 5-77 and 5-78.
Effectiveness
Particulate Removal. There are no particulate emissions
from the above source.
Reduced Sulfur and Sulfur Dioxide Removal. There are no
reduced sulfur emissions from the above source. Both sys-
tems are 95 percent efficient in absorbing sulfur dioxide
from the blow pit stacks, where the SO concentration is
approximately 4 percent during the initial stage of the
blow. This produces recovery water containing 0.85 percent
sulfur dioxide. The recovery water is subsequently used
directly in the fortifying of the cooking acid or the SO
is stripped from the recovery water and used in the manu-
facture of cooking acid.
5-153
-------
§85
o
o
565
Q.
O
=! 55
45
o
3
e" 10
15
20
25
30
4 6
ADT/DAY X 100
8
FIG.5-77 CONTROL METHOD COSTS FOR
BLOW PIT - CONDENSER WITH CYCLONE
AND PACKED TOWER
5-154
-------
^ 40
o
o
o
o 35
r-1
X
to
o
30
g 25
i—t
CL.
" 20
o
- 15
o
o
o
«t 5
oo °
4 6
ADT/DAY X 100
FIG.5-78 CONTROL METHOD COSTS FOR BLOW
PIT - PACKED TOWER
5-155
-------
Operation. System (a) may have more problems than (b) because
of additional equipment and instrumentation. System (b) consists
mainly of a gas header and an absorption tower, and is, therefore,
relatively trouble free in operation.
Summary. The capital cost of System (a) is approximately three
times the cost of System (b), both having a sulfur dioxide recovery
efficiency of 95 percent. Although System (a) has a higher heat
recovery efficiency, it would appear that System (b) ,: Packed Tower,
is more effective.
5.3.8 NSSC SOURCES
Evaluations of control methods applied to NSSC sources are not
included due to the lack of emission data and application experi-
ence. The reader is referred to the previous Kraft discussions
for the following NSSC sources: Washer vents, evaporators, combi-
nation boiler, and power boiler.
5-156
-------
11. Mullen, J. F., "A Method for Determining Combustible
Loss, Dust Emission, and Recirculated Refuse for a
Solid Fuel Burning System," Paper presented at ASME
Winter Annual Meeting, New York, November 29 -
December 4, 1964.
12, Wrist, Peter, Mead Corporation, Verbal Communication,
Liaison Meeting, New York, November 6, 1969.
13. Carlton-Jones, Dennis and Schneider, H.B., "Tall
Chimneys," Chemical Engineering, 75, October 14, 1968.
14. Martin, F., "Secondary Oxidation Overcomes Odor from
Kraft Recovery," Pulp and Paper, 43, June 1969.
15. Industrial Gas Cleaning Institute, "Test Procedure
for Gas Scrubbers," Publication No. 1, IGCI, Box 448,
Rye, New York, 1964.
16. Industrial Gas Cleaning Institute, "Criteria for
Performance Guarantee Determinations," Publication E-P3,
IGCI, Box 448, Rye, New York, 1965.
17. NAPCA, "Tall Stacks - Various Atmospheric Phenomena and
Related Aspects," Publication No. APTD 69-12.
18. Wrist, Peter, Mead Corporation, Personal Communication,
Sept. 2, 1969.
19. Blosser, R. O., and Cooper, H.B.H., "Current Practices
in Thermal Oxidation of Noncondensible Gases in the
Kraft Industry," Atmospheric Pollution Technical
Bulletin No. 34, National Council for Air and Stream
Improvement, Inc., New York.
5-158
-------
5.5 REFERENCES
1. Shah, I. S. and Stephenson, W.D., "Weak Black
Liquor Oxidation System: Its Operation and
Performance," TAPPI 51 (a), 87-, 1968.
2. "Steam, Its Generation and Use," 37th Edition,
The Babcock and Wilcox Company, New York.
3. Fryling, G. R., "Combustion Engineering," Revised
Edition First Impression, Combustion Engineering,
Inc., New York.
4. MacDonald, R. G., Editor, "Pulp and Paper Manu-
facture - Volume I, The Pulping of Wood," Second
Edition, McGraw-Hill, New York 1969
5. Whiteny, R. P., Editor, "Chemical Recovery in
Alkaline Pulping Processes," TAPPI Monograph Series
No. 32, TAPPI, New York, 1968
6. Theon, G. N., DeHaas, G. G., Tallent, R. G., and
Davis, A. S., "The Effect of Combustion Variables
on Release of Odorous Compounds from Kraft Recovery
Furnaces," TAPPI 51 (8), 329-, 1968.
7. Harding, C. I. and Galeano, S. F., "Using Weak Black
Liquor for Sulfur Dioxide Removal and Recovery,"
TAPPI 50 (10), 48A-, 1967.
8. Tomlinson, G. H., Chapter 5, page 419, "Pulp and
Paper Manufacture - Volume I, Preparation and
Treatmenr of Wood Pulp," First Edition, McGraw-Hill,
New York, 1950.
9. Thomas, E., Broadus, S., and Ramsdell, E. W., "Air
Pollution Abatement at S. D. Warren's Kraft Mill in
Westbrook, Maine," TAPPI 50 (8), 81a-, 1967.
10. Hough, G. W. and Gross, L.J., "Air Emission Control
in a Modern Pulp and Paper Mill," American Paper
Industry, 36-, February 1969.
5-157
-------
CHAPTER 6
NEW DEVELOPMENTS IN CONTROL TECHNOLOGY
TfiBEE OF CONTENTS
Page No.
Summary 6-1
Introduction 6-2
General Description of Control Methods 6-2
Conventional Kraft Recovery Systems 6-3
New Kraft Recovery Systems 6-4
Sulfite Sources 6-9
NSSC Sources 6-9
Application, Cost, and Effectiveness of New
Control Methods 6-10
Kraft Sources 6-10
Recovery Systems 6-11
Oxidation Systems 6-26
Smelt Dissolving Tanks 6-28
Digester Relief and Blow, Multiple
Effect Evaporators 6-30
Combination Boilers 6-33
Brown Stock Washer Vents 6-35
Multiple Effect Evaporators 6-38
Miscellaneous 6-39
Sulfite Sources 6-40
NSSC Sources 6-42
References 6-45
6-i
-------
CHAPTER 6
NEW DEVELOPMENTS IN CONTROL TECHNOLOGY
SUMMARY
New control methods evaluated in this chapter are those
which have had limited or no application in the United
States, but which may show promise for more economically
or effectively reducing atmospheric emissions. Only
methods which promise improvement over existing technology
have been evaluated in depth. Due to the indeterminate
characteristics of some of the new methods, only order of
magnitude costs have been estimated in order to make a
relative comparison of the new techniques.
The following systems have been evaluated:
Kraft Recovery Furnace
Black Liquor Oxidation Unit
Smelt Dissolving Tanks
Digester Relief and Blow
Plus ME Evaporators
Combination Boilers
Brown Stock Washers
Ammonium Sulfite Liquor
Conversion to High Solids
System
Conversion to Air Contact
Evaporator System
Sodium Carbonate - Bicarbonate
Scrubber System
High Solids Plus Brine
Scrubber System
Weak Black Liquor Oxidation
(Molecular Oxygen)
Packed Tower Scrubber
Orifice Scrubber
Incineration in Recovery Boiler
Separate Thermal Oxidation
Mechanical Collector Plus
Shave-off Scrubber
Continuous Diffusion Washing
Enclosed Pressure Washing
Incineration
Incineration in Combination Boiler
Incineration in Separate Boiler
6-1
-------
6.1 INTRODUCTION
In this chapter control methods which have had limited or
no application within the United States, but which may show
promise for economically and/or effectively reducing atmos-
pheric emissions are investigated. Only methods which indicate
possible improvement over existing technology will be evaluated
in depth. All methods in this chapter have not been proven in
U. S. practice and should be considered as experimental until
their practical performance has been demonstrated under full-
scale operating conditions.
Due to the indeterminant characteristics of new developments,
capital cost and net annual cost curves have not been prepared.
However, an assessment of capital cost and net annual cost have
been made to obtain order of magnitude costs for a relative
comparison of new developments. This order of magnitude cost
is used to compare the relative control effectiveness of new
developments with the present control methods described in
Chapter 5.
The order of magnitude capital costs and net annual costs are
calculated for the following size mills which are considered
typical for the industry:
Kraft 500 ADT/Day
Sulfite 200 ADT/Day
NSSC 200 ADT/Day
Following the same procedure as Chapter 5, costs for "site
clearance" and "loss of production" are not included in any
calculations in this chapter.
6.2 GENERAL DESCRIPTION OF CONTROL METHODS
For a general description of control equipment, the reader
should refer to Chapter 5. The general descriptions included
in this chapter refer to "New Developments."
While capital costs for the new recovery designs were included
in Chapter 5, a detailed discussion of the two new systems
(High Solids and ACE) are included in this chapter. These two
new systems are classified as new developments; however, the
initial installations of these systems have just recently been
placed in operation (start-up was the last six months of 1969).
6-2
-------
Operating data from these first two systems were not available,
and, therefore, could not be included with this report.
6.2.1 CONVENTIONAL KRAFT RECOVERY SYSTEMS
The following new developments are applicable to a conventional
kraft recovery system with a direct contact evaporator.
6.2.1.1 Black Liquor Oxidation
Supplementing the discussion of oxidation systems in Section
5.3.6.1.5, the West Coast Canadian Kraft Industry reportedly
may sponsor a pilot plant study of a recovery flue gas scrubbing
system using a sodium carbonate-bicarbonate mixture at pH 9,4 and
180°F. However, the system design must also include an elaborate
oxidation system to oxidize the Na S formed since the liquor is re-
circulated to reduce the chemical cost. Until more details of this
work done by British Columbia Research Council are available, no
evaluations of the scrubbing liquor and the system can be made.
6.2.1.2 "Hot" Precipitators
The question has been raised as to whether the so-called "Hot"
precipitators would be of any interest to the U.S. pulp industry.
These precipitators have found acceptance in Scandinavia.
The most common location is downstream of the economizer—"warm"
precipitator^-but in at least twenty-five mills, the precipitator
is located before the economizer—"hot" precipitator.
The hot precipitator provides many advantages for the economizer.
Less cleaning is necessary. The cleaning is done by flushing the
economizer tubes with water. For the warm precipitator, the
maintenance cost of the economizer is high even when using hot
cleaning water. The reduced cleaning also means a longer life of
the economizer. The hot precipitator will reduce the solids in
the flue gas, which will keep the heat transfer coefficient at
the high level as well as the draft loss constant through the
economizer. According to the Scandinavian experience, the econo-
mizer can be made approximately 30 percent smaller when the
precipitator is located upstream (hot) due to constant heat
transfer coefficient.
The resistivity of the dust at the elevated temperature is con-
siderably lower than for conventional precipitator operation.
This would mean that the size of the precipitator could be smaller.
6-3
-------
The gas volume at. this temperature is, however, much larger and
the end result is that the hot precipitator has to be made
larger than the warm unit. The savings in the smaller economizer
will be offset by the larger precipitator.
The layout is somewhat simpler for the hot concept, which will
realize some savings, especially in Scandinavia, where most pre-
cipitator installations are inside the building. The structural
steel, however, has to be increased due to the increased load.
The hot precipitator is larger and heavier and is always located
above the boiler. If both precipitator designs are properly
maintained, the total evaluated cost of the "warm" precipitator
is approximately equal to the total evaluated cost of the "hot"
precipitator.
The high temperature creates more severe operating conditions,
and the dust has more sticky properties than for the warm precipi-
tator. For these reasons, this control method appears to have
limited application in the U.S.A.
6.2.1.3 Precipitator Installed Subsequent to Venturi Evaporator/Scrubber
There has been some discussion about adding an electrostatic
precipitator following a Venturi evaporator/scrubber on a recovery
boiler. This is a somewhat unorthodox approach, and there is
knowlingly none in operation, but one is to start up in 1970 in
the aluminum industry. According to the manufacturer, the corrosion
problem anticipated is overcome by making the precipitator shell in
tile, the collecting plates in treated wood, and all other internals
in 316 stainless steel. Further, there are no rapping mechanisms
because the manufacturer expects that considerable condensation
will take place on the collector plates which will flush down the
precipitated dust. The precipitator is further equipped with wet
bottom. The equipment cost is estimated to be approximately 30
percent higher than for a conventional precipitator installed on
the same recovery boiler.
Operating experience from this first unit will tell if this approach
is feasible. At the present time, this control method must be
classified as questionable.
6.2.2 NEW KRAFT RECOVERY SYSTEM
This section of the report describes the conversion of existing
recovery systems to the new designs which remove the direct contact
evaporator from the flue gas stream. These designs for new installa-
tions were generally described in Section 5.3.6.1.6 with capital costs.
6-4
-------
These new designs have a number of variations; however, they
may be generally classified into the following two groups:
1. High Solids Concentration'in the Multiple Effect
Evaporators. (High Solids System)
2. Air Contact Evaporator System (ACE)
More detailed analyses of the various reduced odor systems
have been reported by Arhippainen (3), Hochmuth (1) , and
Rob erson (4^) .
The description of these recovery systems is discussed as
follows:
High Solids System
ACE System
Multiple Effect Evaporators for High Solids or ACE Systems
Precipitators
6.2.2.1 High Solids Systems
This system is so called because the black liquor solids
leaving the multiple effect evaporators are in the range of 60
percent to 70 percent solids. Currently, the most acceptable
solids concentration is expected to be about 63 percent. The
liquor is then heated by direct steam heaters and fired into the
recovery furnace. (Figure 6-1)
By concentrating the liquor to a firing solids concentration of
63 percent, the conventional direct contact evaporator may be
removed from the flue gas stream. The economizer surface is
increased to reduce the flue gas temperature from approximately
600°F leaving the boiler to 350° - 400°F leaving the economizer.
For an existing unit, the flue gas temperature must be below
400°F. For an existing unit, the flue gas temperature must be
below 400°F if a tile-shell precipitator is on line. This can
usually be accomplished for most installations. However, in
cases of high temperature feed-water (in the order of 350°F and
higher), this gas temperature may be difficult to attain.
The high solids system has found extensive application in Europe
due to the high fuel cost. These systems were originally installed
in Europe to attain high thermal economy.
6-5
-------
6.2.2.2 Air Contact Evaporator System
This system has been described in more detail by Hockmuth (1). It
consists essentially of removing the conventional cascade evaporator
from the flue gas stream and placing it in the heated air stream as
shown in Figure 6-2. In order to reduce the temperature of the flue
gases to a reasonable level a regenerative type air heater is added.
The flue gases from the boiler at approximately 650°F, pass to the
air heater and are reduced to approximately 300 - 350°F. The heat
removed from the flue gases is transferred to the combustion air
which is heated from 80°F to approximately 550°F. The heated com-
bustion air then passes through the air contact evaporator where the
temperature is reduced to approximately 275°F before being introduced
into the furnace.
As the air temperature drops from 550°F to 275°F, the black liquor is
concentrated from 55 percent solids to approximately 65 percent solids
by evaporation. Any odorous gases which are entrained with the com-
bustion air are oxidized to non-odorous compounds in the furnace.
For the best thermal economy, the black liquor is usually concen-
trated to 55 percent solids in the multiple effect evaporators.
The application of the ACE system to existing units will usually
require increasing the building size. Therefore, the cost for site
clearance will probably be very expensive. In many cases, it may
well be impossible to install the ACE system at an existing mill.
6.2.2.3 Multiple-Effect Evaporators
The multiple-effect evaporators at existing mills are usually de-
signed to produce 50 percent black liquor solids„
If the High Solids System is installed at an existing mill, an
additional effect, usually referred to as a "concentrator," is
added to provide 63 percent solids liquor. The flash steam from
the concentrator can usually be introduced into the 3rd or 4th
evaporator effect. As shown in Figure 6-2, the additional effect
is a two-body installation.
If the ACE System is installed at an existing mill, an additional
effect must be added to concentrate the liquor to 55 percent. For
a new installation, this can be incorporated into the original
design of the multiple-effect evaporators. The flash steam from
6-6
-------
PAGE NOT
AVAILABLE
DIGITALLY
-------
PAGE NOT
AVAILABLE
DIGITALLY
-------
the additional effect can usually be exhausted into the
existing evaporators.
The exact design for increasing the capacity of the multiple-
effect evaporators at an existing mill to provide 55 percent
or 63 percent solids will depend upon the particular situation,
and will have to be individually investigated.
In general, 63 percent solids is currently used for evaporator
design. Higher solids concentration (above 63 percent) are
expected to require higher evaporator capital costs and opera-
ting costs. However, more experience is required before the
optimum solids concentration can be determined.
The first U.S.A. installation for 63 percent solids has been
reported by Groce and Harris (2_) . This installation is for a
new recovery system at an existing mill.
6.2.2.4 Precipitators
Precipitators have been discussed in detail in Chapter 5. The
cost data for Chapter 5 are based on a precipitator for a con-
ventional recovery system which usually has a particulate load-
ing in the order of 2 grains per ACF. For these new recovery systems
which omit the direct contact evaporator, the particulate loading
will approximately double to 4 grains per ACF. This is attributed
to the fact that the direct contact evaporator usually removes
approximately 50 percent of the entering particulate.
This higher particulate loading of 4 grains per ACF means that the
precipitator for a given size recovery system will have to be
larger and slightly more expensive for the new recovery systems.
Therefore, a precipitator for a new recovery system will have a
capital cost approximately 10 percent higher than a precipitator
for a conventional recovery system in order to achieve the same
particulate loading leaving either precipitator.
As discussed in Section 6.2.1.2, Hot Precipitators have also been
applied to high solids recovery systems in Scandinavia.
6.2.3 SULFITE SOURCES
Sulfite control methods which are considered in this chapter
include packed towers which have been described in detail in
Chapter 5.
6.2.4 NSSC Sources
Due to the lack of emission data (as described in Chapters), con-
trol methods for NSSC sources were not analyzed.
6-9
-------
6.3 APPLICATION, COST, AND EFFECTIVENESS OF NEW DEVELOPMENTS IN
CONTROL TECHNOLOGY
As stated in the introduction of this chapter, the capital costs
and net annual costs for this chapter should be considered order
of magnitude values which will provide an approximate basis of
comparison for the different control methods.
Following the same procedure as Chapter 5, costs for "site
clearance" and "loss of production" are not included in any
calculations in this chapter.
6.3.1 KRAFT SOURCES
The following new developments were evaluated for kraft sources:
Source Description
Recovery System Case 1 - The conversion of an existing re-
covery system with a conventional
direct contact evaporator (precipitator
located on the ground) to the following
new systems:
a. High solids with large economizer.
b. ACE System.
c. High solids with no change to the
existing economizer.
Recovery System Case 2 - Same as Case I except the existing pre-
cipitator is located on the building
roof.
Recovery System Case 3 - The conversion of an existing Venturi
scrubber to the following new system:
a. High solids system with precipitator.
b. Wet precipitator and black liquor
oxidation.
c. Brine system.
6-10
-------
Source
Recovery System
Oxidation System
Smelt Dissolving
Tanks
Digester Relief
and Blow, Mul-
tiple Effect
Combination Boilers
Description
Case 4 - Molecular oxidation system.
Analysis of new developments to control
emissions from weak and concentrated
oxidation systems.
Analysis of new developments to con-
trol emissions from smelt dissolving tanks.
Analysis of an individual thermal oxi-
dation system.
Analysis of mechanical collectors with
shave-off scrubbers.
Brown Stock Washer
Vents
Multiple Effect
Evaporators
Miscellaneous
Analysis of continuous diffusion
washing, enclosed pressure washing,
and incineration of vent gases.
Discussion of caustic scrubbers.
Discussion of steam stripping of
combined condensate.
6.3.1.1 Recovery System, Case 1
Application
This case involves the conversion of an existing recovery boiler
with a 90 percent AOE precipitator, which is located on the ground,
to a new design which eliminates the direct contact of the flue
gas and black liquor. This case also includes the addition of a
new 99.0 percent guaranteed efficiency Venturi scrubber following
the existing precipitator. Three different types of control
methods were investigated as follows:
a.
b.
Conversion of the existing recovery system to an ACE System,
and adding a concentrator to obtain 55 percent solids liquor.
(Figures 6-2, 6-5, and 6-6)
Conversion of the existing recovery system to a high solids
system (62 percent solids) and adding additional economizer
surface. (Figures 6-3 and 6-4)
6-11
-------
PAGE NOT
AVAILABLE
DIGITALLY
-------
I
H1
Ul
co i — n
o •-•»-«
2£P
00 O
03 TO
m o>
70 73 I
m -F»
I»o
o o
o-er-j
•-• m -<
-I 70 -a
>— < -< i— i
o o
a-
< r-
m ^9
oo
O m
•-H o
"i,
m <;
30 m
• s?
S3
73
ro
CO
NOTE:
LOCATION OF VENTURI SCRUBBERS
AS SHOWN IS FOR CLARIFICATION
COSTS WERE FIGURED WITH THE
LOCATION BEING TO THE SIDE OF
PRECIPITATOR.
ADDED EQUIPMENT SHOWN WITH
HEAVY LINES
TEMPORARY CONNECTION FROM
CASCADE TO PRECIPITATOR FOR
OPERATION DURING CONSTRUCTION
OF THE ECONOMIZER.
-------
I
H
*»
m o TI
x ?> >-<
t-i CO C7J
t/5 o •
-)S»
•-" a
z m CT>
o i
m en
-o <
X3 >
m -o — j
o o -<
•-H ?3 T3
-O 3> t-H
>-i — { O
.
PO m CT5
m ^ m
o -H 2
o c: m
-< CO TJ
00
CO ^3 ^3
O C
»-H CD O
r— oo o
m m z
^3 ^3 •<
m
3> ^3
a co
a >-H
1-1 O
NOTE:
LOCATION OF VENTURI
SCRUBBERS AS SHOWN IS
FOR CLARIFICATION COSTS
WERE FIGURED WITH THE
LOCATION BEING TO THE
SIDE OF PRECIPITATOR.
ADDED EQUIPMENT SHOWN
WIT,11 HEAVY LINES.
-------
TO BOILER
AIR CASCADE
EVAPORATOR
SECTION A-A
AIR HEATER
TO
AIR CASCADE
EVAPORATOR
COLD AIR
INLETS
SECTION B-B
AIR HEATER
FIG. 6-6 TYPICAL ARRANGEMENT FOR CONVERSION TO AIR CASCADE
EVAPORATOR AND VENTURI SCRUBBER ADDITION TO AN EXISTING
PRECIPITATOR - RECOVERY BOILER
6-15
-------
c. Conversion of the existing recovery system to a high solids
system (62 percent solids) with no change to the economizer
surface. (Figure 6-1)
This last method sacrifices thermal efficiency for a lower
capital cost and a reduced down time for converting the re-
covery unit. As shown in Figure 6-1, water is added to the
cascade evaporator in the last control method (ij in order to
reduce flue gas temperature to a more reasonable level (below
400°F for tile shell precipitators) before entering the exist-
ing precipitator. This temperature limit does not apply to
steel shell precipitators; however, there are very few steel
units in operation. The flue gas temperature may also have to
be controlled to remain within the design gas flow volumes for
the precipitator and the I.D. fan.
The reduction of the flue gas temperature by the addition of
water to the cascade is considered more reasonable than spray-
ing water directly into the flue gas stream. If water is
sprayed directly into the flue gas stream, some water droplets
may reach the precipitator resulting in significant corrosion
problems.
This reduction in flue gas temperature by the addition of
moisture will increase the visibility of the plume being
emitted from the stack for control method (c). Conversely,
methods fe)and (b) should give a less visible plume. As men-
tioned previously, all three of these control methods include
the addition of a 99.0 percent guaranteed efficiency Venturi
scrubber added downstream of the existing precipitator.
The selection of the most effective control system will depend
upon a number of factors which must be evaluated individually
for each specific application. For example, additional space
will be required for all three control methods. Further, the
electrical requirements will increase, especially for the ACE
system. This may be an important consideration for some mills.
Costs (500 Ton Size)
Capital cost of these three control systems are as follows:
6-16
-------
Control Method Capital Cost
a. Air Contact Evaporator $ 1,500,000
b. 62% B.L.Solids with Additional 1,000,000
Economizer Surface
c. 62% B.L.Solids, without Addi-
tional Economizer Surface 500,000
The net annual costs for these three control methods are as
follows:
Control Method Net Annual Cost
a. Air Contact Evaporator 400,000
b. 62% B.L.Solids with Additional 275,000
Economizer Surface
c. 62% B.L.Solids without Addi- 200,000
tional Economizer Surface
The net annual cost includes credit for chemical recovery,
resulting from the collection of additional salt cake by the
Venturi scrubber. The salt cake collected by the Venturi
scrubber will be bled into the existing salt cake mix tank.
Effectiveness
Particulate. The particulate removal for all three cases is
considered to be approximately equally effective. Actually the
control method utilizing 62 percent solids liquor with no
change to the economizer would be slightly more effective due
to the fact that the cascade evaporator would have been retained
and would collect approximately 50 percent of the inlet dust.
Sulfur Dioxide and Total Sulfur Removal. The emission of SO
and TRS is considered to be equally low for all three con-
trol methods.
Operation. Maintenance and operating costs are not well defined
at this time since none of these control methods has been in
operation for sufficient time to clearly establish these costs.
6-17
-------
However, if we can assume that higher maintenance and operating
costs are usually incurred with more equipment, as evidenced
by significant differences in capital cost, then the mainten-
ance and operating costs will be in the following relationship:
Highest - ACE System
Moderate - High Solids System, large economizer
Lowest - High Solids System, no change to the economizer
The reliability of all three control methods has not been
clearly defined at this time, but is expected to be reason-
ably equal. The high solids system utilizing 62 percent
solids with no change to the economizer does have one disad-
vantage as compared to the other two methods. This is in
regard to plume appearance. Since additional moisture must
be used to reduce the flue gases leaving the economizer from
approximately 600° F to 300-350° F, the moisture content of
the flue gases will be higher than the other two methods.
This will result in a more visible plume leaving the stack
as compared to methods (a) and (b). The plume visibility for
method (c) would be approximately equal to a conventional
recovery with a direct contact evaporator.
Summary
If only these three control methods are compared, control
method (c) has the lowest net annual cost with an effective-
ness equal to the other two control methods, as described
previously.
Comparison with Chapter 5. If this control method is compared
to concentrated black liquor oxidation and the addition of a
Venturi scrubber (from Chapter 5), the following net annual
costs are applicable:
Uet Annual Cost
Concentrated B.L. Oxidation
(99% Oxidation Efficiency)
99.0% Guaranteed Efficiency
Venturi Scrubber
Total
$ 75,000 (from Fig. 5-40)
100,000 (from Fig. 5-21)
$ 175,000
6-18
-------
This net annual cost of $175,000 compares to $200,000 for the
62 percent solids without additional economizer surface. Con-
sidering the fact that oxidation will require less downtime
for conversion, the concentrated oxidation system appears to
be the best choice for an existing installation. The oxidation
system will probably have a somewhat higher malodorous emission
than the 62 percent solids. For example, the relative differ-
ence in H2S concentration ranges in the stack gases might be
approximately 5-20 ppm for oxidation and 1-10 ppm for 62 percent
solids, when these control methods are applied to recovery
systems of moderate age (within the last 10 years).
6.3.1.2 Recovery System, Case 2
Application
Case 2 considers the same three control methods as Case 1 with
the exception that the precipitator is located on the roof, and
has an AOE of 95 percent.
Costs
All the considerations for Case 1 apply equally to Case 2.
The only difference is that the capital cost and net annual
costs will change. Whether the change is an increase or a
decrease is not significant since the order of magnitude
will be relatively small.
Effectiveness
The effectiveness comparison for this case will be the same
as Case 1.
6.3.1.3 Recovery System, Case 3
Application
Case 3 is based on reducing emissions from an existing single
stage Venturi evaporator/scrubber. The existing single stage
Venturi is assumed to have a particulate removal efficiency of
80 percent AOE. The following control methods have been in-
vestigated:
a. Convert to High Solids Systems, omitting contact between
the flue gas and black liquor, and add precipitator.
6-19
-------
b; Add new design wet precipitator (see paragraph 6.2.1.3)
following Venturi and install new oxidation system.
c. Install Na SO Brine System.
These control methods are described in more detail as follows:
a. Convert to High Solids System, Omitting Contact Between
the Flue Gas and Black Liquor and Add Precipitator. For
this control method, the existing Venturi will be removed
and the cyclone separator converted to a cyclone evapo-
rator which utilizes water rather than black liquor. The
cyclone evaporator is retained for several reasons: 1) to
reduce the gas volume to within fan design limits, 2) to
reduce the gas temperature to within fan and precipitator
design limits, and 3) to provide some particulate collec-
tion which will reduce slightly the cost of the new pre-
cipitator.
Therefore, the flue gas temperature will now be in the
order of 350°F rather than 180°F to prevent corrosion
problems in the precipitator. The salt cake collected
in the cyclone evaporator will be concentrated to a 65
percent solids slurry of water and salt cake. The neces-
sary bleed-off to maintain this solids level will flow to
the existing salt cake mixing tank.
A new precipitator will be added at ground level to achieve
minimum particulate emission. The required removal effi-
ciency of the precipitator may vary depending upon local
regulations. However, for the purpose of this control method
the maximum of 99.9 percent guarantee has been included.
This results in an AOE of 99.5 percent.
This control method provides minimum particulate emission
(99.9 percent guarantee precipitator) and minimum TRS
(total reduced sulfur) emission. The minimum TRS emission
is achieved by eliminating the contact of the flue gas with
the black liquor.
b. Add New Design Wet Precipitator Following Venturi and Install
New Oxidation System. One precipitator designed for 200°F
and following a single stage Venturi is currently being in-
stalled in the aluminum industry (see paragraph 6.2.1.3).
Therefore, this type design has been investigated as a poss-
ible control method. The materials of construction are as
follows:
6-20
-------
Shell - Tile
Collecting Plates - Treated Vfood
Internals - 316L Stainless Steel
There is no rapping system since considerable condensation
will take place on the collector plates. The manufacturer
expects that this condensation will flush down the precipi-
tated dust. The ductwork entering and leaving the precipi-
tator is 316 stainless. A new 250 foot brick chimney with
an acid brick lining is included.
The precipitator for this method has a guaranteed efficiency
of 99.5 percent (99.0 percent AOE) for a combined efficiency
of 99.8 percent AOE which includes the 80 percent Venturi
AOE.
A concentrated black liquor oxidation system at 99 plus
percent efficiency is included to minimize TRS enission.
c. Install Na SO Brine System. Another control method which
was considered was converting a Venturi evaporator/scrubber
to a 15 percent Na SO brine Venturi unit as shown in figure
6-7. This method also eliminates the direct contact of the
flue gas and the black liquor to achieve minimum TRS emis-
sion. The solids concentration in the brine solution is
limited to 15 percent to achieve an expected particulate
collection efficiency of 99 percent. The Venturi pressure
drop will be increased to 30 inches to achieve this 99 per-
cent efficiency.
As shown in figure 6-7, the bleed from the brine system is
concentrated to 65 percent solids and then introduced into
the liquor system at the salt cake mix tank.
Costs (500 Ton Size)
Based on the above descriptions, order of magnitude capital
costs and net annual costs are as follows:
Control Method Capital Cost Net Annual Cost
a. High Solids $ 1,800,000 $ 320,000
b. Wet Precipitator $ 1,000,000 $ 150,000
c. Na SO Brine System $ 1,500,000 $ 330,000
6-21
-------
PAGE NOT
AVAILABLE
DIGITALLY
-------
Effectiveness
Particulate. Particulate removal for the above methods is estimated
as follows:
a. High Solids Adjusted for 50% Collection
in Cyclone Evaporator (Without Cyclone
Evaporator 99.5% AOE) - 99.8% AOE
b. Wet Precipitator - 99.8% AOE
c. Na2SO4 Brine System - Approx. 98.0%
AOE
Sulfur Dioxide. The difference in SO emission for these three
methods is unknown at this time. Based on operation at rated capac-
ity in similar furnaces, the SO emission will probably be approxi-
mately the same for all three methods.
Total Reduced Sulfur. All three methods should be approximately
the same, except for the TRS emissions from the oxidation system
of method (b). No data are available at this time to better
quantify the TRS emission.
Operation. All of these systems are relatively unproven at the
present time. Therefore, any discussion of operating problems
would be hypothetical.
Summary
Based upon the above three control methods, the wet precipitator
control method (b) indicates the best effectiveness with minimum
cost.
Comparison with Chapter 5. A comparison of the wet precipitator (b)
with the two stage Venturi scrubber from Chapter 5 is now in order.
From paragraph 5.3.6.1.4 of Chapter 5 the two stage Venturi was an
effective control method and had the lowest net annual cost. In
order to achieve a TRS emission comparable to the wet precipitator
control method (b), a concentrated black liquor oxidation system
must be installed.
A cost comparison of these two control methods is as follows:
(500 ton size)
6-23
-------
Control Method
Capital Cost
$ 1,000,000
$ 470,000
Net Annual Cost
Wet Precipitator $ 1,000,000 $ 150,000
*Second Stage Venturi $ 470,000 $ 90,000
These two control methods are considered equally effective for SO,
and TRS removal. The particulate removal effectiveness is as
follows:
Wet Precipitator
Second Stage Venturi
99.8% AOE
98.0% AOE
Considering the difference in the above costs, the second stage
Venturi appears to have considerable merit if the 98.0 percent AOE
particulate collection will meet the local air pollution regulations.
Therefore, for recovery systems which include Venturi evaporator/
scrubbers, the addition of a second Venturi scrubber in combination
with a concentrated oxidation system appears to be the most effec-
tive control method for this case. As mentioned previously, the
effectiveness of the concentrated oxidation system applied to a
Venturi evaporator/scrubber needs to be further evaluated.
*Broken down as follows:
Second Stage Venturi
Scrubber
(from Chapter 5)
Oxidation
Capital Cost Net Annual Cost
$ 270,000
$ 200,000
$ 470,000
$ 15,000 (from Fig. 5-34)
$ 75,000 (from Fig. 5-21)
$ 90,000
6-24
-------
6.3.1.4 Recovery System, Case 4
Application
This control method involves the use of molecular oxygen to
oxidize weak black liquor. The method is classified as a
new development since there is only one system which is
currently under construction. This system has been described
in detail by Galeano and Amsden (5) . Oxygen can be introduced
into a pipeline across a pump.
Molecular oxidation can apparently be used with either weak
black liquor or concentrated black liquor. The usual problem
of foaming when oxidizing 100 percent pine weak black liquor
should not exist because all of the oxygen is absorbed.
The oxidation products were reported to be stable up to 12
hours. Therefore, it appears that the reversion of sodium
thiosulfate to sodium sulfide, as reported by Martin (6) may
not occur when using molecular oxidation. If this is, indeed,
proven true in the future, this control method would probably
find more application in the oxidation of weak black liquor,
thereby, reducing the odorous emissions from the multiple-effect
evaporators.
Costs
The capital cost for this method is rather low, and was re-
ported by Galeano to be approximately $75 per ADT. Therefore,
the capital cost for a 500-ton mill would be approximately
$37,000.
The net annual cost would vary widely depending upon the cost of
oxygen. Galeano used an oxygen cost of $8.50 per ton, which
resulted in an operating cost of $0.07 per ADT of pulp. However,
this is rather low, and is based on being located close to a pipe-
line supply of oxygen. From $8.50 per ton, oxygen cost will range
upward to $100 per ton. Therefore, it is impossible to calculate
any meaningful net annual cost due to the wide range of oxygen
costs.
CAPITAL COST - $37,000 NET ANNUAL COST - Wide Fluctuation
Effectiveness
Particulate. This control method has no influence on particulate
removal.
6-25
-------
Sulfur Dioxide and Total Reduced Sulfur Removal. The reduction
of SO and TRS would be similar to the oxidation systems
described in Chapter 5.
Operation. The operation of this control method should be
simpler and require less maintenance and operator attention
than the oxidation systems in Chapter 5.
Summary
In summary, molecular oxidation should be equally as effective
as the oxidation systems described in Chapter 5. Where the
cost of oxygen would result in a net annual cost equal to the
net annual costs of the systems in Chapter 5, molecular oxygen
would possibly be preferred due to ease of operation and mainte-
nance. Of course, this should be analyzed based on any addi-
tional performance data which may be available in the future.
6.3.1.5 Oxidation System
Application
The application of weak black liquor oxidation and concentrated
black liquor oxidation systems to recovery units has been
described in more detail previously in Chapter 5. This section
discusses the application of control methods to the off-gas vent
from the oxidation systems.
Data which have been published in relation to an analysis of
oxidation system off-gas pollutants is very limited. Therefore,
it is impossible to define the range of pollutants from oxidation
systems. However, a general appraisal of some facts can be
stated.
First, the off-gas flow from a weak black liquor oxidation
system will usually be higher than a concentrated system.
Based on a sodium sulfide loading of 10 grams/liter, a weak
liquor system might require approximately 30,000 SCF/ADT of
air; whereas, a concentrated system might be approximately
15,000 SCF/ADT of air. Therefore, the weak liquor system
might have an off-gas flow as high as approximately double
that of the concentrated system. This, in turn, means
the control equipment must also be larger.
6-26
-------
While data analyzing the off-gas pollutants are limited,
Amberg and Walther (8) pointed out that a concentrated
system would emit less total reduced sulfur compounds than
a weak liquor system. This is attributed to the fact that
the quantity of total reduced sulfur compounds in the
liquor is less for concentrated liquor since a significant
quantity of these compounds is vented from the multiple
effect evaporators.
Costs
Costs have not been estimated because of the unknown con-
centrations and composition of the off-gas pollutants.
Control methods which might be considered are as follows:
a. Introduction into recovery F.D, fan for incinera-
tion in recovery furnace,
b. Incineration in lime kiln,
c. Individual thermal incinerator,
d. Packed tower scrubber which uses caustic and
chlorine,
e. Dilution in stack with other boilers.
Methods (a) and (b) are expected to find less acceptance
because of: Increased corrosion problems in F.D. fan duct-
work and F. D. fan itself; and the oxidation off-gas volume
is usually too high to be handled by an existing kiln.
The remaining three methods should be investigated further.
Effectiveness
Since the concentration and quantity of pollutants are
unknown, the effectiveness cannot be precisely defined.
Additional research and development is needed to determine
the most effective control method for the off-gas from
oxidation systems.
6-27
-------
6.3.1.6 Smelt Dissolving Tanks
Two control methods which have baen discussed as possible
applications to the dissolving tank are evaluated in general
terms. In view of the questionable nature of these control
methods, an analysis of cost and effectiveness has not been
prepared. The two control methods are as follows:
Exhaust Vent Into Recovery Flue Ahead of ID Fan. For some
new recovery installations, an arrangement has been discussed
where the dissolving tank vent gases are exhausted into the
recovery boiler flue gas stream ahead of the induced draft
fan. The chemicals in the vent ge.ses are then collected in
the recovery precipitator. Since the dissolving tank vent
stack diameter is approximately 3 feet to 7 feet, the routing
of this stack (or duct) is difficult. For an existing in-
stallation, the rerouting of this stack is not possible with-
out considerable expense. Due to the high capital cost, this
control method is not practical for an existing installation.
For a new installation where the routing of this stack may
be included in the original plans, this may be a practical
control method. The most convenient method of routing the
stack would be from the dissolving tank (located at the gas
outlet side of the boiler) into the ductwork upstream of the
cascade evaporator. However, there is some concern about
the chemicals in the dissolving tank vent causing lignin pre-
cipitation in the cascade evaporator. Therefore, injection
of the dissolving tank vent gases downstream of the cascade
may be the most appropriate location. This location may also
cause problems since the stack must be routed around the cas-
cade to reach the downstream side. Injecting the gases into
the breeching ahead of the ID fan is a potential source of
damage to the fan and breeching during dissolving tank puffs
and explosions.
In consideration of the relatively small capital cost
difference between this control method and the installation
of a scrubber or packed tower, the use of this control
method is somewhat questionable. Also, the aspects of
damage from puffs and the possibility of lignin precipitation,
further detract from this control method. One possible benefit
of this control method is better dispersion of the vent gases.
This is attributed to the fact that the recovery stack is
usually higher than a separate smelt dissolving tank vent stack.
6-28
-------
One mill is currently installing this system. Operating
experience from this mill should help to further evaluate
this method.
At the present time, this control method would have to be
classified as questionable.
Incinerate Vent in Recovery Furnace. The dissolving tank
vent flow will vary depending upon factors such as: shatter
jet steam pressure, stack dimensions, and amount of air
leakage. In general, this flow will approximate 10 percent
of the recovery boiler flue gas flow.
Incineration of the dissolving tank vent gases in the
recovery furnace can be considered as a method for odor
incineration plus recovery of the chemicals in the vent.
The chemicals would then be recovered by the boiler precipi-
tator or Venturi scrubber. Injection of the gases into the
furnace would replace other control methods such as a
scrubber or demister. The vent stack must still extend
above the boiler roof to relieve dissolving tank puffs.
The injection of dissolving tank vent gases into the furnace
is considered to have more disadvantages than advantages.
Some of these disadvantages are:
1. The recovery furnace would be more subject to damage
from dissolving tank puffs.
2. The boiler induced draft fan must handle the additional
10 percent vent flow, thus possibly reducing the combustion
air available to the furnace.
3. The ductwork and fan for injection of the vent flow must
be fabricated of stainless steel. The vent stack itself
would be carbon steel construction.
4. Corrosion problems could be expected where the duct enters
the furnace since this would be a horizontal duct and some
of the chemicals would settle out. This means that
corrosion of the boiler pressure parts may occur.
About the only advantage to this method, as compared to a
scrubber (see Chapter 5), is the possibility of reduced capital
cost.
After weighing the above factors and in consideration of other
control methods, this control method was considered to be un-
desirable.
6-29
-------
6.3.1.7 Digester Relief and Blow, Multiple Effect Evaporators
Application
Supplementing Chapter 5, an additional control method which might
be considered is a separate thermal oxidation unit for these vents.
While thermal oxidation units are used successfully in many appli-
cations , caution should be used for this source since these gases
are explosive.
Therefore, this control method includes a fan for introducing air
into the vent gases prior to their entrance into the thermal
oxidation unit. The dilution air allows the vent gases to be
handled without explosions. For batch digesters, this fan may be
sized to handle the maximum flow from the blow tank without a gas
accumulator. This type system has been used in Sweden (12).
Costs
The thermal oxidation unit is designed to heat the gases to 1300-
1500°F for approximately 0.5 seconds. The costs as shown in
Figures 6-8 and 6-9, include the necessary piping, wiring, etc.,
for a complete installation. The costs for the batch digester
include a gas accumulator. The Swedish system which omits the
gas accumulator would be somewhat less expensive.
Effectiveness
Particulate. There are no particulate emissions from this source.
Reduced Sulfur. Thermal oxidation has an effectiveness approach-
ing 100 percent.
Sulfur Oxides Removal. The reduced sulfur compounds are oxidized
to sulfur oxides; therefore, the sulfur oxides will be emitted.
Operation. Due to the physical location of these vents, the
thermal oxidation units will be installed in areas which are
relatively remote from operator attention. This should be con-
sidered when evaluating this control method.
6-30
-------
§
o
^40
X
•V*
fc 30
o
o
g 20
CL
O
< 10
o
o
o
•i
^1
X
o
o
y
8
7
6
5
6 9
ADT/DAY X 100
12
FIG.6-8 CONTROL METHOD COSTS FOR
DIRECT FLAME INCINERATION OF EVAPORATOR
RELIEF GASES OR CONTINUOUS DIGESTER
RELIEF GASES
6-31
-------
80
o
o
o
£70
C_3
60
£50
O.
40
30
l.*J
o"
o
o
" ?0
j— t L-
-------
Summary
The net annual cost of this control method is very similar
to incineration in the lime kiln which is discussed in
Chapter 5. The sulfur oxides emission from this source
are higher than the control method of incineration in the
lime kiln.
Therefore, the control method of incineration in the lime
kiln will usually be preferred to this control method of
individual thermal oxidizers. Individual thermal oxidizers
will undoubtedly find acceptance in mills where the lime
kilns are remote from the digesters and multiple effect
evaporators.
6.3.1.8 Combination Boilers
Application
This newly developed control method combines a multi-tube
cyclone collector with a shave-off scrubber. About 20 percent
of the total gas flow is exhausted from the hopper of the
new multi-tube collector into the shave-off scrubber. This
system is used to replace existing low efficiency multi-tube
cyclone collector to an overall collection efficiency of 96
percent.
For purposes of this report the following items were assumed
for a system evaluation: boiler energy input of 40 percent
bark and 60 percent oil, collectors and ducts of mild steel,
field erection from shop assembled knock-down components,
and reuse of existing sand classifiers. The material collected
in the shave-off scrubber were assumed to be discharged to the
existing ash handling system. The dust collector has a
pressure drop of 2.5 inches WG and the scrubber requires an
additional 6.0 inches WG for 20 percent of the total gas flow.
Costs
Cost curves have been based on the system gas handling capacity,
since there is no direct relation between gas flow and mill tonnage.
These curves are presented in Figure 6-10.
6-33
-------
— . JOU
o
o
o
-~ 300
X
~ 250
to
o
o
_j 200
X
X
X
/
10 20 30 40
GAS VOLUME (CFM X 10,000)
FIG.6 - 10 CONTROL METHOD COSTS FOR REPLACING
MULTI-TUBE COLLECTOR WITH NEW MULTI-TUBE COLLECTOR
AND SHAVE-OFF SCRUBBER - COMBINATION BOILER
6-34
-------
Effectiveness
Particulate. This method has an expected guaranteed efficiency
of 96 percent, which is equal to two multi-tube collectors in
series as described in Chapter 5.
Sulfur Dioxide and TRS Removal. This control method does not
remove and SO or reduced sulfur compounds.
Operation. The operation of the multi-tube collector of this
control method will be the same as any other multi-tube collec-
tor. However, the operation of the shave-off scrubber is not
clearly defined. One consideration that may be of concern is
the handling and disposal of the bark char from the scrubber.
A strainer has been included for this control method; however,
the handling and disposal of the scrubber effluent must be in-
vestigated for each individual application. This control
method may create a water pollution problem.
Summary
For new installations, two multi-tube collectors in series
(Chapter 5) will have a net annual cost approximately equal to
this method. Since the two collectors in series are a more
tried and proven design, the two collector methods will prob-
ably find more acceptance than the collector and scrubber
method for new installations.
It is possible that the application of the shave-off scrubber
may find more acceptance on existing installations where space
is limited. Of course, this would have to be investigated for
each application.
6.3.1.9 Brown Stock Washer Vents
The following three systems have been investigated to handle
the high volume, low concentration gases from this source:
a. Continuous Diffusion Washing
This is a patented system in which pulp flows axially upward
through the unit, while wash water diffuses radially. The
wash water, distributed by nozzles, diffuses through the
stock and is extracted through a strainer. The clean washed
stock is discharged with the help of a scraper at the top of
the unit. Because the unit is enclosed, no malodors should
escape to the atmosphere.
6-35
-------
Order of Magnitude Capital Cost:
500 ADT/Day - $ 450,000
The above price is for a single stage of diffusion washing
which follows four hours of diffusion washing in a continuous
digester. Prices include equipment and its erection, building,
equipment foundations, instrumentation, piping and valves,
filtrate tank and pump, motors, motor controls, and wiring.
Indirect costs of 30 percent are included, consisting of
contingency, engineering, general construction overhead,
spare parts, and sales tax.
b. Enclosed Pressure Washing
The design of the pressure washer differs from the conventional
vacuum filter in that the pulp mat is formed and the washing
liquor is displaced from the mat by a difference of positive
pressure instead of vacuum. The washer is completely enclosed
with an airtight hood.. An external blower provides pressure
under the hood, and the air pressure forms the mat on the drum
wire and passes into the drum interior. The air and vapors then
discharge from the drum interior at the take-off, lifting the
pulp mat off the drum. The vapors then pass to the suction of
the blower to be recirculated back into the drum.
Order of Magnitude Capital Cost:
Batch Digesters. For 500 tons per day, following batch
digesters and assuming three (3) washers in series, the
installed cost will be approximately $1,700,000. The cost
includes equipment and its erection, building, equipment
foundations, instrumentation, piping and valves, filtrate
tanks and pumps, motors, motor controls and wiring. Indirect
costs of 30 percent are included,
Continuous Digesters. For 500 tons per day, following a
continuous digester having four hours of diffusion washing
and assuming one (1) washer, the installed cost will be
approximately $650,000. The cost will include those items
listed for washing after batch digesters and contain indirect
costs of 30 percent.
6-36
-------
c. Incineration of Vent Gases
A control method which has been suggested is the incineration
of the brown stock washer vent in the recovery furnace (9).
Flow from the brown stock washer vent would be introduced"
into the suction of the recovery forced draft fan. Since
the forced draft air requirement for a recovery system is
approximately 200,000 SDCF/ADT of pulp, the brown stock
washer vent flow of approximately 90,000-150,000 ACF at
120° F/ADT of pulp can be accommodated by the recovery system.
The brown stock washer vent gases are usually very corrosive
and will require special consideration in the selection of
forced draft fan and duct materials.
For discussion, if a brown stock washer vent flow of 100,000
ACF/ADT is assumed then the following duct sizes result
based on 2000 FPM velocity:
Mill Size ADT/Day
300 500 1000
Approx. Flow - ACFM 21,000 35,000 70,000
Approx. Duct Diam. 3.7 4.7 6.7
In considering the application of this control method to
an existing installation, the problem of material changes
to prevent corrosion and the problem of physically instal-
ling a large duct in an existing facility, require that
each installation would have to be evaluated individually.
This control method could be more easily applied to a new
mill or a major expansion at an existing mill.
The method is a new development which has never been
installed.
Summary
At present none of the three developments has been applied
within the United States. There is one installation of
diffusion washing of brown stock in Canada and all three
methods have had limited application in Europe.
6-37
-------
Of the three methods the practice of incineration in
the recovery furnace will probably cause the most
concern among operators because of possible resultant
upsets in furnace operation and corrosion problems.
There are several installations of pressure washers in
Europe and one in Japan. In theory, because the vapors
are being recirculated within the washer, this system
can reduce emissions to the atmosphere. The gas flow,
if any, would be very small and, therefore, can be
incinerated in the lime kiln.
Brown stock diffusion washing systems have been installed
in Europe and Canada. Because the pulp is washed in
complete absence of air there should be no emissions to
the atmosphere.
Either pressure washing or diffusion washing will probably
increase the sodium sulfide content of the liquor as
compared to conventional vacuum washers. This is expected
because conventional washers expose the liquor to contact
with the air resulting in some oxidation of the sodium
sulfide. The effect of this increased sodium sulfide
concentration on the capacity of an existing oxidation system
should be considered when selecting these new design washers.
Diffusion washing and pressure washing appear to have
significant future potential for the reduction of emissions
to the atmosphere and should be investigated further.
6.3.1.10 Multiple Effect Evaporators
Caustic Scrubbers. Caustic scrubber systems are favored on
the European continent, particularly in Sweden where this
system was invented.
The system consists of first a scrubber, then a condenser
(heat exchanger), and finally a cyclone separator. Non-
condensible gases leaving the multiple effect evaporators
pass to the scrubber. The scrubbing medium may be white or
weak wash liquor or a weak caustic soda solution. Ninety-nine
percent of the hydrogen sulfide and some methyl mercaptan
is absorbed in the caustic solution. From the scrubber the gases
6-38
-------
and liquor are discharged to a condenser, where heat is
recovered. The gases and liquor then continue to a
cyclone separator, the gases being exhausted through
the evaporator vacuum system and the liquor returned
to the causticizing system. Because the dimethyl
sulfide and the dimethyl disulfide are not absorbed
in the caustic solution it follows that these two gases
will escape to the atmosphere.
Order of magnitude costs for an 800 ton per day mill will
be $75,000-00, including erection and indirect costs. All
material in contact with the liquid is stainless steel
Type 304.
6.3.1.11 Miscellaneous
Steam Stripping of Combined Condensate. Liquids obtained
by condensing the vapors from blow heat recovery and digester
relief systems and the evaporation of the spent kraft liquors
will contain steam-volatile organic compounds. These will
include the sulfur-bearing compounds, later released from the
mill effluent disposal system. Steam stripping the effluent
condensates in a stripping column and then collecting the
volatile compounds for disposal will eliminate or reduce the
discharge to the atmosphere of these kraft odors. The stripped
condensate could be purified to a degree that it could be
returned for reuse in operations.
A typical steam stripping system will include a feed tank for
storing the condensate, a heat exchanger for pre-heating the
feed condensate, a fractionating or stripping column in which
the feed condensate is passed counter-current to steam flow,
a stripped condensate storage tank, a collection system for
recovering and storing the stripped volatiles, and pumps,
valves and instrumentation.
The sulfur bearing compounds that are stripped from the conden-
sate consist of hydrogen sulfide, methyl mercaptan, dimethyl
sulfide and dimethyl disulfide. In addition, some methanol
and turpentine are volatized. Because the boiling points of
dimethyl sulfide and dimethyl disulfide are approximately
100 to 120°C, the stripped condensate from the column may
contain small amounts of these two compounds.
6-39
-------
6.3.2 SULFITE SOURCES
6.3.2.1 Ammonium Liquor Incineration
Application
Incineration of ammonium liquor is practiced in relatively
few mills; however, it is anticipated that more mills will
utilize incineration in the future. This will occur as
a result of efforts to decrease stream pollution which may
result from the discharge of waste ammonium liquor.
Currently, there are two known acid sulfite mills (T) which
incinerate liquor and both of these incinerate in combination
boilers burning pulverized coal. There are additional
ammonium NSSC mills and possibly some ammonium bisulfite
mills, which incinerate ammonium liquor but this control
method is confined to an acid sulfite mill. Two cases were
considered as follows:
a. Incineration in Combination Boiler
Since the basis for the consideration of this control
method is incineration of the ammonium liquor in a
combination boiler, it appears reasonable to consider
the recovery of SO from this unit by scrubbing the flue
gases with an ammonium hydroxide solution. However, there
are several disadvantages which preclude the application
of this method. The major disadvantages are:
Ash. Ash from the burning coal and bark will require high
efficiency dust collection equipment to prevent the
injection of ash into the cooking acid.
Higher Gas Volume. Recovery of SO would require the
absorption system to be designed for the flue gas volume
leaving the combination boiler. While this volume will
vary for each particular installation, it will be substan-
tially higher than the gas volume produced from the incinera-
tion of the ammonium liquor only. An order of magnitude
would be approximately twice as much gas volume for a
combination boiler as compared to ammonium liquor incineration
only. Therefore, the absorption system would have to
handle this higher gas volume and to remove the SO in
a more dilute concentration.
6-40
-------
High Excess Air. The combustion of fuels such as coal
and bark usually requires a large amount of excess air
(20 percent to 40 percent). As excess air increases,
the formation of SO in relation to SO also increases.
This in turn will produce higher quantities of sulfates
in the scrubbing liquid, and increase the amount of
ammonia make-up required for the system. The formation
of sulfates with the consequent increase in ammonia
make-up could be minimized by using a gas cooling tower
and discharging the effluent to waste, however this would
increase the waste treatment cost. Further, this
cooling water would absorb SO , thus reducing the SO
available for recovery.
For the above reasons, the recovery of SO from a combina-
tion boiler by scrubbing with an ammonium solution is not
considered to be a practical method.
b. Incineration in Separate Boiler
To avoid the technical difficulties of SO recovery from
combination boiler flue gases, a separate ammonium liquor
boiler and flue gas SO recovery system were considered.
Estimates were prepared for a 200 ADT per day ammonium
(acid sulfite) mill. A full scale system utilizing ammonium
liquor incineration and recovery of SO is not in operation
anywhere in the world so far as the contractor knows.
Therefore, this control method must be considered unproyen
at the present time in so far as the exact SO recovery
efficiency is concerned.
Since it was assumed the liquor was previously fired in the
combination boiler, no credit or debit has been allowed for
the recovery boiler steam.
In designing the sulfur dioxide recovery system it is
necessary to know the desired SO concentration which may
be emitted from the stack and the SO concentration at the
entrance to the recovery system. Both of these concen-
trations will vary from mill to mill, depending upon the
quantity of sulfur in the ammonium liquor, stack height,
stack exit velocity, meteorological conditions, topography,
and other physical conditions peculiar to the individual
mill. However, it appears desirable to design the S02
recovery system for an efficiency in the order of 95 percent.
Based upon pilot plant work by Palmrose and Hull (11), an
6-41
-------
SO_ recovery efficiency of approximately 95 percent when
incinerating ammonium liquor appears to be possible.
The AOE which might be expected is unknown.and must await
the development of actual installations and test data from
these installations.
The sulfur dioxide recovery system for this control method
consists of four components:, a complete boiler instal-
lation, a gas cooling tower, an SO absorption tower, and
a water cooling tower. The absorption of SO by the cooling
water and the gas cooling tower will result in a low pH
liquid. Since this low pH liquid may damage a conventional
water cooling tower, a stainless steel heat exchanger has
been included as shown in Figure 6-11.
The required amount of makeup water is introduced into the
SO absorption tower to minimize the los,s-.of NH from the
system and control the concentration.of .the combined SO
in the acid. , . , • :•,-
This control method does not consider special applications
such as mills receiving salt waterborne logs. For these
specialized cases, individual control methods and cost
calculations must be prepared for each case.
(200 Ton Mill)
Costs
Capital Costs $ 2,500,000
Net Annual Cost $ 270,000
6.3.2.2 Improved SO Recovery in Magnesium Sulfite Process
Application
A potential modification of the magnesium sulfite process
is a reduction of SO emission by upgrading the efficiency
of the SO recovery system. The basic process considered
for this evaluation includes an MgO recovery boiler and
an existing 95 percent efficiency SO recovery system.
As reported by Kleinegger (10) , an SO removal efficiency
6-42
-------
AIR
HEATER
CD
OJ
AMMONIUM
LIQUOR
NHj
MAKE-UP
TO FORTIFICATION
TOWER AND PULP MILL
MAKE-UP
H.J.S.
NEW DEVELOPMENT
AMMONIUM LIQUOR INCINERATION
AND RECOVERY OF S02
EXHIBIT NO.
FIG. 6-II
SYSTEMS ANALYSIS STUDY OF
EMISSIONS CONTROL IN THE WOOD PULPING INDUSTRY
CONTRACT NO. CPA 22-69-18
FOR
DEPARTMENT OF HEALTH. EDUCATION AND WELFARE
CONSUMER PROTECTION AND ENVIRONMENTAL HEALTH SERVICE
NATIONAL AIR POLLUTION CONTROL ADMINISTRATION
ENVIRONMENTAL ENGINEERING, INC.
GAINESVILLE. FLORIDA
J. £. SIRRINE COMPANY, ENGINEERS
GREENVILLE, S. C.
-------
of approximately 95-97 percent may be expected from
an existing installation. (Recent work indicates that
these SO_ levels may be in question.) This efficiency
may be increased by up-grading the existing SO recovery
system by the addition of a packed absorption tower
which may increase the overall SO removal efficiency above
95 percent. The SO concentration entering the recovery
system and the SO_ concentration leaving the stack will
vary for the same reasons previously stated under
paragraph 6.2.2.1 b.
The cost for this control method was developed for a 200
ADT per day mill. It was assumed that adequate space was
available for the installation of this additional packed
tower.
Costs (200 Ton Mill)
Capital costs include a packed tower and a booster I.D
fan with the necessary pumps, piping, ductwork and electrical
wiring.
Capital Cost $ 200,000
Net Annual Cost 50,000
6.4 NSSC SOURCES
Due to the lack of emission data (as described in Chapter
5), control methods for NSSC Sources were not analyzed.
6-44
-------
REFERENCES
1. Hochmuth, F. A., "An Odor Control System for Chemical
Recovery Units," Pulp and_Paper Magazine of Canada, 70,
57-66, (1969).
2. Groce, A. B., Jr. and Harris, A. D., Jr., "Evaluation of
Kraft Liquor Concentrator Operation," presented at TAPPI
Alkaline Pulping Conference, October 14-17, 1969.
3. Arhippainen, Bengt and Westerberg, E, N. , "Kraft Odor
Controls - Its Effect on Mill Operating Parameters and
Costs," Pulp and Paper Magazine of Canada, 69, April
19, 1968.
4. Roberson, J. E., "How Does Recovery Odor Control Affect
a Kraft Mill Energy Balance," Pulp and Paper, November,
1969.
5. Galeano, S. F., and Amsden, C. D., "Weak Black Liquor
Oxidation With Molecular Oxygen," presented at the 62nd
APCP Meeting in New York, New York, June 22, 1969.
6. Martin, F., "Secondary Oxidation Overcomes Odor from
Kraft Recovery," Pulp and Paper, 43, June 1969.
7. Clement, J. C. and Sage, W. C., "Ammonium Liquor Burning
and SO,, Recovery," presented at 23rd Engineering Conference
of TAPPI, November 6, 1968.
8. Amberg, H. R. and Walther, J. E., "The Status of Odor
Control in the Kraft Pulp Industry," presented at
National AICHE Meeting, Portland, Oregon, August 24-27,
1969.
9. Suda, Stanley, "Kraft Recovery Odor Reduction System,"
presented at 23rd TAPPI Alkaline Pulping Conference,
Jacksonville, Florida, October 14-17, 1969.
10. Kleinegger, J. C., "Relative Absorptive Efficiency of^
Packed Towers in Magnesium Acid Bisulfite Production,"
TAPPI, 52_, 1291-, July 1969.
11. Palmrose, G. V. and Hull, J. H., "Pilot Plant Recovery
of Heat and Sulfur from Spent Ammonia Base Sulfite
Pulping Liquor," TAPPI, _35_, May 1952.
12. Tucker, W. G., Personal Communication, February 9, 1970.
6-45
-------
CHAPTER 7
CRITICAL REVIEW OF CONTROL TECHNOLOGY
•CABLE OF CONTENTS
Page No.
Summary 7— 1
Introduction 7- 2
Kraft Process 7- 3
Precipitators 7- 6
Venturi Scrubbers 7- 8
Cyclonic Scrubbers 7-10
Recovery Systems Without Direct Contact
Evaporators 7-11
Black Liquor Oxidation 7-12
Orifice Scrubbers 7-14
Mechanical Collectors 7-14
Incineration in Lime Kiln 7-15
Enclosed Pressure Washing 7-17
Continuous Diffusion Washing 7-17
Sulfite Process 7-18
Packed Tower Scrubbers 7-19
Ammonium Liquor Incineration and SO Recovery 7-20
Improved SO Recovery in Magnesium Sulfite
Recovery 7-20
NSSC Process 7-21
7-i
-------
CHAPTER 7
CRITICAL REVIEW OF CONTROL TECHNOLOGY
SUMMARY
In Chapters 5 and 6, control methods presently in
use and new developments as applied to various sources were
analyzed. The relative merits and specific limitations of the
most effective and economical control methods are summarized
in this chapter. An example also is analyzed of the ability of
selected kraft process configurations to meet selected emission
limitations.
Within each process, control devices are evaluated
as to applicable emission sources, efficiency, flexibility,
economics, reliability, and adaptability.
7-1
-------
7.1 INTRODUCTION
Chapters 5 and 6 analyzed various control methods applied
to different sources. The relative merits and specific
limitations of the most effective and economical control
methods are summarized in this chapter in general terms.
Control methods are frequently selected for new plants
or new additions to existing plants while the new plant
or addition is in the design stage. Therefore, parameters
such as gas flow rates, temperature, composition, density,
and viscosity, as well as the particulate concentration,
size distribution, composition, shape, and density are all
designer estimates. Once the facility is in operation,
the real values for these variables become evident and
may differ somewhat from design values.
Guaranteed efficiencies are often quoted by equipment
manufacturers for particulate removal. These guarantees
are based upon assumed values for the above parameters and
must be adjusted based upon test conditions which may differ
from those anticipated. The procedures for converting from
guarantee to test conditions are often complicated and
usually involve both the physical laws and empirical formulas.
As indicated in Section 5.3.3 most equipment manufacturers
base their guarantee of collector efficiency on procedures
specified by the Industrial Gas Cleaning Institute (IGCI).
(See also Section 9.2.2.) The method is based on a determi- .
nation of the average particulate concentration at the inlet
and outlet of the collector. Isokinetic sampling at specified
locations in the ducts for specified times is required at
steady operating conditions. The particle collector is speci-
fied as a filter of any material and form which has a collec-
tion efficiency in excess of 99.0 percent for particulates of
the approximate size distribution to be encountered during the
test. Collectors other than filters are acceptable if it can
be demonstrated that they have an efficiency exceeding 99.0
percent at test conditions.
7-2
-------
The design of control methods for the removal of gaseous
pollutants is less precise than the control methods for
particulate removal. In actual practice, it is usually
the exception rather than the rule that guarantees are
made for gaseous pollutant removal.
Since changes to control methods to meet guarantees are
usually costly and interrupt plant operation, control
methods must be judiciously specified for both guaranteed
performance and flexibility in design and operation to
satisfy possible future process changes. When pollutant
removal guarantees cannot be obtained from equipment
suppliers or designers, the plant often finds that deci-
sions must be made to insure conservative design.
In order to meet more stringent air pollution regulations
in the future, it is possible that pilot plant test data
may have to be collected and extrapolated for the design
of full scale control methods. Situations of the type
cited in the preceding paragraphs require the cooperation
and appreciation of all segments of industry, government
and the public.
7.2 KRAFT PROCESS
The kraft pulping process has received the most attention
in recent years regarding emissions to the atmosphere.
Since the kraft pulping process produces approximately 70
percent of the total chemical pulp in the U. S. A., it is
expected that the major portion of expenditures for air
quality control will be required by the existing kraft
mills.
Prom Kraft Flow Diagram No. 10 in Chapter 3, it is apparent
that the emissions from a new modern kraft mill can be nominal
with the possible exception of the brown stock washers. Even
these emissions from the brown stock washers can be expected
to be drastically reduced as additional technology and experi-
ence are developed for diffusion washing and pressure washing
as described in Chapter 6.
An attempt was made to compare emissions from older kraft
mills (15 + years), older mills which have been upgraded to
place emphasis on emission control, new mills (designed within
the past 5 years), and mills designed in 1970 with the latest
tested control technology. The emissions from kraft mills
7-3
-------
vary widely depending on many factors including product being
produced, type of wood being pulped, production rate, process
variables, operating variables, and combination of process
equipment used.
Two older mills, operated essentially as they might have been
constructed more than 15 years ago/ are typified by Kraft Flow
Diagrams No. 1 and No. 5. Emission control equipment in the
chemical recovery system was limited to an electrostatic pre-
cipitator on the recovery boiler having an 88 percent AOE and
a scrubber on the lime kiln with an 80 percent AOE. A mill of
similar vintage but upgraded to reduce emissions is illustrated
by Kraft Flow Diagram No. 3. Emission control systems include
a 92 percent AOE precipitator on the recovery boiler, a 99
percent AOE scrubber on the lime kiln, a secondary scrubber
following the precipitator having an AOE of 80 percent, weak
black liquor oxidation, and incineration of noncondensibles.
Mills of less than five years in age are typified by Kraft
Flow Diagram No. 4. Emission control systems include a 97
percent AOE precipitator on the recovery boiler, a 99 percent
AOE scrubber on the lime kiln, concentrated black liquor oxi-
dation, mesh pads in the dissolving tank vent, and a scrubber
on the off gases from the multiple effect evaporators. A
mill which might be designed in 1970 incorporating the latest
technological developments in emission control is illustrated
by Kraft Flow Diagram No. 10. Emission control systems include
elimination of the direct contact between flue gases and black
liquor (thus making black liquor oxidation unnecessary), a
precipitator on the recovery boiler having an AOE of 99 percent,
a 99 percent AOE scrubber on the lime kiln, incineration of
noncondensibles, and a 95 percent AOE scrubber on the dissolving
tank vent. It must be recognized that these examples were
selected because descriptions were readily available to the
reader. The conditions are not the same in all cases and it
must not be inferred that the systems cited represent the only
way to achieve the desired reductions.
With cognizance of all of the uncertainties involved, the
following tabulation is presented to compare typical emissions
in terms of pounds per air dry ton of unbleached pulp. Auxiliary
boilers are not included in the total.
7-4
-------
_TRS Particulate
Old Mill Total 18-22 18-25
Recovery Boiler 12-16 10-11
Upgraded Old Mill Total 10 3
Recovery Boiler 5 2
Relatively New Mill Total 4 4
Recovery Boiler 2 3
Latest Design Mill Total 2 3
Recovery Boiler 0.1 2
Recent emission control regulations for kraft mills such as
in the states of Washington and Oregon have limitations
broadly tabulated as follows (pounds per air dry ton of pulp
produced):
1975
TRS Recovery Furnace
Noncondens ibles
Particulates Recovery
Furnace 4 4
Particulates Other Sources 1.5 1.5
It should be noted that of the examples cited, which do not
represent all possibilities, only the relatively new mill
and the latest design mill can meet the 1972 TRS standards
and only the latest design mill can meet the 1975 TRS stand-
ards . All except the old mill can meet the particulate
standards. This points out that the state of the art of
particulate collection is more advanced than that of odor
reduction. Note also that the proportion of TRS produced
from the recovery boiler system (including the DC Evapo-
rator) has been reduced from 65 - 70 percent of the total to
about 5 percent. As the major sources of TRS are brought
under control, sources formerly considered as insignificant
become of importance.
7-5
-------
Ths impact of standards for the TRS emissions from kraft
recovery boiler systems could be dramatic. Apparently,.
there is a consensus that standards set for 1972 (in
Washington and Oregon) can be met by the utilization of
BLO or by conversion to one of the "new" design evaporator
systems. Standards proposed for 1975, however, might only
be met by scrapping existing process equipment and replac-
ing with a completely new system which eliminates direct
contact between the flue gas and black liquor. If BLO can
be utilized effectively, or if a conversion of existing
facilities can be successfully achieved, the capital costs
(for a 500 TPD null) could range from $200,000 (for BLO)
to $1.5 million (for an air contact evaporator). If it is
determined, however, that a new recovery boiler, per se,
is required, then the capital costs could approach $8
million.
7.2.1 PRECIPITATORS
For a more general discussion of precipitators, the reader
is referred to Chapter 5.
Applicable Emission Sources. Within the chemical wood pulp-
ing industry, precipitatcrs have found extensive applications
to recovery systems; however, application to other sources in
the industry is very limited. Other than direct contact
evaporator recovery systems, specific precipitator applica-
tions are reported for a Venturi scrubber recovery system
(See page 6-6) and a bark and coal fired combination boiler.
The bark and coal fired application is the first reported
installation on a bark fired boiler. This unit is expected
to start up in 1970. The ability of the precipitator to
collect bark char is of interest to the industry. Bark char
is expected to be very difficult to collect due to the high
carbon content which results in a low electrical resistivity.
Precipitators have found limited application to coal fired
boilers in the industry. This is attributed to the fact
that most of the coal fired units are stoker fired and their
emissions heretofore have not been considered excessive.
While the most modern coal fired units are pulverized fired,
their capacity is usually below 500,000 Ibs./hr. steam capacity
and mechanical collectors have been considered sufficient con-
trol. It is expected that future coal fired units .(both
stoker and pulverized) will have provision for a precipitator—
either initially or allowance provided for future installation.
This will result from more stringent air pollution regulations.
7-6
-------
Efficiency. Precipitators are limited to the removal of
participates only. SC>2 and TRS are not affected. Where
precipitators are installed as a primary collector, their
efficiency may be guaranteed as high as 99 + percent for
either recovery systems or coal fired boilers. More
common practice for modern design recovery systems is a
precipitator efficiency of approximately 99 percent.
More important than the extremely high efficiency is the
reliability of operation which one buys with a modern unit
of this type.
For coal fired units provisions are usually made for possible
future installation. It must be recognized that regulations
requiring the burning of low sulfur coal may have a deleterious
effect on precipitator efficiency by altering the condition of
the flue gases.
Flexibility. Precipitators are very flexible in handling load
variations provided the design capacity is not exceeded. Even
when the design capacity is exceeded, the precipitator will
still remove particulate at a reduced efficiency.
For recovery systems, the flue gas must be at a temperature
no less than 275°F - 300°F. At lower temperatures, excessive
corrosion may occur in the precipitator.
For coal fired boilers a low sulfur content of the fuel may
significantly reduce the efficiency of a precipitator.
Economics. Precipitators are expensive control methods (See
Chapter 5). However, they are economically justified on
recovery boilers up to a total annual incremental cost which
does not exceed the value of the annual collected salt cake
($30 per ton has been used in this report for the cost of
salt cake). In the analysis performed in Chapter 13, it will
be seen that this break-even point is in the range of 90 - 95
percent for a new modern installation.
Capital costs and net annual costs, for secondary precipitators
or replacement precipitators to upgrade particulate collection
efficiency for an existing installation, generally range as
follows per ADT/Day:
Capital Cost Net Annual Cost
$1000 to $2000 $300 to $600
7-7
-------
Reliability. Modern precipitator design for recovery systems
utilize double or triple chamber units, multiple electrical
fields, and sophisticated solid state controls. All of these
provide high reliability.
Even with high reliability, unforeseen circumstances may still
occur that result in high short term emissions. For example,
a transformer rectifier failure may result in approximately a
10 percent decrease or more in precipitator efficiency until
it can be replaced.
Adaptability. While precipitators are adaptable to a variety
of applications throughout the U. S. A., their application in
the wood pulping industry is primarily limited to recovery
systems because of cost, corrosion, and process considerations.
7.2.2 VENTURI SCRUBBERS
Applicable Emission Sources. Venturi scrubbers have been
applied to recovery systems and lime kilns in the pulp and
paper industry.
For modern design, single stage Venturi scrubbers are used
almost exclusively for lime kilns.
Single stage and two stage Venturi scrubbers have been used
for recovery systems, with the vast majority of units being
single stage. Approximately 30 recovery systems utilize single
stage Venturi scrubbers while there are approximately three
two stage installations. Due to relatively low particulate
collection efficiency and plume visibility, Venturi scrubbers
are seldom installed today for kraft mill recovery furnace
applications.
Efficiency. For lime kilns, Venturi scrubber units designed
for as high as 99 + percent collection of lime solids and
90 - 95 percent collection of soda fume are frequently applied
for modern installations.
For recovery systems, single stage Venturi scrubber units
with design efficiencies up to 94 percent particulate collec-
tion are in operation. Several two stage Venturi units have
been installed for design efficiencies in the order of 99
percent particulate collection; however, test data from these
installations have not been published.
7-8
-------
Design efficiencies may vary considerably from the AOE
(Annual Operating Efficiencies) of an installation. This
is attributed to the fact that the collection efficiency of
a Venturi scrubber is dependent upon the pressure drop
across the scrubber. This in turn is dependent upon
operating the unit reasonably close to design flow conditions,
or providing an adjustable Venturi throat. Since existing
recovery installations do not incorporate variable Venturi
throat designs, the AOE may vary as much as 10 percent from
the design efficiency; depending upon the operating variables
at a particular mill.
The low flue gas temperature and high moisture content of the
flue gases from a Venturi scrubber may be a disadvantage be-
cause of plume rise and plume visibility considerations.
Venturi scrubbers may also be effective in reducing gaseous
emissions; however, this has yet to be clearly demonstrated
in the industry. For lime kilns, the scrubbing medium may
affect the odorous emission. Therefore, fresh water rather
than contaminated condensate should be used to minimize
odorous emissions due to stripping.
For recovery systems with Venturi scrubbers which utilize
black liquor for a scrubbing medium, odorous gases may be
either absorbed or released depending upon factors such as:
concentrations of malodorous gases in the flue gases
entering the scrubber, pH of the black liquor, and the
sodium sulfide concentration in the black liquor. In view
of the current emphasis on minimizing pollutant emissions
from recovery systems, these Venturi scrubbers which utilize
black liquor are not expected to receive significant con-
sideration in modern practice.
Flexibility. Venturi scrubbers which incorporate variable
throat designs are very flexible providing the design capacity
is not exceeded. Flows above design capacity are usually
limited due to horsepower limitation of the induced draft fan.
Economics. Venturi scrubbers are one of the lowest cost con-
trol methods, however,some specific installations may require
expensive materials of construction to resist corrosion and
thereby negate some of this low cost. Capital cost and net
annual cost ranges per ADT/Day are:
Capital Cost Net Annual Cost
On Recovery Furnaces $300 to $600 $130 to $170
On Lime Kilns $150 to $250 $ 45 to $ 80
7-9
-------
Reliability. The reliability of the Venturi scrubber is
usually very high with most outages being attributable to
auxiliary equipment such as the induced draft fan.
Adaptability. Venturi scrubbers are adaptable to collection
of any particulate which may be handled in a wet system.
Further, they may be adapted to the control of gaseous emis-
sions. An example of this is the use of caustic as a scrubbing
medium as discussed in Section 6.3.1.10.
7.2.3 CYCLONIC SCRUBBERS
Applicable Emission Sources. Cyclonic scrubbers are most
effective as a secondary control method added to an existing
particulate control device. Methods considered in Chapter 5
include the addition of a cyclonic scrubber following a
recovery boiler precipitator and following a bark boiler
mechanical dust collector.
Efficiency. The efficiency of cyclonic scrubbers is a maximum
of approximately 85 percent on relatively coarse particles and
drops off considerably for particles less than 2 microns.
Cyclonic scrubbers may also be effective as gas absorbers, de-
pending upon the particular application and the scrubbing liquid.
Flexibility. Scrubber performance will generally be affected by
variations from design pressure drop. In addition, variations of
particle sizes may have a pronounced effect on scrubber perform-
ance.
Economics. Costs for cyclonic scrubbers are generally quite low
(See Chapter 5) as tabulated below per ADT/Day:
Capital Cost Net Annual Cost
On Recovery Furnaces $300 to $600 $65 to $90
On Slakers $ 30 to $ 60 $10 to $17
Reliability. Cyclonic scrubbers have high reliability with little
possibility of plugging.
Adaptability. The adaptability of cyclonic scrubbers is generally
good; however, on large capacity units, size of the equipment may
become an important consideration.
7-10
-------
7.2.4 RECOVERY SYSTEMS WITHOUT DIRECT CONTACT EVAPORATORS
Applicable Emission Sources. Recovery systems which eliminate
direct contact between the flue gas and black liquor may be
applied to existing recovery systems, or may be substituted
for existing recovery systems. These are systems which re-
ceive black liquor at a concentration of 62 percent solids
or recovery systems which utilize the air contact evaporator.
These systems minimize the emission of odorous gases but have
no effect upon the particulate matter. In fact, the particu-
late concentration in the flue gas will increase because the
direct contact evaporator, which has a capability of removing
approximately 50 percent of the entering dust loading, has
been omitted. Therefore, the particulate dust loading at the
precipitator inlet for a system which eliminates direct con-
tact between the flue gas and black liquor will be approxi-
mately double the dust loading for a conventional recovery
sys tern.
Efficiency. The efficiency of these new recovery systems is
difficult to define in percentage terms. Currently, both of
the new recovery systems marketed in the U. S. are being
designed for low hydrogen sulfide emissions.
As stated previously, the particulate loading for these sys-
tems is expected to increase by a factor of approximately two
at the precipitator inlet.
Flexibility. Since recovery systems which eliminate the direct
contact of flue gas with black liquor are relatively new in the
U. S., and the first one or two installations have been started
up only late in 1969, their flexibility in operation is yet to
be conclusively demonstrated and documented. However, the new
recovery systems should approximate the operating flexibility
of the conventional recovery systems.
Economics. The economics of these systems are more completely
discussed in Chapters 5 and 6; however, of all the control
methods reviewed, these new recovery systems are certainly the
most expensive.
Capital costs range from $9,000 to $18,000 per ADT/Day for a
new recovery system as reported in Chapter 5. For conversion
of an existing recovery system (See Chapter 6) to the new
design, the capital costs are $1,000 to $3,000 per ADT/Day.
Because of variations in chemical recovery costs from mill to
mill, net annual cost ranges are not tabulated.
7-11
-------
Reliability. Since these new recovery systems have only
recently been placed in operation, the reliability has yet
to be demonstrated and documented.
While these units have been designed for low H S emissions,
upsets in operation or equipment malfunctions can certainly
be expected to increase emissions on an annual basis. For
example, if upsets should occur for a total aggregate hours
during a year equivalent to 14 days at a concentration of
500 parts per million H S; then the annual average concen-
tration of H S might be increased by a factor of two or more.
7.2.5 BLACK LIQUOR OXIDATION
Applicable Emission Sources. Oxidation systems are either
the concentrated black liquor type (approximately 50 percent
solids) or the weak black liquor type (approximately 15 per-
cent solids). These systems are used primarily to reduce
malodorous emissions from the direct contact evaporator of
the conventional recovery system. The weak black liquor type
will also reduce malodorous emissions from the multiple effect
evaporators. There is no effect on particulate emissions.
Efficiency. Either oxidation system can be designed for the
same high efficiency of oxidation. The efficiency is based
on the amount of sodium sulfide which is converted to sodium
thiosulfate. This in itself is not a true indicator of the
level of sodium sulfide in the black liquor which is received
by the direct contact evaporator. A sodium sulfide level of
approximately 0.10 grains per liter is desired to minimize
odorous emission from the direct contact evaporator. This
level of 0.10 grams per liter is considered to be the lowest
concentration that can be accurately analyzed and guaranteed.
It should be remembered that the oxidation units themselves
may be a source of odorous emissions in the exhaust gases.
The effectiveness of BLO should be evaluated in terms of the
TRS reduction realized at the DC evaporator, minus the TRS
emitted at the oxidation unit itself.
Flexibility. The concentrated black liquor oxidation system
provides more flexibility than the weak black liquor oxidation
system. This is due to the fact that the weak system utilized
high air flow with minimum liquor storage; while the concen-
trated system utilized approximately 3-4 hours of liquor
storage with less air flow. Since the weak system has less
retention time, the oxidation efficiency is primarily a
7-12
-------
function of the air flow. Therefore, if higher than design
idlei 'loadings* in the xweak \Liqtfor ; are .introduced
ak-vsyst]effiy< a higher sodium isul>fider 'loading^ willl
: the Uiii-.ti?Y^tbfe , addMibnal - 'reientioA fvdlume; of ithe < con-
Sy'item p'isovides L reaction ; tiraeof ore 'fluctuations cin
sttdiuift : sulf idfe- "loadiKgs^and 'the i influenced >of i'iodium ; sulfide
As shown in Section 5.2.7, the weak system is also subject
to an increase in the sodium sulfide concentration as the
liquor flows through the multiple effect evaporators. Since
the concentrated system is installed following the multiple
effect evaporators, the concentrated system is not subject
to this increase in concentration of sodium sulfide.
Economics . Installed costs for oxidation systems are detailed
in Chapter 5.
Capital costs and net annual costs range from $300 to $800 per
ADT/Day and $130 to $260 per ADT/Day, respectively.
For mills which operate with high sulfur losses, oxidation
systems may receive a significant economic credit for chemical
make-up savings in the form of salt cake. For more modern
mills, oxidation systems should be classified as strictly air
pollution control expenditures; since the sulfur losses are
usually minimal insofar as chemical savings are concerned.
Even some older mills which generate substantial quantities of
by-product sulfur compounds could not achieve any chemical
savings with oxidation systems.
Reliability. Adequate data are not available to assess the
reliability of oxidation systems. However, if both types of
systems — weak and concentrated — are designed on a comparable
basis, the reliability would also be comparable.
Adaptability. With proper design, either system is considered
to be equally adaptable. The concentrated system, however,
has more flexibility or adaptability for handling varying
sodium sulfide concentrations for a specific application. For
foaming liquors, only the concentrated system can be used for
oxidation .
In general, oxidation systems are considered to be the most
promising control method to reduce malodorous emissions from
recovery systems in older mills .
7-13
-------
7.2.6 ORIFICE SCRUBBERS
Applicable Emission Sources. The orifice scrubber is an
effective means of controlling emissions from dissolving
tank vents. This application has been discussed in detail
in Section 5.3.6.2.
Efficiency. Orifice scrubbers have a maximum efficiency of
approximately 99 plus percent on particulate. Depending on
the scrubbing liquid used, orifice scrubbers may also be
effective on gaseous sulfur compounds. Scrubbing liquids
are discussed in considerable detail in paragraph 5.3.6.2.a.
Flexibility. Operation of the orifice scrubber is closely
related to the pressure drop (approximately 8 inches), and
variations from design may have a significant effect on
efficiency.
Economics. Costs have been developed in Chapter 5 and are
moderate for most dissolving tank installations.
Capital costs and net annual costs range from $60 to $120
per ADT/Day and $35 to $50 per ADT/Day, respectively.
Reliability. The reliability of the orifice scrubber is
usually very high with most outages attributable to auxiliary
pumps and fans.
Adaptability. Orifice scrubbers may be used for other sources
where moderate pressure drop control methods are acceptable.
The orifice scrubber is adaptable to most dissolving tank vent
configurations.
7.2.7 MECHANICAL COLLECTORS
Applicable Emission Sources. Mechanical collectors are
effective control methods for boilers burning bark alone or
in combination with other fuels. Bark char is difficult to
collect, and the application of mechanical collectors to this
material has been discussed in detail in Chapter 5. Existing
installations normally utilize large diameter cyclones or
multi-tube collectors arranged for single or double stage
collection. Several new designs for mechanical collectors
for bark have been proposed during the last few years. One
such design is the addition of a shave-off scrubber as dis-
cussed in Chapter 6.
7-14
-------
Efficiency. The effectiveness of mechanical collectors is
limited to the collection of participate with no effect on
gaseous emissions. Efficiency for bark collection ranges
from a maximum of 92 percent for single stage units to a
maximum of 96 percent for a two stage collector or a single
stage with shave-off.
Flexibility. Mechanical collectors are sensitive to particle
size and density, and fluctuations in either can be expected
to affect collector performance. In addition, variations
from design pressure drop may significantly affect perform-
ance.
Economics. Costs for mechanical collectors have been developed
in Chapters 5 and 6. These costs are usually low for normal
installations. As demonstrated by the flow diagrams in Chapter
3, the required capacity of these units varies widely. Ranges
of costs per ADT/Day are as follows:
Capital Costs Net Annual Costs
Single Stage $100 - $350 $25 - $: 85
Two Stage $200 - $750 $50 - $175
The cost of these devices is entirely for air quality control
since there are no credits.
Reliability. The reliability of mechanical collectors can be
adversely affected by plugging and hopper fires; however, these
problems can generally be precluded with proper design.
Adaptability. With many arrangements available, the adaptability
of mechanical collectors is very good.
7.2.8 INCINERATION IN LIME KILN
Applicable Emission Sources. Incineration of noncondensible,
malodorous gases in lime kilns is a control method which has been
developed by mills on the West Coast. This control method is
usually applied to noncondensible gases from the multiple effect
evaporators, the digester relief and blow, and the turpentine
condenser.
All three of these sources have a relatively low total gas
volume (in the order of 200 to 300 cubic feet/ADT). Existing
lime kilns have the capability of handling these volumes. Other
7-15
-------
sources might be considered for this control method; however
the quantity of gases which may be handled by an existing
kiln is fixed due to the physical limitations of the installed
equipment. Therefore, any additional sources must be small in
volume and should be judiciously investigated.
Efficiency. Incineration does not remove any particulate matter.
The malodorous gases of H S and TRS are almost completely
destroyed by incineration. Thermal incineration converts these
malodorous gases to SO and water vapor. Limited field studies
by industry engineers indicate that the concentration of SO
is removed.in. the lime kiln system.
Flexibility. If the incineration system is properly designed,
adequate flexibility is provided for day-in, day-out operations.
Of course, during malfunctions or shutdowns of the lime kiln,
these gases must be vented to the atmosphere or a second auxili-
ary control method must be provided.
Economics. More detailed costs for this control method are
provided in Chapter 5. The cost for this control method is
entirely an air quality control cost since credits for recovered
chemicals are insignificant. Capital cost and net annual cost
per ADT/Day generally range as follows:
Capital Cost Net Annual Cost
Batch Digesters $100 - $200 $40 - $75
Continuous Digesters $ 50 - $100 $20 - $35
Reliability. In the past, this control method has only been
moderately reliable. However, due to engineering refinements
and more experience from those mills where it is practiced, this
control method may now be considered as a reasonably tried and
proven method which will provide adequate control of gaseous
emissions from the above sources.
Adaptability. As previously stated, this control method is
limited to relatively small gas volumes.
7-16
-------
7.2.9 ENCLOSED PRESSURE WASHING
Applicable Emission Sources. Enclosed pressure washing is a
patented system based on operating the washer cylinders with
pressure rather than under a vacuum. This system eliminates
the need for suspended hoods over the washers and the high
exhaust gas flows.
Efficiency. An efficiency cannot be applied to pressure washers
because this system would be used only as an alternative to a
system using vacuum washers. Because the pressure washer is
completely enclosed with an airtight hood, emissions due to
washing can be practically eliminated.
Flexibility. The flexibility of pressure washing has not been
documented and is unknown.
Economics. The cost of a pressure washing system is approxi-
mately the same as for a vacuum washing system. As discussed
in Chapter 6, the capital cost is in the order of $1,000 to
$3,000 per ADT/Day. Net annual costs were not calculated.
Reliability. Because there are no known installations of
pressure washers in the U. S., it is not possible to evaluate
the reliability of pressure washers.
Adaptability. Pressure washing is expected to be more applicable
to a new mill installation or to an expansion at an existing mill.
For existing mills, pressure washing will probably receive more
acceptance than diffusion washing; however, this must be investi-
gated further.
7.2.10 CONTINUOUS DIFFUSION WASHING
Applicable Emission Sources. Continuous diffusion washing is a
patented system designed to replace the existing system of washing
with vacuum cylinders with their suspended hoods and resultant
high exhaust gas flows.
Efficiency. An efficiency cannot be applied to diffusion washing
because this system would be used only as an alternative to a
vacuum cylinder washing system. Because the diffusion washer is
enclosed and because only washed stock, relatively free of mal-
odors, is exposed to air at the top discharge, emissions due to
washing can be practically eliminated. Should washing not be
complete because of overload, the small quantities of gases from
the vent can possibly be piped to a lime kiln or furnace for
incineration.
7-17
-------
Flexibility. The flexibility of continuous diffusion washing
has not been documented and is unknown.
Economics. The diffusion washer can be installed on the top
of a high density storage tank and, therefore, there is no
need for an operating floor or washer building such as re-
quired with vacuum cylinder washers. In addition, large seal
tanks are not required. This results in significant capital
savings. Capital cost is in the order of $1,000 per ADT/Day.
Net annual costs were not calculated.
Because consistency can be kept high in the washing system,
the horsepower required for predilution to vacuum washers is
eliminated, creating a savings in operating cost.
Reliability. The first brown stock continuous diffuser-washer
was put into operation in Sweden in 1964. The second diffuser-
washer was installed in Canada in 1968 where problems were
encountered due to cold weather. Minor operational upsets
were encountered.
Adaptability. Continuous diffusion washing is expected to
be more applicable to a new mill installation or to an expan-
sion at an existing mill.
For existing mills, numerous factors such as space, operation,
and existing equipment would probably preclude the installation
of continuous diffusion washing.
7.3 SULFITE PROCESS
Control methods applied to the sulfite pulping process consist
primarily of methods to remove sulfur dioxide gases. The
quantity of SO gases that can be released to the atmosphere
from any one mill must be determined on an individual mill
basis. This is a function of the ambient air quality standards,
existing concentration of SO in the atmosphere prior to the
construction of the mill, meteorology, topography, and mill
design.
Work is currently underway in some states to establish the
acceptable level of emissions from a sulfite mill. Certainly,
additional testing, sampling, and development work appear to be
required in the field of sulfite SO emission control. Reliable
monitors need to be developed for tnese sources.
7-18
-------
7.3.1 PACKED TOWER SCRUBBERS
Applicable Emission Sources. Packed tower scrubbers are
effective control methods for reducing gaseous emissions and
recovering SO from sulfite blow pits. In addition, packed
towers have found application to kraft dissolving tank vents.
Efficiency. The packed tower is a good gas absorber with an
efficiency of approximately 95 percent on SO absorption as
stated in paragraph 5.3.7.2. Collection of particulate is
limited because of possible dust build-up and consequent
plugging.
Flexibility. Packed towers are usually very flexible, with
the gaseous absorption efficiency increasing for reduced
flows. However, carryover of the scrubbing medium may occur
when the gas flow exceeds the design flow.
Economics. A major advantage of packed tower scrubbers is
their low cost. In many sulfite applications, the value of
recovered SO more than offsets the equipment and operating
costs resulting in a net savings(See Chapter 5).
Capital costs and net annual costs (savings) per ADT/Day
range as follows:
Capital Cost Net Annual Cost
(Savings)
Packed Tower for Acid
Tower
$ 12 - $ 20
$ 6 - $ 9
Packed Tower for
Blowpit
$ 450 - $ 750 ($ 75) - ($125)
Packed Tower and Con-
denser for Blowpit $1200 - $2000 ($300) - ($550)
Reliability. The potential for plugging may seriously affect
the reliability of packed towers when particulate is present.
Therefore, packed towers must be used with care in situations
where insoluble particulate is present.
7-19
-------
Adaptability. Due to their simple design and reasonable size,
packed tower scrubbers are easily adapted to most installations.
7.3.2 AMMONIUM LIQUOR INCINERATION AND SO RECOVERY
Applicable Emission Sources. This control method is applicable
only to ammonium base sulfite mills. This control method would
probably be used when an ammonium base sulfite mill is ordered
to remove the effluent from liquid waste streams and provide
some other means of disposing of this waste. The SO recovery
system which is a part of the ammonium liquor incineration sys-
tem would also be applicable to existing ammonium base mills
which presently incinerate their liquor.
Efficiency. While some particulate will be removed, the
quantity is not significant.
In regard to sulfur dioxide, this control method is expected to
remove approximately 95 percent of the entering SO , as described
in more detail in Chapter 6.
Flexibility. The flexibility of this control method has yet to
be demonstrated and documented since it is not in operation in
the U. S.
Economics. As pointed out in Chapter 6, this control method is
expected to be very expensive. The capital cost and net annual
cost are in the order of $12,000 per ADT/Day and $1,300 per
ADT/Day, respectively.
Reliability. Since this control method is not in operation in
the U. S., the reliability is an unknown factor.
Adaptability. Adaptability is also an unknown factor with the
exception that a significant amount of space will be required for
the installation.
7.3.3 IMPROVED SO RECOVERY IN MAGNESIUM SULFITE PROCESS
£
Applicable Emission Sources. This control method involved
addition of an absorption unit to an existing MgO recovery
boiler with an SO recovery system.
7-20
-------
Efficiency. The concentration of particulate in this flue
gas stream is present only in trace amounts. Therefore,
particulate removal efficiency is not of concern. This
control method would be primarily installed for additional
removal of SO gases. Since the installation of this control
method following an existing SO recovery system which operates
at approximately 95 percent efficiency is not in operation in
the U. S., the SO removal efficiency is unknown.
Flexibility. The flexibility of this control method must also
await an actual installation and testing.
Economics. Since the amount of SO removal by this control
method would be relatively low, the cost would almost totally
be attributable to air quality control. Capital cost and net
annual cost are in the order of $1,000 per ADT/Day and $250 per
ADT/Day, respectively.
Reliability. Since this control method has yet to be installed,
the reliability is unknown; however, the reliability should be
comparable to the existing absorption units at a particular mill.
Adaptability. The adaptability of this control method for
various mills will depend upon existing space conditions. While
absorption towers are adaptable to other sources, this particular
control method is based on the application to only one source
which is an MgO recovery boiler which incorporates an SO
recovery system.
7.4 NSSC PROCESS
As pointed out in Chapter 5, data are not available on the
pollutants from NSSC pulping processes. Therefore, the develop-
ment of control methods for this pulping process must await
additional data from sampling and testing.
7-21
-------
CHAPTER 8
POWER BOILER SULFUR RECOVERY
TABLE OF CONTENTS
Page No.
Summary 8-1
Introduction 8-2
Limits on S02 Emissions 8-3
Requirements on Make-Up Sulfur in the Industry 8-5
Estimates of Fuel Usage and SO Provided by
Power Boilers 8-7
Make-Up Sulfur/Sulfur Loss Ratio 8-14
Fuels Consumed - Paper and Allied Products Industries 8-15
Flue Gas Desulfurization Technology 8-19
Process Feasibility Considerations 8-32
R S D Efforts 8~38
References 8-39
8-i
-------
CHAPTER 8
POWER BOILER SULFUR RECOVERY
SUMMARY
Many mills operate on-site steam-electric generating
plants in the 2 - 100 megawatt range to provide required
additional energy over and above that provided by the
recovery furnace. Where such power boilers are utilized,
the fuels consumed are usually coal, oil, gas, or a com-
bination of fuels. With coal and oil as the fuel, the
potential for sulfur dioxide emissions is present. The
combination of on-site power boilers and a process demand
for soluble sulfur compounds presents a potentially unique
advantage to the industry.
By mid 1970, Air Quality Regions called for by the Air
Quality Act of 1967 will be designated in all of the states.
Sulfur dioxide standards will be among the first to be
promulgated in each state. Thus, the application of strin-
gent sulfur dioxide emission standards to pulp mill power
boilers appears inevitable. It is estimated that if the
mix of fuels remains essentially as it is today, a portion
of the make-up sulfur requirements could be provided from
this source and theoretically all of the recovered sulfur
could be used in the pulping process.
Several flue gas desulfurization processes are under
active development at the present time. Six of these were
considered as having potential application to this situation.
Extensive technical and economic evaluation indicated that
none of the processes had advanced beyond the pilot plant
stage, and high capital and operating costs were predicted
for application on this scale. It was concluded, therefore,
that none of the processes was feasible for application to
this situation in the foreseeable future.
8-1
-------
8.1 INTRODUCTION
The complete combustion of coal and fuel oil, which contain
inorganic sulfides and sulfur-containing organic compounds,
results in the emission of sulfur dioxide into the atmosphere.
Other oxides of sulfur may also be emitted but in quantities
that are small by comparison; for example, about 40 to 80
parts of sulfur dioxide to one part of sulfur trioxide are
emitted from fossil-fueled power plants.
Several techniques for the control of sulfur oxide air pol-
lutants are under intensive investigation. These investi-
gations have been brought about because of concern for the
oxides of sulfur which are produced by such major industries
as smelters and sulfuric acid producers as well as by the
combustion of fossil fuels in steam-electric generating
plants. The bulk of the research activity and the economic
studies to date have been looked at from the point of view
of the electric utility. From this vantage point it is seen
that the control of sulfur oxides is achieved through concern
for air quality. The material removed from the stack gases
may or may not have a market value, and even if it does have
value, the logistics of getting the material to a place where
it can be used often makes the total scheme impractical.
In proceeding chapters of this report, heat and material
balances for the pulping processes of concern were presented.
It is generally agreed that the energy requirements for the
entire kraft pulping and recovery systems can be supplied
from the energy generated by the recovery furnace itself. If
this is true, the questions logically to be asked are . . .
why do pulp mills have power boilers and why consider power
boiler sulfur recovery?
The answer to the first question is related to the way
the industry operates. In Chapter 2 it is pointed out
that most pulp is made by integrated companies and consumed
captively without moving through the market place. About
ten percent of the total pulp produced is, however, made by
independent pulp producers without their own paper making
facilities or by integrated companies producing surpluses
for market.
In either case, energy is required to operate paper machines
and/or to operate pulp dryers. This additional energy (over
and above that required for the recovery operation) must be
8-2
-------
supplied from some external source. The usual approach is
for the company to operate steam-electric generating plants
(power boilers) on-site, although some mills purchase their
energy.
Where power boilers are utilized, the fuels consumed are
usually coa-l-y oil, natural gas, or a combination of fuels.
With coal, and oil as fuel, the potential for sulfur oxide
emissions is present the same as with large steam electric
generating plants. Control of sulfur oxides in this instance
though can be thought of in terms of sulfur recovery. The
combination of on-site power boilers and a process demand
for soluble sulfur compounds presents a potentially unique
advantage to the wood pulping industry.
8.1.1 LIMITS ON SULFUR DIOXIDE EMISSIONS
In January, 1969, "Air Quality Criteria for Sulfur
Qxides" was released by the National Air Pollution
Control Administration. Air quality criteria are an
expression of the scientific knowledge of the rela-
tionship between various concentrations of pollutants
in the air and their adverse effects on man and his
environment. Air quality criteria are descriptive/-
that is, they describe the effects that have been
observed to occur when the ambient air level of a
pollutant has reached or exceeded a specific figure
for a specific time period.
The document "Air Quality Criteria for Sulfur Oxides"
indicates that "adverse health effects were noted
when 24-hour average levels of sulfur dioxide ex-
ceeded 0.11 ppm for three to four days. Adverse
health effects were also noted when the annual mean
level of sulfur dioxide exceeded 0.04 ppm. Visibility
reduction to about five miles was observed at 0.10
ppm; adverse effects on materials were observed at
an annual mean of 0.12 ppm; and adverse effects on
vegetation were observed at an annual mean of 0.03
ppm."
When air quality criteria are incorporated into control
rules and regulations they become air quality standards.
8-3
-------
Air quality standards are prescriptive. They prescribe
pollutant exposures which a political jurisdiction
determines should not be exceeded in a specified
geographic area, and are used as one of several fac-
tors in designing legally enforceable emission standards.
It is reasonable and prudent to conclude that, when
promulgating ambient air quality standards, consid-
eration should be given to requirements for margins
of safety which would take into account long-term
effects on health, vegetation, and materials occurring
below the criteria levels specified above.
One can intuitively appreciate that the level of sulfur
dioxide observed in the ambient air is indeed influenced
by the amount of sulfur dioxide being emitted from some
source or sources. It may seem then that all one would
have to do is to limit the amount of sulfur dioxide
being emitted to that level which keeps the ambient
level below the prescribed amount. If only one source
of sulfur dioxide is evident, this is a relatively
straightforward exercise. There are a number diffusion
equations which can be used to predict what the resulting
ground level concentration will be for a given emission
of sulfur dioxide. When multiple sources are consid-
ered, however, diffusion models must be developed
and these suffer the indignities which only Mother Nature
can explain. Then too, the nagging question always
remains. . . How much of the total allowable emission
shall we reserve for future sources?
Needless to say, the establishment of emission standards
for sulfur dioxide is not easy. Nonetheless, emission
standards have been set in some areas of the country.
A typical expression for the limit of sulfur dioxide
emissions is to wit: it shall be unlawful for any
person to cause or permit the emission of sulfur
compounds, calculated as sulfur dioxide, of more than
2,000 ppm.
Another approach to setting limits on sulfur dioxide
emissions (especially directed toward fossil fuel
fired combustion units) is to set a limit on the sulfur
content of the fuel itself. The rationale for this
approach being that it is easier to analyze for sulfur
in the fuel than for sulfur dioxide in the stack gas,
8-4
-------
and that if you cut the sulfur content of the fuel by
some percentage, you will automatically cut the sulfur
dioxide concentration by an equivalent percentage. A
corollary to this might be to provide that the concen-
tration of sulfur dioxide in the stack gas must not
exceed that which would be obtained by using a fuel
with a specified sulfur content. This approach would
allow a fuel with a high sulfur content to be burned
but would require that at least a portion of the result-
ing sulfur dioxide be removed or recovered.
By mid 1970, Air Quality Control Regions called for
under the Air Quality Act of 1967 will be designated
in all of the states. Since the SO Criteria and
the Control Documents have been published, the clock
is running in those Regions already designated. It
will begin to run in other states when regions are
designated in them. Therefore, it is inevitable that
more stringent sulfur dioxide emissions standards will
be imposed in many areas within the year. That these
will be applied to pulp mill power boilers also would
appear inevitable.
8.1.2 REQUIREMENTS ON MAKEUP SULFUR IN THE INDUSTRY
The sulfur makeup requirements, as obtained from visits to
various operating mills, show a range from 22 to 32 pounds
per ton of air dry pulp for the kraft industry. The sulfur
makeup requirements of NSSC and sulfite processes, limited
though the data are, show a reasonable correlation between
the flow diagram sulfur makeup requirements and the mill
visit data. Flow diagram data should not be used to esti-
mate sulfur makeup for kraft because not all sulfur loss
points are shown. The flow diagrams are oriented to atmos-
pheric emissions of sulfur and do not have complete data
for evaluation of sulfur losses in other process streams.
It is assumed that the sulfur makeup requirements for the
kraft, sulfite, and NSSC processes are as follows:
1. Kraft - 26 pounds per ton A.D. pulp
2. NSSC - 60 pounds per ton A.D. pulp
3. Sulfite - without recovery - 243 pounds per ton A.D. pulp
4. Sulfite - with recovery - 51 pounds per ton A.D. pulp
These sulfur makeup requirements may be high or low for
an individual mill, but are suggested as typical for the
industry. These values are summarized in Table 8-1.
8-5
-------
TABLE 8-1
SUGGESTED VALUES FOR SULFUR MAKEUP - POUNDS PER TON A.D. PULP
Kraft NSSC Sulfite
Wi thout Without With
Recovery Recovery Recovery
26 60 243 51
The form in which the makeup sulfur is supplied depends on the
pulping process employed. Of the three pulping processes under
consideration, the kraft process is the most flexible as to the
form in which the sulfur can be supplied. The most common means
of sulfur addition is as sodium sulfate just prior to the re-
covery boiler. Other forms in which sulfur may be introduced
are elemental, sulfuric acid, SO , SO , sodium sulfite, and
sodium sulfide. This flexibility of the form in which sulfur
may be added is because of the highly alkaline nature of
the black liquor which ties up any of the acid forms of sulfur
as the various sodium salts and the reactions in the recovery
boiler which cause the reduction of almost any sodium sulfur
compound to the desired sodium sulfide.
The flexibility of sulfur addition applies only to the form
of sulfur and the usual sodium-sulfur balance must be main-
tained to ensure the required sulfidity in the pulping liquor.
Large amounts of some forms of sulfur added may require additional
makeup chemicals to maintain the proper sodium-sulfur ratio.
This is determined by normal mill process controls.
The NSSC pulping process requires that the makeup sulfur be
added as elemental sulfur (to be burned to SO ), SO or sodium
sulfite.
The sulfite process also requires the sulfur makeup as elemental
sulfur (to be burned), S0_ or the sodium, calcium, magnesium,
or ammonium sulfite depending on the pulping process under
consideration.
In the sulfur makeup requirements, no consideration has been
given to sulfur additions which may be supplied from satellite
processing such as crude tall oil production, bleaching, spent
acid streams, other pulping processes in the same mill, or other
8-6
-------
operations which could conceivably supply more sulfur than the
makeup requirements and thereby cause a particular pulp mill
to be long on sulfur. Supplies of sulfur from such captive
sources external to the pulping process can drastically alter
the sulfur recovery economics. Such situations are specific
to individual mi lls.
An example of the sulfur supplied from the processing external
to the pulp mill is crude tall oil production.
Typically, about 180 pounds of black liquor skimmings (soap)
are produced per ton of air dry pulp made from pine. This is
the average of the range of 120 pounds to 240 pounds which is
the soap yield variation due to geographical factors. The
same species of pine grown in the extreme Southeast (north
Florida for example) has a much higher fatty acid-rosin acid
content than the trees grown more north and west. The higher
fatty acid and rosin acid content yields more soap.
Normal soap requires about 295 pounds of 78 percent sulfuric
acid (range 280 pounds to 310 pounds dried) per ton of soap to
produce crude tall oil. The spent acid is returned to the
pulp mill chemical recovery.
Assuming no sulfur losses in the tall oil processing, this
would return to the mill as spent acid the equivalent of 6.. 8
pounds of sulfur per ton of pine pulp or about 26 percent of
the 26 pounds of makeup sulfur.
It must be remembered that this factor of sulfur returned
applies only to the pulp produced from pine and must be
apportioned on the pine-hardwood ratio fed to the digesters.
All kraft mills have the capability of producing tall oil and
any sulfur recovery projections should consider this factor.
Furthermore, some mills with crude tall oil production facilities
may be purchasing soap from outside sources either consistently
or sporadically, which will increase the sulfur return to the
mill beyond that calculated from their own soap processing.
Other external sources of sulfur must be considered where
applicable.
8.1.3 ESTIMATE OF FUEL USAGE AND SO PROVIDED BY POWER BOILERS
In 1969, the National Council for Air and Stream Improvement
in a cooperative study with NAPCA issued a steam and power
8-7
-------
boiler survey questionnaire to the pulp and paper industry. One
of the questions asked dealt with the annual average consumption
of coal, oil, gas, and bark. Of the questionnaires mailed out,
282 were returned.
Table 8-2 contains a summary of the fuel consumption data reported
in the questionnaires. The data are stratified by use category to
show the "mix" of fuels used in the industry.
The calculation for equivalent BTU for each fuel was made on the
basis of heating values as follows:
Coal 14,000 BTU/Lb.
Oil 140,000 BTU/Gal.
Gas 1,000 BTU/Ft.3
Bark 4,500 BTU/Lb.
8-8
-------
TABLE 8-2
SUMMARY OF ANNUAL FUEL CONSUMPTION DATA
NCASI-NAPCA STEAM & POWER BOILER SURVEY
USE
CATEGORY
Coal (Only)
Coal s Oil
Coal & Gas
Coal & Bark
Coal, Oil &
Gas
Coal, Oil &
Bark
Coal , Gas &
Bark
Coal, Oil, Gas
& Bark
Oil (Only)
Oil & Gas
Oil S Bark
Oil, Gas &
Bark
Gas (Only)
Gas & Bark
Bark (Only)
No Information
Available •
TOTALS
EQUIVALENT BTU
NO. OF
REPLIES
65
6
3
11
6
10
4
4
67
23
16
18
17
23
0
9
282
do12)
COAL OIL 10 FT3 BARK
TONS BBL TONS
4,163,075
443,450 717,500
129,850 — 3,395
1,955,800 — — 1,181,600
282,100 118,333 2,414
1,589,000 2,260,200 — 840,000
473,900 — 4,540 153,650
388,850 77,933 4,442 259,350
11,509,825
1,039,950 20,657
7,177,400 — 1,911,350
6,310,700 74,696 3,188,325
22,631
75,883 2,510,550
—
9,426,025 29,211,841 208,658 10,044,825
263.9 17.2 208.6 90.4
BASED ON 350 DAYS PER YEAR
8-9
-------
It is immediately obvious that data were supplied by more mills
than are listed in Chapter 2. This means that some of the fuel
reported in Table 8-2 was consumed at mechanical pulp mills and
at paper and paperboard mills which were not on-site with a
pulp mill.
Table 8-3, therefore, contains data taken from questionnaires
on which a production of chemical pulp was indicated. It is
seen from Table 8-3 that useable information was received'from
120 mills with a combined nominal capacity of 70/263 tons per
day of A.D. pulp. This represents data from 60 percent of the
chemical pulp mills accounting for 64 percent of the 1968 pulp
tonnage capacity.
8-10
-------
T"A B L E 8-3
ANNUAL FUEL CONSUMPTION DATA - CHEMICAL PULP PRODUCERS
NCASI-NAPCA STEAM & POHER BOILER SURVEY
CATEGORY
Coal (Only)
Coal & Oil
Coal s Gas
Coal & Bark
Coal, Oil s
Gas
Coal, Oil &
Bark
Coal , Gas &
Bark
Coal, Oil,
Gas & Bark
Oil (Only)
Oil & Gas
Oil & Bark
Oil, Gas &
Bark
Gas (Only)
Gas & Bark
Bark Only
No Information
Available
TOTALS
Equivalent BTTJ \
CAPACITY
TONS
A.D. PULP
PER DAY
2,348
1,466
240
3,443
185
8,265
1,380
2,820
2,561
2,421
9,747
17,270
2,596
15,521
—
4,440
74,703
non)
NO. OF
REPLIES
12
4
2
9
2
10
2
4
10
8
14
18
4
21
0
3
123
COAL OIL 106 BARK
TONS BBL FT TONS
1,598,800
351,400 490,000
71,750 — 2,275
1,749,300 — -- 1,146,600
149,800 35,833 1,687
1,589,000 2,260,117 — 840,000
439,600 — 3,507 172,900
388,850 77,933 4,442 259,350
3,696,967
23,718 12,191
6,944,650 — 1,911,525
6,281,300 72,403 3,178,350
9,770
73,307 2,330,300
—
—
6r338,500 19,810,518 179,582 9,839,025
177.5 11.6 179.6 88.6
Based on 350 Days Per Year
8-11
-------
Taking -che totals from Table 8-3 and scaling them upward on the
basis of 1968 pulp production ratioed to 80 percent of the
nominal capacity reported (86, 571/56, 210) we get an estimate
of the total quantities of fuel consumed in power boilers in
1968 in locations where chemical pulp is produced on-site. This
estimate is shown in Table 8-4. A figure of 80 percent of nominal
capacity reported was used because a comparison of data in Table
2-1 with data in Table 2-4 suggests that on the average, mills
were operating at 80 percent of capacity.
TABLE 8-4
ESTIMATED FUEL CONSUMPTION - 1968
AT MILLS WHERE CHEMICAL PULP IS PRODUCED
Coal
9,761,290 Tons
273.3 x 1012BTU
Oil
30,508,198 Bbl.
179,4 x 1012BTU
Gas
276,566 x 106Ft3
276.6 x 10l2BTU
Bark
15,152,099 Tons
136.4 x 1012BTU
Again, taking data from Table 8-3 one can generate Table 8-5 which
represents the quantities of the various fuels consumed per year per
ton of A.D. pulp per day. Please note that these figures do not
necessarily equate themselves on an energy basis. They merely
represent historical data that say, for example, that 113 tons of
coal were consumed in 1968 for each ton of chemical pulp produced per
day.
TABLE 8-5
RATIO OF FUEL CONSUMED PER YEAR PER TON OF
CHEMICAL PULP PRODUCED PER DAY
Coal
113 Tons
3.2 x 109BTU
252 Bbl.
2.1 x 109BTU
Gas
3.2 x 106Ft3
3.2 x 109BTU
Bark
175 Tons
1.6 x 109BTU
If the "mix" of fuels remains constant through 1980, then the data
in Table 8-5 can be coupled with data in Table 2-3 to estimate fuel
consumption for the years 1975 and 1980. These estimates are given
in Table 8-6.
8-12
-------
TABLE 8-6
PROJECTION OF FUEL CONSUMPTION
- CHEMICAL PULP PRODUCERS -
coal Oil Gas Bark
1975 14,270,092 Tons 44,451,968 Bbl. 404,109 x 106Ft3 22,099,700 Tons
399.6 x 10 BTU 261.4 x 1012BTU 404.1 x 1012BTU 198.9 x 1012BTU
1980 17,563,364 Tons 54,710,656 Bbl. 497,370 x 106Ft3 27,199,900 Tons
491.8 x 1012BTU 321.7 x 1012BTU 497.4 x 1012BTU 244.8 x 1012BTU
It is now a straightforward calculation to estimate the
amount of SO generated by the combustion of these fuels in pulp
mill power boilers. If one assumes a sulfur content of two
percent for both coal and oil—this is taken as being representa-
tive of the values given on the steam and power boiler survey
questionnaires—then the values in Table 8-7 can be generated.
TABLE 8-7
TONS OF SULFUR DIOXIDE PRODUCED BY POWER BOILERS
- ALL CHEMICAL PULP PRODUCERS -
Year
1968
1975
1980
Coal
390,452
570,804
702,535
Oil
192,202
280,047
344,677
Basis 2% S for coal and oil
Tons Coal x 0.02 x 2 = Tons SO
BBL Oil x 42 x 7.5 x 1/2,000 x .02 x 2 = Tons
8-13
-------
8.1.4 MAKEUP SULFUR - SULFUR LOSS RATIO
The sulfur makeup requirements presented in Table 8-1 can be
coupled with data in Tables 8-7 and 2-1 to give a national
picture of the balance between sulfur makeup requirements
and the sulfur which conceivably could be supplied from power
boiler flue gases. This relationship is shown in Table 8-8.
TABLE 8-8
INDUSTRY-WIDE DATA ON MAKEUP SULFUR REQUIRED VERSUS
SULFUR AVAILABLE FROM POHER BOILERS
Process Sulfur Required - Tons
Kraft 315,900
NSSC 105,000
Sulfite Without Recovery 218,700
Sulfite With Recovery 17,800
TOTAL 657,400
Potential Sulfur Available291,327
It is readily apparent from Table 8-8 that from the national
standpoint, recovery of all sulfur lost in power boiler flue
gases would still not supply the total required makeup sulfur.
The question might well be asked then, "How does this relate
to specific mill circumstances?" Is it conceivable that a
given mill could lose more sulfur through power boiler flue
gases than is required for makeup, or does the national
picture prevail even at the individual mill level?
To determine if the sulfur makeup requirements derived above
are compatible with the potential sulfur available from power
boiler flue gas, the following analyses were made:
1) Kraft — Looking at the energy balance provided for each
kraft flow diagram one sees that the range of fuel required
is from 386 pounds/ADT to 1180 pounds/ADT. Assuming the coal
to have a sulfur content of two percent, this equates to a
potential sulfur loss of 7.7 pounds to 23.6 pounds/ADT.
8-14
-------
2) NSSC — The range of fuel required for the NSSC processes
is from 1192 to 1550 pounds/ADT. Again, this equates to a
potential sulfur loss of 23.8 pounds to 31 pounds/ADT.
3) Sulfite — The range of fuel required for the various
sulfite processes is from 507 pounds to 1630 pounds/ADT.
This equates to a potential sulfur loss of 10.1 pounds to
32.6 pounds/ADT.
In each case presented above, the potential sulfur recovery
(sulfur loss from power boilers) is less than the corresponding
sulfur makeup requirement for the specified process. It is
concluded then, that if regulations are passed which require
the control of S02 emissions from power boilers, sulfur recovery
could be considered because the potential amount of sulfur to
be obtained could all be utilized in the pulping processes.
The examples cited above were based only on using coal as a
fuel. For equivalent energy, less oil than coal is required
so that the corresponding sulfur loss would also be less. Since
all the sulfur available from coal burning can be utilized, it
is apparent that all sulfur lost from oil burning can likewise
be utilized.
8.1.5 FUELS CONSUMED - PAPER AND ALLIED PRODUCTS INDUSTRIES
As a comparative exercise, it was decided to estimate the
total amount of fuel utilized by the paper and allied products
industry (SIC Group 26). Contact was made with the National
Coal Association, the American Petroleum Institute, and the
American Gas Association regarding compilations of fuel usage
data. All three organizations reported that they do not compile
specific data on types of industries. The American Gas Institute
did have available gas consumption data for the paper and allied
products industries for selected years. All three organizations
referred to the Census of Manufacturers of the U. S. Department
of Commerce.
The data reported in the Census of Manufacturers are given in
'detail only once every five years; the latest available being
the 1963 report covering fuel usages in 1962. In addition to
the detailed Census of Manufacturers, the Department of Commerce
publishes an Annual Survey of Manufacturers in which data are
presented for total purchase price paid for fuels by industry
classification, but it does not report a breakdown of cost by
types of fuels.
8-15
-------
Table 8-9 reports the total purchased cost of fuels versus the
total chemical pulp production for selected years from 1954
through 1966 plus the calculated cost per ton of pulp. It must
be remembered that this total cost reflects the entire paper
and allied products category and not just pulp mills. Because
of fluctuations in this cost, the five year average for 1962 -
1966 of $12.07/ton pulp was chosen as a basis for projection.
The projected values of dollars to be spent on purchased fuels
for the years 1975 and 1980 are also shown in this table. The
projected values were obtained by multiplying the cost per ton
figure of $12.07 by the projected tonnage of chemical pulp.
Table 8-10 was prepared showing the breakdown of this total cost
into types of fuel assuming that the national distribution by
types of fuels will remain unchanged from the 1962 values.
An interesting comparison can be made by viewing the estimates
of Table 8-6 with the estimates in Table 8-10. This comparison
reveals that of the estimated fuel consumed by the paper and
allied products industry, 48 percent of the coal, 91 percent
of the oil, and 87 percent of the gas is utilized by chemical
pulp producers.
8-16
-------
TABLE 8-9
TOTAL PURCHASED COST OF FUEL vs. CHEMICAL PULP PRODUCTION
FOR SELECTED YEARS
YEAR TOTAL PURCHASED
FUEL COST
($1000)
1954
1958
1959
1960
1962
1963
1964
1965
1966
PROJECTED
1975
1980
220
266
274
281
302
305
311
334
355
555
679
,036
,444
,797
,064
,885
,464
,554
,336
,607
,220
,541
CHEMICAL PULP
PRODUCTION
(MILLION TONS)
16
17
19
20
23
24
27
28
30
46
56
.0
.6
.8
.7
.1
.6
.2
.8
.1
.0
.3
COST
$/TON PUL
13
15
13
13
13
12
11
11
11
12
12
.75
.13
.87
.57
.11
'"I
.45 !
.60
.81-
.07
.07
8-17
-------
oo
i
00
1980
TABLE 8-10
FUEL USAGES PAPER AND ALLIED PRODUCTS
1975 and 1980
TOTAL
PTTPPWLCP
YEAR COST
($1000)
1975 555,220
C
COST 1
($1000)
239,855
0 A L
QUANTITY
(1000 Tons)
29,215
O I
I
COST ;
($1000) ;
127,145
i
L ;
QUANTITY
(1000 Bbl)
48,715
G A
COST .
($1000)
161,569
S
QUANTITY
(106Pt3)
464,279
OTHER* - ,i
COST
($1000)
26,651 ;
j
j
679,541 ; 293,562
35,757
155,615 • 59,623 197,746 \ 568,236 ' 32,618
*Gasoline, LPG, Wood and Purchased Steam
BASIS: (1) : Projection of total purchased cost from Table 8-8
(2) Assumption that distribution by type of fuel and the
cost per unit of each type will remain unchanged from 1962
-------
8.2 FLUE GAS DESULFURIZATION TECHNOLOGY
Many flue gas desulfurization processes have been proposed
and a number of them are currently being actively developed.
All of these processes are primarily directed toward the
removal of sulfur oxides from flue gas for air pollution
control. The recovery of sulfur in any marketable form is
a secondary feature which would help defray the cost of
achieving the primary objective. We are concerned only
with the recoverable sulfur processes.
Of the many processes being proposed for the removal of sulfur
oxides, the state of development of most must still be con-
sidered as experimental despite the fact that some are being
advertised as commercial processes. Though the technology
may be sound, as demonstrated by pilot plant and semi-works
scale operations, full commercial size plants have not
been built as yet.
8.2.1 A CRITICAL REVIEW OF APPROPRIATE PROCESSES
.The following processes are considered as having a
potential application for the recovery of sulfur
from power boiler flue gas.
1. Catalytic oxidation
2. Beckwell Process (Wellman-Lord)
3. Stone and Webster Process
4. Reinluft Process
5. Alkaline Scrubbing
6. Kiyoura - T.I.T.
8-19
-------
8.2.1.1 Catalytic Oxidation (2_, 3)
A promising system producing sulfuric acid
from the SO in the flue gases involves
catalytic oxidation of SO to SO . Some
research on catalytic oxidation has been
done by Bituminous Coal Research, Inc.,
an arm of the National Coal Association,
with laboratories in Monroeville, Penn-
sylvania, but currently attention is focused
on work now being done at the prototype
plant built by Monsanto Company and Metro-
politan Edison Company at the latter's
Portland, Pennsylvania, station along the
Delaware River.
The process consists of oxidizing catalyti-
cally the sulfur dioxide found in flue
gases to sulfur trioxide, which is con-
densed to sulfuric acid and can be used
commercially.
In 1962 the Pennsylvania Electric Company,
Air Preheater Company, Research Cottrell
Corporation, and Monsanto participated in
the construction and operation of a small
pilot plant at Penelec's Seward, Pennsylvania,
generating station.
The project proved the technical feasibility
of the process, but a number of questions are
left unanswered. Monsanto and Metropolitan
Edison decided to build and operate a pro-
totype plant at Portland to obtain economic
and engineering information necessary for
the design of commercial units. Construction
was begun in 1966 and completed in August 1967.
The prototype plant takes only a portion of
the total flue gas from the generating station,
but it incorporates commercial type equipment
about the size used on many industrial boilers.
The development program required an expenditure
of between ?4 and ?5 million.
8-20
-------
Monsanto believes results are so encouraging
that the process can be made available commer-
cially in the near future.
In the process, flue gases direct from the
boiler are first passed through mechanical
and electrostatic collectors to remove in
excess of 99.9 percent of the particulates.
This efficiency is required to prevent foul-
ing of the catalyst. The gases are then
passed through a converter where more than
90 percent of the SO is oxidized to SO
with a vanadium pentoxide catalyst.
The SO rich stack gas then enters a rotary
heat exchanger where incoming combustion
air is used to cool the gas from about
900 degrees F to 200 degrees F. During
cooling the SO combines and condenses with
the moisture normally present in the stack
gas and forms 70 percent sulfuric acid.
Most of the acid is collected on the cool
surfaces of the condenser. Acid mist re-
maining in the gas is collected in a second
electrostatic precipitator. A typical
effluent for this process would contain
200 ppm SO by volume and 20 ppm SO .
Figure 8-1 shows the Catalytic Oxidation
SO recovery process.
8.2.1.2 Beckwell (Wellman-Lord) (12, 13)
A pilot installation, owned and
operated by Wellman-Lord, Inc., of Lakeland,
Florida, performed satisfactorily for several
months at one of the Tampa Electric Company's
Gannon Plant units. It is reported to remove
more than 90 percent of the sulfur oxide and
all of the flyash that remains after the gases
pass a precipitator.
8-21
-------
AIR
t
TO STACK
200° F
900° F
00
I
tO
IO
t
ELECTROSTATIC
DUST
COLLECTOR
900° F
t. C<
CATALYTIC
REACTOR
FLYASH
AIR
HEATER
ACID
CONDENSER
ELECTROSTATIC
MIST
COLLECTOR
FURNACE.
FIGURE 8-1
CATALYTIC OXIDATION
S02 RECOVERY PROCESS
-------
The process involves a reactor for scrub-
bing SO from the flue gas with an aqueous
solution of potassium sulfite. The SO reacts
to form potassium bisulfite. About 99 per-
cent of the flyash and SO are removed, along
with 90 to 95 percent of the SO . This portion
of the pilot plant operated continuously with
little difficulty and minimal pressure drop.
The flue gases are discharged from the reactor
to the stack at 140 degrees F with a velocity
of 70 to 90 feet per second. The flue gas
is 70 percent saturated and plumes will not
be visible until the temperature drops below
45 degrees F. No need for stack reheat is
predicted.
The reacted scrubbing liquor is pumped to
a special treating area where the potassium
bisulfite is crystallized out of solution
as potassium pyrosulfite, is pumped to
the top of the stripping column in the recovery
area, and flows countercurrent to steam in-
jected in the bottom of the column. The
potassium pyrosulfite in the heated solution
reverts to potassium sulfite and releases SO .
Approximately four pounds of steam reportedly
are required per pound of SO produced. More
recent data indicate that this figure may be
low. The potassium sulfite is recycled to the
reactor scrubber.
Steam from the column is condensed and recycled
from the K.O. drum to the special treatment area.
The SO vapor from the K.O. drum is compressed
and fed to a distillation column and is con-
densed overhead with water. A portion of the
liquid SO is refluxed to the column and the
remainder fed to storage.
8-23
-------
A demonstration plant was built at the
Crane Station of the Baltimore Gas and
Electric Company to recover the SO re-
leased from 25,000 kw of power generation,
Results of this demonstration should be
available in early 1970. Figure8-2.
shows the Weliman-Lord,' Inc. SOij-:recovery
process. "••
8.2.1.3 Stone and Webster Process (19,,' 15)
Stone and Webster Corporation and Ionics,
Inc., have developed a process the basis of
which is an electrolytic cell which regen-
erates two necessary process fluids. The
process utilizes a high efficiency wet
scrubber of proven design and can produce
dry high purity sulfur dioxide. The process
is adaptable to existing boilers. This
process was piloted on flue gas from coal-
fired boilers at the F. T. Gannon station
of the Tampa Electric Company and is now
being planned for a demonstration plant.
Further process details are unavailable
at thi s time.
8.2.1.4 Reinluft (_3, 9^, 10, 15)
This process, developed in Germany, uses
activated char to absorb SO from the flue
gases, and can produce sulfuric acid. In
the process, the activated char is in a
fixed, slowly moving bed. Flue gas, at
a temperature above its dew point, enters
the bottom of the absorber, where the SO
is then drawn off, cooled to 220 degrees F and
returned to the absorber at a higher level.
8-24
-------
TO STACK
03
N)
Ul
SOLUTION
REACTOR
FURNACE
FLYASH
AND S03
I
SPECIAL
TREATMENT
AREA
TO S02
COMPRESSOR
STRIPPING
COLUMN
STEAM
FIGURE 8-2
WELLMAN-LORD
SO, RECOVERY PROCESS
-------
Sulfur dioxide in the gas is oxidized to
SO and adsorbed as sulfuric acid along
with moisture in the char. The char, with
its adsorbed sulfuric acid, drops to the
regenerating section where at 700 degrees
F the sulfuric acid disassociates. The
resulting SO is reduced to SO . The SO
gas leaving the generator at 300 degrees F
is heated to 700 degrees F and returned to
the base of the generator. A side stream
removes the SO product gas.
Although the process seemed almost ready
for full-scale commercial application,
technical difficulties developed concerning
the high consumption rate and ignition of
activated char, and its applicability has
become questionable. Reinluft, Inc. has
recently signed over full rights to the
process to Chemieban- Zieren.
One reference (16) states that three full-
scale Reinluft plants either are in oper-
ation or about to be completed in West
Germany. A smaller plant has been operating
at the Volkswagen works for several years
and Volkswagen management is installing the
first of the three aforementioned plants in
expectation of reducing their pickling
acid purchases. Figure 8-3 shows the
Reinluft, Inc. process.
8.2.1.5 Alkaline Scrubbing
A system reported by Galeano and Harding (11)
uses soda ash liquor to scrub SO from the
power boiler flue gases, resulting in a liqui<
suitable for use in NSSC cooking liquor.
8-26
-------
TO STACK
FURNACE
300° F
300° F
GAS
HEATER
700° F
t k 215° F
i
ex.
LU
02
CC
o
oo
o:
o
CD
220° F
GAS
290° F
s
-R
_r\
LU
_j
o
o
eg
LU
CO
ce
0
to
o
-------
The efficiency of base scrubbing for removing
SO from boiler stack gases generally
depends upon the concentration of the
base, liquid/gas ratio, temperatures,
pressure drop, and the particular scrub-
ber. Operating the system for most
efficient removal of SO produces a
liquor volume exceeding the expected
daily makeup requirements (9) . To match
the liquor produced to that needed, the
concentration and liquid/gas ratio are
the most easily changed variables.
Sulfur dioxide can be removed from a
gas stream by putting it in contact with
a solution of any base. Bases suitable for
sulfite pulping include, Ca(OH) , NaOH,
NH OH, and Mg(OH) . A solution of any of
these can be used to effectively scrub
SO from the gases and produce a bisulfite
solution. This weak acid can be incorpor-
ated into the acid production system of
the sulfite mills, for a net savings in
elemental sulfur requirements.
It is possible that alkaline scrubbing could
cause some conversion of sulfite and bisulfite to
to sulfate. The result of this would be the
introduction of an inert impurity into the
pulping liquor.
8.2.1.6 Kiyoura - T.I.T. (12)
This process uses catalytic oxidation followed
by a gaseous ammonia reaction to produce
ammonium sulfate. It is similar to the catalytic
oxidation (Monsanto-Metropolitan Edison) method
in that the hot flue gases from the boiler are cleaned
of essentially all particulates before passing
through a converter where the SO is oxidized
to SO with a vanadium pentoxide catalyst.
Gaseous ammonia is added after the catalytic
converter, when the stack gas, now rich in SO
has been cooled to 425 degrees F to 500 degrees
F. At these temperatures, the SO will not
8-28
-------
have condensed, and ammonium sulfate is
formed. The ammonium sulfate is collected
in a second electrostatic precipitator.
The sulfate crystals are extremely pure
and can be used directly as a fertilizer
material. More than 70 percent of the
flue gas SO is recovered (2). Figure
8-4 shows the Kiyoura process.
8-29
-------
TABLE R-ll
SUHMARY OF SULFUR OXIDE RECOVERY PROCESSES
Process
Name
Catalytic
Oxidation
Type of
Process
Uses V 0
Cat. 0
Oxidize SO,
to SO_ "
Form of
Recovered
Sulfur
V°4
State of
Development
Offered
Commercially
Possible
Pulp Process
Application
Kraft
Remarks
Requires high degree
of particulate removal
from gas (99.9 percent)
co
Beckwell
(Wellman-
Lord)
Scrub SO
into Pot.
Sulfite
Decompose Pot Pyro
sulfite to rec.
SC-
SO,
Demonstration
Plant built
Kraft
NSSC
Sulfite
Stone &
Webster
Process
Scrubbing into
process fluids
electrolytically
regenerate
solutions
SO,
Demonstration
plant planned
Kraft
NSSC
Sulfite
Reinluft
Alkaline
Scrubbing
Adsorb SO
on activated
char, oxidize
to SO and desorb
Scrub SO
from flue gas
with alkaline
solution
H2S°4
Sulfite
Salts
Conflicting
reports from
technical diffi-
culties to full
plant operation
Pilot Plant
Kraft
Char fires common
Kraft
NSSC
Sulfite
Use in pulping process
depends on scrubbing
medium chosen
Kiyoura
T.I.T.
Special case of
Cat. oxidation
so"4'2
S°4
Pilot Plant
None
Produces salable
-------
TO STACK
AMMONIA
CD
I
W
DUST
COLLECTOR
800° F
CATALYTIC
REACTOR
750° F
AIR
HEATER
ELECTROSTATIC
DUST
COLLECTOR
450° F
AIR
HEATER
AMMONIUM SULFATE
FURNACE
FIGURE 8-4
CATALYTIC OXIDATION - AMMONIA
S02 RECOVERY PROCESS
-------
8.3 PROCESS FEASIBILITY CONSIDERATIONS
There are many schemes, either proposed or being
developed, for the removal or recovery of sulfur
oxides from flue gas from utility power boilers.
Progress in developing suitable flue gas desulfurization
processes has been slow because of the magnitude and
complexity of the problem. Flue gas desulfurization
process development is further complicated by a wide
variation in the size of utility plants.
The technical and economic feasibility of most processes
is closely related to plant size (15).
It is unlikely that a single flue gas desulfurization
method will be developed that is capable of controlling
effluents from all types of sources. The recovery
technique employed will depend on factors such as boiler
size and configuration, age, load pattern, fuel character-
istics, and utility of by-product sulfur or sulfur com-
pounds .
In evaluating the feasibility of the several processes
outlined in order to select the most attractive, the
controlling criteria are technical feasibility and
economic considerations.
The major problem associated with the evaluation of the
processes is the fact that most of the systems have not
been developed commercially and projected cost data reported
by the developers should not be taken at face value. Cost
analyses based on published data seldom provide sufficient
detail with which the reported economics can be verified.
Letters were written to the developers of the several
processes under consideration requesting the most recent
and detailed capital and operating costs that they would
feel free to furnish. No replies were received from any
of these inquiries therefore evaluation and comparison of
the systems must be based on the limited data which appear
in publications.
8-32
-------
In addition to the lack of dependable economic
data, process details necessary to reasonably
estimate capital and operating costs are un-
available, presumably, because of the proprietary
nature of the information. Here also, we must
rely on published articles as a gauge of the
technical feasibility of the systems.
Much of the published data, both economic and
process, is repetitive and articles published
by the developers tend to be more optimistic
about the merits of their process than do the
articles published by independent writers.
A recent article (17) classifies some of the
various sulfur oxide recovery processes by the
following criteria:
1. First Generation
a. Advanced pilot plant studies
b. Active research and development
c. Adaptability to U. S. market
d. Available data on economic assessment
2. Near First Generation
a. Meeting first generation criteria
except:
1. Fewer advanced pilot studies or
insufficient information
2. Better adaptability to foreign
operations
3. Second Generation
a. Essentially bench studies of new concepts
b. Pilot studies incomplete
c. Economic assessment not practical
8-33
-------
The considerations used in the classification
were:
1= Technical feasibility
2. State of development
3. Process applicability
4. Economic factors
5. Relative advantages
6, Problem areas.
Using this system of ranking, the six processes
discussed in 8.2.1 may be classified as follows:
1. First Generation
a. Catalytic Oxidation
b. Reinluft
2. Near First Generation
a. Beckwell (Wellman-Lord)
b. Kiyoura - T.I.T.
c. Alkaline Scrubbing
3. Second Generation
a. Stone and Webster - Ionics, Inc.
The determination of the most feasible process
for the recovery of sulfur oxides from power boiler
flue gases depends on the following factors:
1. Utility of recovered product
2. Technical feasibility
3= State of the development of the process
4. Economics of the process
8-34
-------
In Section 8.2.1, the number of processes appearing
attractive was found to be six, based on the cri-
terion of recovered product utility in one or more
pulping processes.
In this section, based on the criteria of technical
feasibility and the state of development, the num-
ber of processes to be considered as attractive has
been reduced to two which have been listed as First
Generation Processes. These processes yield
recovered products suitable only to the kraft pulp-
ing process.
8.3.1 Sulfur Recovery and Reuse - Economic Evaluation
The literature was reviewed in an effort to determine
the parameters necessary to make an economic evalua-
tion of the processes under consideration. All
available publications appear to be based on public
utility stations of 800 MW size (2.7 x 109Btu/hr) and
new plant installations. Of the information gleaned
from the literature sources, there is a similarity in
capital and operating costs which implies that the
data may have been extracted from a limited number of
original estimates. This situation is created by the
lack of economic data sources available to independent
authors. The economic data which are available, in
general, come from the developer of the particular
process, and such economic data are sparse.
This study is specifically oriented to the power
boilers in the pulp industry and few, if any, of these
are beyond a size of 100-200 MW equivalent. (See
Appendix B for distribution of boiler sizes in the
pulp industry.) It becomes apparent, then, that any
published capital cost will suffer in accuracy by
reducing from an 800 MW to a 100-200 MW basis since
the relationship of capital cost to unit size is not
known. It is felt that the application of the
standard "six-tenths power" factor normally used for
scale-up effect on capital cost should not be used
because of the uncertainties of the published capital
cost data and the lack of information on the equip-
ment details of the various processes. This is further
complicated by the fact that all data have been projected
from relatively small pilot plant installations. The
8-35
-------
literature gives capital costs for 800 MW installations
in the range of $10 - 20 million or $10 - 22 per kilowatt.
Even with a reasonable reduction in capital cost attribu-
table to scaling down from 800 to 100 - 200 MW, the
investment for these processes would still be substantial
and could approach approximately 10 percent of a new
800 - 900 TPD mill investment. It is felt, however, that
the chances of a substantial reduction in investment are
marginal because all of the investment figures shown are
based on new plant installations and there are several
comments in the literature pertaining to the difficulty or
impracticability of putting these sulfur recovery processes
into existing power plant installations due to lack of
available space „
In existing installations , the generation of space sufficient
to accommodate the sulfur recovery process units would
insure a substantial cost which could conceivably approach
the cost of the unit installation. This situation is par-
ticularly prevalent in many existing pulp mills.
In view of the above, it is obvious that under any circum-
stances, the installation of the sulfur recovery processes
would result in a very substantial capital cost.
Operating cost data for the three processes are subject
to the same limitations as are the capital cost data in that
the information appears to be repetitive and probably
from the same parent sources. The operating costs stated
in the literature sources show a new operating cost ranging
from about $0.75 to $1.50 per ton of coal and $0.40 to
$0.50 per barrel of oil, including credits for the sulfur
by-products recovered.
It is felt that these operating costs must be considered
as minimum costs because the factors included in operating
costs will not be reduced proportionately to plant size
and, in fact, the operating costs may increase per unit of
fuel as the plant size is reduced,, Also, it should be
noted that any credits included above are based on sulfur
prices in the range of $35 per long ton- Recent publications
(23) have indicated that this sulfur price may no longer be
8-36
-------
accurate and, in fact, may be substantially higher
than future sulfur prices. Also, these credits do
not include any allowances for costs to supply
sodium to correct any imbalance created in the
sodium-sulfur ratio.
In view of the fact that these sulfur recovery
processes are not advanced beyond the pilot plant
stages, and, in addition, with the high capital and
operating costs predicted, it must be concluded that
the processes will not be feasible for the control
of sulfur oxide emissions from utility boiler flue
gases in the foreseeable future.
The state of development of the non-recovery processes
is no further advanced than that of the recovery
processes. These processes also require additional
development at the utility boiler scale before they
can be applied to pulp mills.
Thus, it appears that the only feasible alternate for
immediate application available for the control of
sulfur oxide emissions from utility boiler flue gases
for the pulp industry must be based on fuel substitution.
8-37
-------
8.4 R & D EFFORTS
The fact that the sulfur removal processes have
not been developed to a feasible stage does not
preclude the interest in continuing the research
and development efforts in these and similar
sulfur removal processes as methods of control
of sulfur oxide emissions from industrial power
plant installations utilizing fossil fuels.
It is estimated that as many as eighty to one
hundred separate processes or modifications are
being investigated by many agencies. Currently,
most of this work is being directed toward
emissions control from large power generating
stations (800 MW). It is suggested that these efforts
should be continued and supported with additional
emphasis being placed on the smaller power generating
stations found in many industrial plants as auxiliary
operations. In this continuing R S D program, we
should not lose sight of the various fuel desulfurization
programs which are being intensively studied. Some
support is justified in this area as well as flue
gas treatment.
Another area of needed research is to search for
methods to protect the equipment used in these sulfur
recovery processes against corrosion. This considera-
tion is a continual one which affects both process and
emission control equipment.
Work must also be done to devise ways to prevent process
contamination by carbon when sulfur recovery processes
are considered. It has been found that carbon, which is
removed in particulate form from the flue gas, passes
through the other process steps to show up at the paper
machines. This quite obviously impairs the quality of
paper being produced. Sophisticated filtration processes
must be used to remove the carbon and these should be
researched further.
8-38
-------
8.5 REFERENCES
1. Libby, C. E., Pulp and Paper Science and Technology,
_!, Pulp, (1962) .
2. Bovier, Ralph F., "Sulfur-Smoke Removal Systems," Proceedings
of the American Power Conference, 26-, 135-143, (1964).
3. Moody, J. E., "Research Must Meet the Demand," Combustion,
(May 1968).
4. Galeano, S. F., "Removal and Recovery of SO in the Pulp Mill
Industry/" Dissertation, University of Florida (August 1966).
5. Harding, C. I. and Galeano, S. F., "Utilization of Weak
Black Liquor for SO Removal and Recovery," TAPPI 50, (10),
(October 1968). X
6. Shah, I. S. and Stephenson, W. D., "Weak Black Liquor
Oxidation System: Its Operation and Performance," TAPPI
_51, (9) , (September 1968) .
7. Murray, F. E., "Oxidation of Kraft Black Liquor," in
Atmospheric Emissions from Sulfate Pulping, (E. R.
Hendrickson, Ed.), (April 1966).
8. Aires, R. S. and Newton, R. D., Chemical Engineering Cost
Estimation, McGraw-Hill Book Company, Inc. (1955)."""
9. Kaiser, Alex, "Air Pollution Control in the Power Industry,"
presented at Annual Meeting, Florida Section APCA, October
7, 1968.
10. "Outside U.S., Tough Laws Spur SO Removal Efforts," Chemical
Engineering, 84-88, (November 4, 1968).
11. Galeano, S. F. and Harding, C. I., "Sulfur Dioxide Removal
and Recovery from Pulp Mill Power Plants," Jour. APCA, 17^, (8) ,
(August 1967).
12. Terrana, J. D. and Miller, L. A., "New Process for
Recovery of SO from Stack Gases," presented at American
Power Conference, Chicago, April 1968.
8-39
-------
13. "Profit in Stack Gas?", Chemical Week, 53-54, (July 20, 1968).
14. Galeano, S. F. and Amsden, C. D., "Weak Black Liquor
Oxidation With Molecular Oxygen," presented at 62nd
Annual Meeting of the Air Pollution Control Association,
New York, June 22, 1969.
15. "Control Techniques for Sulfur Oxide Air Pollutants,"
U. S. Department of Health, Education and Welfare,
NAPCA Publication No. AP-52, (January 1969).
16. Frankenberg, T. T., "Sulfur Removal: For Air Pollution
Control." Mechanical Engineering, (August, 1965).
17. Cortelzon, C. G., "Commercial Processes for SO Removal,"
Chemical Engineering Process, (September 1969).
18. Persh, E. A., Rusanowsky, N. P., Young, N. W., "An Appraisal
of Air Pollution in the Power Industry," presented at
American Power Conference, Chicago, April 1968.
19. "Research Goal: Solution of the Sulfur Problem," Coal Research,
Bituminous Coal Research, Inc.,, Vol. 25, (Winter 1968).
20. Reid, W. T., "Sulfur Oxides Control in Central Station Power
Plants," Heating, Piping and Air Conditioning, 153, (March
1968).
21. Kiyoura, R., "Studies on the Removal of Sulfur Dioxide from
Hot Flue Gases as a Measure to Prevent Air Pollution," presented
at 59th Annual Meeting of the Air Pollution Control Association,
San Francisco, June 20-24, 1966.
22. Katell, S., "Removing Sulfur Dioxide from Flue Gases,"
Chemical Engineering Progress, 62, (10), (1966) .
23. "Market Newsletter," Chemical Week, 105, (13), 43-44,
(September 27, 1969).
8-40
-------
APPENDIX B
Appendix B consists of two tables which show the distri-
bution of steam generating capacity by on-site power boilers
in the chemical wood pulping industry.
Table B-l summarizes the total quantities of steam generated
by the industry.
Table B-2 presents information on the steam generating
capacity of individual power boilers in the industry.
B-l
-------
TABLE B-l
DISTRIBUTION OF TOTAL QUANTITIES OF
STEAM GENERATED BY POWER BOILERS AT
289 PULP AND/OR PAPER MILLS IN THE U.S.
Total Annual Average Steam Generation
1,000's
0 -
21 -
51 -
101 -
201 -
501 -
1,000 -
> -
IbsAr
20
50
100
200
500
1,000
1,500
1,500
MW Equivalent
0-1.8
1.8-4.5
4.5-9.0
9.0-18.0
18.0-45.0
45.0-90.0
90.0-135.0
> 135.0
Number
of Mills
24
48
47
53
75
35
7
0
Percent of
Total
8.5
17
16
18
26
12
2.5
0
289 100
Source: NCASI - NAPCA questionnaire survey of steam and power boiler
in the pulp and paper industry (1969).
B-2
-------
TABLE B-2
DISTRIBUTION OF STEAM GENERATING CAPACITIES OF
INDIVIDUAL POWER BOILERS AT
282 PULP AND/OR PAPER MILLS IN THE U.S.
Steam Generating Capacity
1,000's
0 -
21 -
41 -
61 -
81 -
101 -
151 -
201 -
301 -
401 -
501 -
601 -
701 -
Ibs/hr
20
40
60
80
100
150
200
300
400
500
600
700
800
MW Equivalent
0-1.8
1.8-3.6
3.6-5.4
5.4-7.2
7.2-9.0
9.0-13.5
13.5-18.0
18.0-27.0
27.0-36.0
36.0-45.0
45.0-54.0
54.0-63.0
63.0-72.0
Number of
Boilers
155
187
159
119
82
111
67
66
20
15
7
0
1
989
Percent of
Total
16
19
16
12
8.2
11
6.8
6.7
2.0
1.5
0.7
0
0.1
100.0
Source: NCASI - NAPCA questionnaire survey of steam and power boilers
in the pulp and paper industry (1969)
B-3
------- |