CONTROL OF SALT-LADEN PARTICULATE

   EMISSIONS FROM HOGGED FUEL BOILERS
                   by

      Michael Szabo, Samir Kothari,
   Craig Doolittle, and Richard Gerstle
        PEDCo Environmental, Inc.
            11499 Chester Rd.
         Cincinnati, Ohio  45246
         Contract No. 68-01-4147
               Task No. 51
 EPA Task Manager:  Michael M. Johnston
             US EPA Region X
            1200 Sixth Avenue
       Seattle, Washington  98101
              Prepared for

  U.S. ENVIRONMENTAL PROTECTION AGENCY
Division of Stationary Source Enforcement
         Washington, D.C.  20460
                May 1979

-------
                           DISCLAIMER
     This report was prepared for the United States Environmental
Protection Agency, Region X by PEDCo Environmental, Inc.,  in
fulfillment of contract No. 68-01-4147,  Task Order No.  51.   The
contents of this report are reproduced herein as received  from
the contractor.   The opinions, findings,  and conclusions expressed
are those of the authors and not necessarily those of the
Environmental Protection Agency.

-------
                            ABSTRACT

     This report presents results of an evaluation of salt emis-
sions from hogged fuel boilers, and defines the technical
problems, compliance prospects, and costs of various control
systems.  The control measures considered are fuel pretreatment,
combustion modifications, use of conventional control devices
(electrostatic precipitators, fabric filters, and scrubbers), and
several novel particulate control devices.
                                111

-------
                            CONTENTS
Disclaimer                                                   _ 1;L
Abstract                                                     ii:L
Figures                                                        v
Tables                                                       vii
Acknowledgement                                               !x
Summary and Recommendations                                    x

     1.    Introduction                                         1
               Scope of this study                             3
     2.    Definition of the Salt Emissions  Problem             5
               Current usage and geographic distribution
                 of hogged fuel boilers                        5
               Characteristics of hogged fuel                   9
               Fuel storage and pretreatment                  13
               Combustion of hogged fuel                      20
               Particle size distribution of salt
                 emissions                                    25
               Effects of salt-laden hogged fuel boiler
                 emissions on ambient air quality             28
               Effects of salt-laden particulate on
                 corrosion rates
     3.    Control Technology to Reduce Salt Emissions
            from Hogged Fuel Boilers                          42
               Fuel handling and pretreatment                  42
               Combustion modifications                       46
               Conventional particulate control devices       49
               Novel fine particulate control  devices         83
     4 .    Cost Assessment of Control Technology for
            Salt-laden Particulate Emissions from
            Hogged Fuel Boilers                               91
               Comparison of estimated and  actual costs       91
               Financial hardship                             93

Appendices

     A.    Salt-laden Hogged Fuel Boilers in Washington,
            Oregon, and Alaska                                97
     B.    Types of Hogged Fuel Burning Furnaces              101
     C.   " Case Histories of Secondary Collectors for
            Control of Salt-laden Particulate from Hogged
            Fuel Boilers                                     119
     D.    Performance of Novel Control Devices Applicable
            to Salt Particulate Emissions                    126

                               iv

-------
                             FIGURES
Number                                                      Page

 2-1      Some methods of energy conversion using direct
            combustion of residue materials                   6

 2-2      Number of wood-fired boilers by state               8

 2-3      Storage of logs by flat raft and bundle            16

 2-4      Log storage time in Oakland Bay for Simpson
            Timber Company                                   16

 2-5      Cross-section of a typical hogging machine         17

 2-6      Effect of fuel moisture on steam production        19

 2-7      Inclined-grate stoker furnace at a paper and
            pulp mill in British Columbia                    24

 2-8      Bark-fired cyclone type furnace                    26

 2-9      Particle size distribution at Weyerhauser Co.,
            North Bend, Oregon, plant                        27

 2-10     Suspended particulate data, Shelton, Washington    30

 2-11     Suspended particulate data, Port Townsend,
            Washington                                       34

 2-12     Suspended particulate data, Coos Bay, Oregon       37

 3-1      Relation of time in saltwater to absorption of
            water and salt                                   43

 3-2      The effect of fuel moisture on steam production
            as reported by Johnson                           45

 3-3     .Increase in apparent density of smoke with
            increasing chimney diameter                      48

 3-4      Simplified diagram of a multiple cyclone           50
                               v

-------
                       FIGURES (Continued)


Number                                                      Page

 3-5      Cyclone collector for particles in flue gases      51

 3-6      Relation of particle size to collection effi-
            ciency of cyclones                               54

 3-7      Typical electrostatic precipitator with top
            housing (courtesy of Research Cottrell,  Inc.)    59

 3-8      Research-Cottrell flooded disc venturi scrubber    65

 3-9      Reverse air or shaker type                         75

 3-10     Pulse jet type                                     76
                               VI

-------
                             TABLES
Number                                                      gage

 1-1      Hogged Fuel Boilers with Excessive Salt
            Emissions                                         2

 2-1      Uses of Process Steam in Forest Product Manu-
            facturing Plants                                  7

 2-2      Analyses of Selected Wood Refuse Burned as  Fuel     10

 2-3      Major Salt Compounds in Seawater                   12

 2-4      Techniques for Measuring Salt Content of Fuel
            and Flue Gas from Hogged Fuel Boilers            14

 2-5      Advantages and Disadvantages of Three Major
            External Drying Systems                          21

 2-6      Particle Size Data from Hogged Fuel Boilers
            with Excessive Salt Emissions                    29

 2-7      Particulate Ambient Air Quality Standards          31

 2-8      Summary of Suspended Particulate Data, Shelton,
            Washington                                       32

 2-9      Summary of Total Suspended Particulate Data
            Obtained Near Crown Zellerbach Boiler            33

 2-10     Summary of Total Suspended Particulate Data,
            Coos Bay Sampling Station                        36

 3-1      Effects of Cyclone Dimensions on Performance
            and Cost                                         56

 3-2      Effects of Physical Properties and Process
            Variables on Efficiency                          57

 3-3     " Design Power Density                               62

 3-4      Design Parameters and Design Categories for
            Electrostatic Precipitators                      64
                               VII

-------
                       TABLES (Continued!
Number
                                                            Paqe
 3-5      Design Parameters for Baghouses on Hogged Fuel
            Boilers                                          71

 3-6      Characteristics of Various Baghouse Fabrics        73

 3-7      Performance of Conventional Secondary Collectors
            on Salt-ladenParticulate from Hogged Fuel
            Boilers                                          82

 3-8      Novel Fine Particulate Control  Devices Applied
            to Boilers Burning Salt-laden Hogged Fuel        86

 4-1      Capital and Annual Operating Cost Estimates for
            Secondary Collectors at Weyerhauser Co. North
            Bend Plant                                       92

 4-2      Summary of Cost Information on  Secondary Col-
            lectors Applied to Salt-laden Hogged Fuel
            Boilers                                          94
                               VI 11

-------
                         ACKNOWLEDGEMENT

     This  report was  prepared  for the U.S. Environmental Protec-
,ton Agency,  Region  X,  Seattle, Washington, by PEDCo Environmental
 Inc.,  Cincinnati, Ohio.
     The project was  directed  by Mr. Richard W. Gerstle, and
 managed by Mr.  Michael F.  Szabo.  Principal authors were Mr.
 Szabo, Mr. Samir P. Kothari, Mr. Craig C. Doolitle, and Mr.
 Gerstle.
     Mr. Michael M. Johnston was the task manager  for U.S. EPA,
 Region X,  and  the authors  appreciate his cooperation.
     We also thank  officials of the Crown Zellerbach Corp. and
 Simpson Timber Company,  who were very cooperative  and helpful
 during visits  to their plants  and in subsequent discussions;
 Mr. Dick Blanchard  of the  Olympic Air Pollution Control Authority,
 Olympia, Washington,  for his valuable assistance during these
 visits; the  State environmental control agencies in Washington,
 Oregon, and  Alaska; and  Dr. Richard W. Boubel, for providing
 valuable technical  data  on salt-laden hogged fuel  boilers.
                                IX

-------
                   SUMMARY AND RECOMMENDATIONS

     Although the data base is small, test operations have demon-
 strated that salt-laden participate emissions from hogged fuel*
 boilers can be controlled to comply with stringent particulate
 and opacity regulations by application of currently available
 control technology.  Whether this technology should be applied to
 each affected hogged fuel fired boiler should be determined on a
 case-by-case basis.  The following information summarizes salient
 aspects of this report and the conclusions drawn therefrom.
 Recommendations are given for further study of reducing salt
 emissions.

 SIGNIFICANCE OF THE SALT EMISSION PROBLEM
     Salt emissions from hogged fuel boilers are a significant
 problem, principally because the salt portion of the particulate
 is primarily submicron in size.  Particles larger than 5 ym are
 deposited in the nasal cavity or naso-pharnyx.  The smaller
 particles, however, are deposited in the lungs, including over 50
 percent of the particles from 0.01 to 0.1 ym diameter that
 penetrate the pulmonary compartment.  The tendency of particu-
 lates to penetrate the respiratory systems and be captured is
mainly a function of their geometry rather than their chemical
properties.
     Health effects of the captured fine particles depend largely
on their chemical or toxic qualities, except for long fibrous
materials,  whose physical qualities also provide potential for
  A term derived from the machine that processes the mixture of
  wood and bark before it is burned and that is called "a hog"
  or "hogger".

-------
irritation of tissue.  Because of the many unknown factors, it is
unwise to generalize concerning health effects of fine salt
particulates from hogged fuel boilers.
     The other important aspect of the problem is that salt-laden
particulate emissions, after passing through a conventional
mechanical collector, are usually in violation of both particu-
late emission and opacity regulations.  A plume from these opera-
tions is often highly visible and aesthetically objectionable in
the community.

DEFINITION OF THE SALT EMISSION PROBLEM
     The salt concentration in hogged fuel varies widely, and the
salt content of wood fuel or flue gas can be expressed in several
ways:  as NaCl, equivalent chloride, or total sea salt.  No
standard method of analysis is in use, and the method of analysis
often is not reported.  The need is apparent for a consistent
method of collecting, storing, and analyzing salt samples from
wood fuel or flue gas.
     To evaluate the effects of salt-laden particulate emissions
on local ambient air quality, we analyzed data from three sites.
At none of them were there violations of ambient air regulations
for suspended particulates, during the period when mechanical
collectors were used for particulate control.  Secondary collect-
ors have been installed at two of the sites, but data are not
sufficient to indicate whether these installations have led to a
measurable decrease in ambient particulate levels.
     The absence of ambient air violations on a monthly or yearly
basis does not mean that ambient particulate levels are not in-
creased on a short-term basis.  With low winds and inversion
conditions, the plume from a salt-laden hogged fuel boiler equipped
only with mechanical collector could significantly increase
ambient salt levels over a short term.  However, just one of the
three sites analyzed in this study has shown violations of the 24
hour maximum average standard, and that was only in 1970.
                               XI

-------
     An increase in corrosion due to salt emissions would be
possible if the salt aerosol were present in sufficient quanti-
ties to deposit on surfaces.  Tsang and Stubbs, however, found no
deleterious corrosive effects in a 2-year period downwind of a
boiler fired with salt-laden hogged fuel in a coastal environ-
ment .

CONTROL TECHNOLOGY TO REDUCE SALT EMISSIONS
Handling and Pretreatment of Fuel
     Among various preventive methods for reducing salt particu-
late emissions from hogged fuel boilers, the most obvious is not
to transport the logs via saltwater; in most cases this is not
possible.  Transport by flat rafting rather than in bundles will
reduce the salt content of hogged fuel, as will reducing the
duration of storage in salt water.  About half of all salt ab-
sorbed in 6 months is absorbed in the first 2 or 3 weeks of
contact with seawater.
     Hydraulic debarking of logs with fresh water is reported to
have reduced salt content in one instance but not in another.
     Bark pressing can reduce moisture content by 50 percent and
remove substantial quantities of salt in the process.  This can
result in lower opacity and particulate emissions because of
increased boiler efficiency and fewer fine particles of salt.
The disadvantage, however, is serious water pollution from the
bark pressate wastewater.
     Other methods of predrying fuel do not reduce the salt con-
tent,  but do reduce combustible emissions by improving boiler
efficiency.
Combustion Modifications
     Combustion modifications can increase boiler efficiency and
thereby reduce total particulate emissions, but do not reduce
salt emissions because salt is not combustible.
                               xn

-------
     About 65 to 85 percent of the salt in bark passes out of the
chimney as particulate emissions; the remainder is retained in
the boiler, on the grates, or on boiler surfaces, reducing
efficiency of the boiler and disrupting uniform air flow through
the fuel beds.  Plant engineers therefore should try to reduce
both the salt and moisture content of the bark at reasonable
cost.  The resulting improvement in fuel quality will reduce
consumption of auxiliary fuel and boiler cleaning requirements.
Additional research is needed into ways to slow the quenching
effect on vaporized salt leaving the combustion zone in hogged
fuel boilers.  This would generate large particles of salt that
would be more easily collected by a secondary control device.  In
most operations, however, additional control measures will be
necessary to reduce salt-laden particulate emissions to accept-
able levels.
Conventional Secondary Collectors
     Secondary conventional collectors operating on salt-laden
emissions are fabric filters and wet scrubbers.  Electrostatic
precipitators  (ESP's) are not used because the manufacturer's
guarantee will cover outlet emission levels and in-stack opacity,
,but not visible opacity.  The dry scrubber has not been success-
ful on a full-scale boiler burning salt-laden hogged fuel.
     Performance data show that the best available control system
is a mechanical collector followed by a fabric filter.  This is
as expected, in view of the variation in salt content of wood
fuels and the fine-particle collection capability of the fabric
filter.  Operation of a fabric filter is not affected by changes
in salt content to the same extent as wet scrubbers and presum-
ably ESP's, which should be designed for the worst-case condi-
tion.  For the scrubber, the design must provide for a higher
pressure drop to account for increases in fines when the salt
content of the wood fuel increases.  This then entails higher
operating costs due to power consumption and component wear.   For
the ESP, the design must provide enough collection plate area  to

                               xiii

-------
compensate for the reduction in corona current at a given voltage
because of space charge phenomena caused by high fines content.
This disadvantage is compensated for somewhat by the low resisti-
vity (and unburned carbon)  of the bark ash, which causes higher
power inputs;  however,  the  low resistivity also can cause severe
reentrainment problems.  The cost of additional collection plate
area may make the ESP noncompetitive with the fabric filter.
     Although no fractional efficiency data are available from
the presently operating fabric filter and wet scrubber installa-
tions,  the low opacity (<5%)  and low outlet loading (0.02 g/m  or
0.01 gr/dscf)  of the Simpson Timber Co.  fabric filter indicate
that it is very effective in removing the salt fines.   The
scrubber installation at Crown Zellerbach's Port Townsend Mill
generates an estimated opacity of 35 percent and cannot meet the
emission regulation of 0.23 g/m  (0.10 gr/dscf), when the salt
content of the fuel is greater than 1 percent, because pressure
drop is limited to 51 cm (20 in.) of water.  Performance at these
two installations is consistent with test data on fabric filters
and wet scrubbers in other  applications.  ESP's have potential
for efficient removal of fines, but are less effective than
fabric filters.
     Performance of a pilot ESP on a boiler fired with salt-laden
hogged fuel at Victoria Sawmill Division of B.C. Forest Products,
Ltd., ranged from excellent to poor, with outlet loadings of 0.11
to 0.42 g/m   (0.05 to 0.20  gr/dscf).  Salt particles were removed
to an acceptable level as long as the ESP was maintained and
operated within precise limits; otherwise, performance deterio-
rated markedly.
     The disposal of salt-laden ash from secondary collectors  is
a problem sometimes overlooked in evaluating the salt emissions
problem.   Landfilling of the ash is complicated by the presence
of salt,  which presents a potential leachate problem.  Return  of
a slurry containing salt and ash to a bay or ocean is unaccept-
able because of the ash content, and the ash cannot be sold for
cement  or asphalt production because of the salt.
                              xiv

-------
     Acceptable disposal practices include providing an imperme-
able liner for an ash pond or diluting the salt concentration
by mixing the ash slurry with other plant wastewater streams or
with municipal wastewater and processing the ash slurry by con-
ventional wastewater treatment.  Wet treatment seems to be more
effective in terms of efficiency and economy.  Either method,
however, increases capital and operating costs.
Novel Devices
     Devices that combine the principles of an ESP and a wet
scrubber offer potential for effective control of salt-laden
particulate emissions from hogged fuel boilers.
     A  full scale test of one of these devices, the University of
Washington electrostatic scrubber, was recently performed on a
salt-laden hogged fuel boiler in the Pacific Northwest.  The
exact location is unknown and test details are sketchy, but
reports cite emission levels of 0.114 g/sdm  (0.05 gr/dscf) and
20 percent opacity.  Data from pilot tests of the Ceilcote
ionized wet scrubber and the A.P.S. electrotube show outlet
loadings of 0.09 g/m  (0.044 gr/dscf) and lower.
     Performance of the various types of electrostatic scrubbers
indicates that these units should be tested further in full-scale
applications.

COSTS
     Cost information on application of conventional control
devices to boilers burning salt-laden hogged fuel is minimal.
Weyerhauser Co. has developed cost estimates that allow compari-
son of a fabric filter and ESP at a level where the fabric filter
would comply with opacity regulations (40 percent), but the ESP
would not [0.42 g/m  (0.2 gr/dscf)].  For an installation with
throughput of 5042 m /min (180,000 acfm), the capital cost of the
fabric filter at $372 per m /min  ($11.72/acfm) is approximately
18 percent higher than that of an ESP.  Increasing the plate area
for the ESP to comply with the opacity regulation of 40 percent

                              xv

-------
would probably increase its cost over that of a fabric filter.
The capital cost for a comparable venturi scrubber is somewhat
less at $241 per m /min ($6.74/acfm).
     Estimates of annual costs for the same installations show
that the fabric filter would  be more costly to operate than the
ESP, mainly because of bag replacements.   Annual costs of operat-
ing the venturi scrubber are  still higher because of the high
power requirement of the fan.
     The actual capital costs  of the Simpson Timber Company's
fabric filter and Crown Zellerbach's venturi scrubber, when
escalated at 7.5 percent per  year, agree  well with the Weyerhauser
estimates, which are in 1978  dollars.   Actual operating costs for
Simpson Timber's fabric filter, however,  are considerably lower
than the Weyerhauser estimates, which do  not include costs of ash
disposal.

RECOMMENDATIONS FOR EVALUATION OF AFFECTED FACILITIES
     The following are recommendations for evaluation of the
boilers burning salt-laden hogged fuel in EPA Region X and cur-
rently violating applicable particulate emission and opacity
regulations.
     (1)  Analyze each plant  on an individual basis.  Contact the
affected companies to determine their plans for reducing salt
emissions to achieve compliance with regulations.
     (2)  If the company has  no plan for  compliance, visit the
site to evaluate the situation.  Perform  a detailed assessment of
plant fuel handling practices  and combustion techniques to
determine whether improvements could reduce emissions of salt and
ash.  Determine the amount of  space available at the site and
difficulty of retrofitting a  secondary control device.
     (3)  Estimate costs of preventive measures and of installing
secondary- collection equipment; develop a construction schedule
for installation of the equipment.  Concurrently, conduct an
economic feasibility study of  the affected company to determine
                               xvi

-------
whether it can afford to install the needed pollution control
equipment.
      4)  If the company is financially unable to install the
equipment, consider other alternatives such as closing the facil-
ity and evaluate the impacts of such alternatives on the employ-
ment situation in the local community.  If the plant is not
violating local ambient air quality standards and citizen com-
plaints are not numerous, investigate the possibility of a
variance or some other compromise, such as partial treatment of
the flue gas.
     If the affected facility is in an area where further degra-
dation of air quality is not permitted, approach other companies
wishing to establish plants that would cause additional local
pollution with respect to providing some or all of the money
needed to install pollution equipment on the affected hogged fuel
facility.  Such an approach could reduce both emissions and
ambient pollutant levels while encouraging industrial growth.
                               xvii

-------
                            SECTION 1
                          INTRODUCTION

     In some areas of the country, wood fuel that is fired in
industrial boilers comes from logs that have been transported or
stored in sea water, from which the bark may absorb substantial
amounts of salt.  When the fuel is burned, the noncombustible
salt particles, which are primarily submicron in size, contribute
to opacity and particulate emissions.  Most hogged fuel boilers
are controlled only with mechanical collectors, which are not
efficient in collecting fine particulate; the emissions therefore
usually violate state and local regulations of particulates and
opacity.
     Excessive salt emissions from hogged fuel boilers occur
almost exclusively in the coastal states of the Pacific Northwest.
There are 16 salt-emitting installations in Washington, Oregon,
and Alaska.  This study deals only with those boilers, which
together account for 4 percent of the total hogged fuel boilers
in these three states, as shown in Table 1-1.  Additional infor-
mation on these boilers is given in Appendix A.
     Depending on the salt fraction of the fuel and the type of
control device used, salt particles can constitute 30 to 90 per-
cent of stack emissions from these boilers.  Firing of oil con-
currently with the hogged fuel can also contribute to stack
opacity.  Although emissions from these boilers violate regula-
tions set forth in the State Implementation Plans  (SIP's) and
local regulations,  available ambient air data obtained near
several of these plants have not shown violations of ambient air
quality standards.
     In responding  to the salt emissions problem, some companies
have  cited potentially high costs and technical problems

-------
TABLE 1-1.   HOGGED FUEL BOILERS WITH  EXCESSIVE SALT EMISSIONS


Washington
Oregon
Alaska
Total
Total no. of
hogged fuel
boilers
98

No. of boilers
emitting salt
10
318 2
10
426
4
16

% of
total
10
0.6
40
4
Approximate
heat input
of boilers
emitting salt,
106 J/h (106 Btu/h)
1,050,700 (995)
362,210 (343)
785,660 (744)
2,198,570 (2082)

-------
 associated with control of these emissions with  secondary  control
 equipment.  The smaller companies, particularly, regard control
 of  salt emissions as a financial burden.

 SCOPE  OF THIS STUDY
     The purpose of this study is to examine the salt emissions
 problem, and to define the technical problems, compliance  pros-
 pects, and costs of various control techniques.
     Section 2 presents background information on the distribu-
 tion of hogged fuel boilers.  It describes some  characteristics
 of  hogged fuel, including the range of salt concentrations, and
 methods used to measure salt content.  It outlines methods for
 pretreatment and storage of hogged fuel, types of furnaces and
 combustion techniques, characteristics of the resulting flue gas,
 and the effects of salt particle size distribution on opacity.
     Section 3 reviews the control technology available for
 reducing salt emissions from hogged fuel boilers, considering
 both preventive measures, such as fuel pretreatment and combus-
 tion modifications, and remedial measures, such  as use of  con-
 ventional or novel particulate control equipment.  Basic design
 parameters are presented for each conventional control device,
 along  with evaluation of its applicability for control of  sub-
 micron salt particles.  Common operation and maintenance problems
 associated with each control device are reviewed, with emphasis
 on  those problems that are aggravated by the salt particles.
 Techniques for disposal of the salt-laden particulate are  also
 discussed.  Section 3 continues with a discussion of novel
 control devices and their potential for control  of salt emissions
 from hogged fuel boilers.
     Section 4  summarizes cost information available in the
 literature on conventional control devices, giving actual  and
estimated- capital  and annual costs.
     Appendix A lists the hogged fuel boilers in Washington,
Oregon, and  Alaska,  with excessive salt emissions.

-------
     Appendix B describes the types of furnaces in which hogged
fuel is burned.
     Appendix C presents two case histories of application of
secondary control devices to salt-laden particulate from hogged
fuel boilers.
     Appendix D describes some novel control devices that have
been tested on salt-laden particulate emissions from hogged fuel
boilers, giving test results; it also provides information on
commercial availability and cost of these devices.

-------
                            SECTION 2
            DEFINITION OF THE SALT EMISSIONS PROBLEM

CURRENT USAGE AND GEOGRAPHIC DISTRIBUTION OF HOGGED FUEL BOILERS
     The greatest use of wood residue and bark as fuel is by  the
industries that generate the wood wastes:   lumber and plywood
mills, paper mills, and particle board and hardboard mills.
These industries originally burned wood residue as a fuel out of
necessity.  Today these industries can use this relatively  low-
cost fuel to generate electricity and process steam.  In some
cases, they can generate a surplus of electricity for sale  to an
electric utility or for use in the electric system of the "com-
pany town."  In the Pacific Northwest and other areas, the  forest
products industry has been rapidly installing new wood-burning
boilers to replace those that burn oil and gas.
     Since oil prices have increased, wood fuel has become  so
desirable that wood products industries are saving it for their
own use rather than selling it on the open market.  One utility
in Oregon that uses wood wastes to generate electricity, the
Eugene Water and Electric Board, had to forego expansion because
local wood product industries, in a period of about a year,
completely reevaluated the wood fuel situation and chose to use
this fuel themselves rather than sell it.
     Figure 2-1 summarizes the basic ways of using wood fuels
directly to generate energy in the form of electricity, process
steam,  or hot gases.   The uses for process steam are summarized
in Table 2-1.

-------
      _RESIDUE_
      "MATERIAL
HOT GASES    T0 HEAT

         """ LOAD
           A.  HEAT UTILIZATION WITH COMBUSTION GASES
RESIDUE
MATERIAL
FURNACE
HOT GASES

BOILER
STEAM ^

                                                                 TO HEAT
                                                                  LOAD
           B.  HEAT UTILIZATION WITH STEAM
      RESIDUE
      MATERIAL
FURNACE
HOT GASES

BOILER

STEAM

STEAM
TURBINE
ELECTRIC
GENERATOR
ELECTRICITY TO ELECTRIC
LOAD
\ STEAM TO HEAT
LUAU
           C.  ELECTRICITY PRODUCTION WITH STEAM
     .RESIDUE_
      MATERIAL
FURNACE

HOT GASES

GAS
TURBINE
ELECTRIC
GENERATOR

ELECTRICITY ^JC
\ HOT EXHAUST ^_
GAS
ELECTRIC
LOAD
unit FR

STEAM TO HEAT
LOAD
           0.  ELECTRICITY PRODUCTION WITH A GAS TURBINE
Figure 2-1.   Some  methods  of  energy conversion,using direct  combustion
                                 of  residue materials.

-------
       TABLE 2-1.  USES OF PROCESS STEAM IN FOREST PRODUCT
                      MANUFACTURING PLANTS1
Type of plant
or operation
Use of steam from wood-fired boilers
                                    a
Dimension lumber
Plywood mill
Particle board
 and hardboard
Paper mill
Furniture
 manufacture
Kiln for drying lumber and
 "Shotgun carriage" (old, but still used)
Veneer dryer and hot press
Steam-heated particle dryer and
 hot press
Digester and paper machine dryer
Hot press and wood steaming system
alt is assumed that all facilities use wood fuel to supply heat
 and hot water for plant and offices.

     Hot flue gases can be used directly for drying of wood,
veneer, or particles.  The hot gas may be generated directly by a
wood-fired furnace without a boiler; or boiler flue gas can be
used instead of exhausting it through a stack.
     Because wood-fired boilers are traditionally located near
the fuel source, most are in the states with large forest pro-
ducts industries.  Figure 2-2 indicates the number of boilers and
weight of wood residue consumed in those boilers in each state.
Boubel  obtained the data for Figure 2-2 in a mail survey of
state air pollution control agencies.  For states not replying,
he estimated the number of boilers by a linear regression equa-
tion based on replies received and on wood usage as reported by
Suprenant.    in spite of discrepancies in the data, he believes
that they, are probably as reliable as any that can be obtained.
For example,  although reference 5 states that no wood is burned
industrially or commercially in Arizona or Michigan, the agency
in  Arizona  reported 14 wood-fired boilers and that in Michigan
listed  27.

-------
00
                     ^•-.^
                      (3771)  /  v

                     ••'
                                                                        *  ' >' y ST-• -

                                                                       ~X£f'^\"
                                                                      . -r-'  /7n\
                                                                     fcTNtuCKt (70)
                                                           '.       /    	T- f soHOL'^-
                                                           |   100  A—T"--»B*MVwHO.A\.   32

                                                                            102

                                                                                      UNITED STATES
             Figure 2-2.   Number of wood-fired boilers by state.   (Values in  parentheses

               indicate annual wood consumption in thousands of  tons;  1 ton  =  0.907 Mg.)

-------
     It is predicted that conversion of wood residue to energy
                                    2
will increase by 60 percent by 1985;  if this occurs, most of the
growth probably will occur in a few states having wood resources
that are not utilized today, such as Oregon, Washington, Idaho,
and California.  Although some of the other states may increase
the use of wood for fuel, they do not have enough unused resources
to double their usage in 10 years.

CHARACTERISTICS OF HOGGED FUEL
     The properties of wood residues and bark fuels can vary so
widely that no standard specification is possible.  The dif-
ferences should be recognized and accounted for in the engineer-
ing and operation of wood-fueled systems.
     Many species of wood can be used as fuel, but some are
better than others.  Wet cedar bark, for example, is stringy and
difficult to reduce in size.  By comparison, dry Douglas fir bark
is considered a very desirable fuel.   Table 2-2 summarizes the
analyses of several wood species used as fuel.
     Other types of wood fuel are used to supplement the hogged
fuel.  Among these are sawdust, wood chips and shavings, wood
waste material, and sanderdust.  Sludges and spent liquors from
wastewater treatment processes are burned occasionally, more as
a means of disposal than for any advantageous fuel character-
istics.  At some installations where downtime of the hogged fuel
boiler is critical to plant operations, the system incorporates
provisions for the burning of an alternative fuel, such as No. 6
fuel oil.
     The supplemental fuels are used in various quantities and
combinations depending largely upon the function and design of
the plant.   Except for sludge and spent liquors, the supplemental
fuels have  more desirable burning characteristics  (e.g., lower
moisture -content and higher heating value per equivalent mass)
than does the  hogged fuel.   The supplies of hogged fuel, however,

-------
TABLE 2-2.  ANALYSES  OF SELECTED  WOOD REFUSE
                    BURNED AS FUEL1
Item
Proximate analysis, percent
Ash
Volatile
Fixed carbon
Ultimate analysis, percent
Car her.
Hydrogen
Sulfur
Nitrogen
Ash
Oxygen (by difference!,
Heat value (bone dry) ,
J/g (Btu/lb)

Ash analysis, ppm
SiO
Al 0
Fe2°3
CaO
CaCO
MgO
MnO
P20
'K 0
Mn 0
TiO.
SO/
Fusion point of ash, °C (°F)
Initial

Softening

Fluid

Weight (bone dry) , kg/m
(lb/ft3)
Jack pine

2.1
74.3
23.6

53.4
5.9
0
0.1
2.0
38.C

20,718
(8930)

16.0
6.3
5.0
51.6
4.9
5.5
1.6
2.8
4.1
3.1
0.2
2.6

1343
(2450)
1510
(2750)
1516
(2760)
46
(29)
Birch

2. 0
78.5
19.2

57 .4
Maple

4.3
Western
hemlock

2.5
76.1 72.0
19. C

2L . 5

50.4 ! 53.6
6.7 | 5.9 : 5.8
0 0 i 0
0.3
1.8
33.8

20,578
(8870)

3.0
0
2.9
58.2
13.0
4.2
4.6
2.9
6.6
0.5
4.1
39.1

19,000
(8190)

9.9
3.8
1 .7
55.5
1.4
19.4
1.0
1.1
5.8
1.3 2.2
Trace Trace
3.2

1488
(2710)
1493
(2720)
1499
(2730)
59-70
(37-44)
1.4

1454
(2650)
1549
(2820)
1554
(2830)
50-67
(31-42)
0.2
2.5
37 .9

20,613
(8885)

10.0
2.1
1.3
53.6
9.7
13.1
1 .2
2.1
4 .6
1.1
Trace
1.4

1516
(2760)
1521
(2770)
1527
(2780)
42-46
(26-29)
Average moisture of  about  50 percent as received at firing
equipment.   Adapted  from information compiled by the Steam
Power Committee of the  Canadian Pulp and Paper Association.
                         10

-------
are usually more dependable;  also, it- is available in much larger
quantities and at comparatively lower costs.
Salt Content of Hogged Fuel
     In the northwest coastal areas of the United States, logs
transported by water can be subjected to a salt or saline environ-
ment for long periods of time (sometimes weeks or months),
especially if they are stored in saltwater.  Such storage allows
ample time for the deposition of the various salts and other
chemicals present in seawater upon and in the bark of the logs.
     Hogged fuel that is transported or stored in salt or saline
waters and used in power boilers has been shown to contain
anywhere from 0.09 to 2.2 weight percent salt (as sodium chloride,
NaCl).   The primary factors influencing salt content are length
of time the logs remain in saltwater and concentration of salt in
the water.  These factors, in turn, are affected by plant loca-
tion with respect to the wood source, log supply and demand, type
of storage (i.e., dry or wet, fresh or salt water), and, in
tidal estuaries or rivers, the amount of mixing of seawater with
freshwater.  MacLean and MacDonald  found that the average salt
content of bark from hemlock  after 6 months flotation was 1.44 to
3.13 percent, of which roughly half was absorbed in the first 3
weeks.
     Since the concentrations and proportions of the various salt
fractions in seawater vary greatly, it is difficult to predict
the total emissions of particulate and salt from hogged fuel
boilers.   Table 2-3 lists the ranges of the principal salt com-
pounds  in seawater,  with their respective boiling and melting
point s.
                              11

-------
          TABLE 2-3.  MAJOR SALT COMPOUNDS IN SEAWATER'
Compound
NaCl
MgCl2
Na2S04
CaCl2
Misc .
Percent
by weight
68.08
14.44
11.36
3.20
2.92
Melting
point, °C (°F)
801 (1474)
708 (1306)

772 (1422)

Boiling point,
O /i / O pi \
*-* V -^ /
1413 (2575)
1412 (2574)

1600 (2912)

     The effect of mixing of fresh and salt waters can be seen by
comparing the salinity at high and low tides in a coastal river.
Salinity at low tide would tend to be lower because there is a
greater flow of fresh water past a given point as the tide moves
out.  At high tide, the reverse is true.  Typical values are 9800
ppm NaCl at low tide and 13,000 ppm at high tide.
Salt Measurement Techniques
     Salt content of wood fuel or flue gas can be expressed as a
percentage of NaCl, of equivalent chloride, or of total sea salt.
This can lead to confusion in interpretation of data, since no
standard method is consistently applied to determinations of salt
content and the method of analysis often is not given.  The
amounts of salt reported as NaCl, all chloride salts, and total
sea salts can differ significantly (recall Table 2-3).
     In this regard the Council of Forest Industries of British
Columbia has stated the following:
     "The salt content of particulate matter is commonly ex-
     pressed as equivalent sodium chloride.  This value is ob-
     tained by multiplying the chloride ion concentration by
     1.58, the factor representing the weighted average of the
     ratios of chloride salt molecular weight to the chloride
     content of the molecule.  Actually, the value so obtained
     does not represent sodium chloride (which would be obtained
     by multiplying chloride by 1.66) but the total chloride
     salts.  A further correction actually is required for
                               12

-------
     accuracy because chloride salts represent about 90 percent
     of sea salts and so the factor to get total sea salts  from
     chloride content is 1.76.  This latter correction rarely  has
     been made."
     This statement emphasizes the importance of identifying the
basis upon which salt concentrations are presented and the  need
for a standard reference method.  Table 2-4 lists the methods
used at several facilities for measuring salt content of hogged
fuel and boiler emissions.
     Some procedures suggested by Boubel and Junge  for deter-
mining salt contents in wood fuels and flue gas emission samples
are summarized below.
     Because of the wide variation in salt content of fuel  ma-
terials, it is important to obtain a good representative sample.
In fuel supplies that appear to be fairly homogeneous, 10 or 12
samples should suffice.  Where the fuel supplies show a wide
variation, as many as 24 or 36 samples may be needed.  Approxi-
mately 900 g (2 Ib)  of fuel should be collected for each sample.
     Next, the fuel sample should be ground until the size  of the
particles is about 0.3 cm  (1/8 in.)  or less.   Approximately 100
to 200 grams of the ground sample is then used, and the sodium
and/or chloride portion is extracted by a standard technique for
subsequent analysis.
     The salt content of stack emissions is determined in a
similar manner.   After the fractional masses have been determined
(using EPA Method 5), the individual fractions are subjected to a
standard procedure for extraction of the sodium- and chloride-
based salts.   The total extracted sample is then ready for
analysis.

FUEL STORAGE AND PRETREATMENT
Fuel Storage
     Saltwater  transport and storage of logs is common practice
at forest  products plants in the U.S. northwest coastal region.
Typical  storage  by flat raft and bundle methods is shown in

                              13

-------
    TABLE 2-4.  TECHNIQUES FOR MEASURING SALT CONTENT OF FUEL
              AND FLUE GAS FROM HOGGED FUEL BOILERS
     Source
    Measurement technique
Weyerhauser Co.
Crown Zellerbach
Simpson Timber Co.
Washington State
 Department of Ecology
Total salt as Nad; calculated by
determining total chlorides of a
water extract of the sample and
assuming all chlorides in the form
of NaCl (multiply Cl by 1.66)

Total salt as NaCl; calculated by
determining total chlorides and
assuming that all clorides are in
the form of NaCl

Total salt as NaCl; calculated by
determining total chlorides and
assuming all chlorides in the form
of NaCl
Total salt as NaCl; calculated
stoichiometrically by determining
Na by atomic absorption and Cl by
the mercuric nitrate method
                              14

-------
Figure 2-3.   The bundle type storage allows more salt to permeate
the bark than flat raft storage because hydrostatic pressure
holds portions of the bundle well below the water line.
     Bark is generally removed with a hydraulic debarker, which
is a high-pressure saltwater jet that peels the bark from the
logs.
     Most of the processed wood waste is stored outside the plant
because covered storage space is limited.   At the Simpson Timber
Company, in Shelton,  Washington, logs are  stored by flat raft in
Oakland Bay.8  As shown in Figure 2-4, the log storage time in
the Bay ranges from 1 to 10 months.  The logs arrive by barge
during spring and summer.
     Plants that depend on one or more auxiliary fuels for con-
tinued operation provide separate storage  facilities for aux-
iliary fuels.
Fuel Pretreatment
     The fuel is usually uniformly sized before any further pre-
treatment.   If the delivered fuel is not as uniform as required,
additional  sizing is  done  at the plant.  The usual way to reduce
the size of large chunks of wood and bark  is with a hogging
machine, shown in Figure 2-5.  If still further size reduction is
required, it is usually done with a hammermill, which also is
often used  to treat bark directly after the debarker.
     All bark fuels contain grit, which should be removed prior
to burning  or be discharged from the burner with the hot gases.
It is also  possible to remove grit as a molten slag from the
burner.
Predrying Systems for Fuel
     Predrying is an  important pretreatment for saltwater-borne
hogged fuel.  This treatment does not greatly reduce the salt
content of  the hogged fuel, but it improves the efficiency of
combustion  and reduces overall emissions from the stack.  The
following discussion  of predrying is extracted in large part from
Reference 1.

                               15

-------
             TYPICAL FLAT RAFT
                                         TYPICAL BUNDLE
     Figure 2-3.  Storage  of logs by flat  raft and bundle
                  10
                 o
                 c:
                                  AVERAGE 3.6 mo
                                   10
                           NUMBER OF LOG RAFTS
20
Figure  2-4.   Log storage time in Oakland Bay for Simpson Timber Co.


                                 16
                                                                        8

-------
               DOUBLE
               BREAKING
                PLATE
COVER DIVIDES  HERE
                                                 METAL TRAP
Fiaure  2-5.   Cross-section  of a typical  hogging machine.
                             17

-------
     Systems for predrying wood fuels are relatively new.  They
were developed to overcome two serious shortcomings related to
moisture content.  The first problem is the extreme variability
in moisture content of hogged wood, sawdust, bark, and even other
"dry" fuels.  The moisture content is affected by species,
handling, storage conditions, and similar factors.  Drying the
fuel outside the furnace allows both manufacturers and operators
to deal with a more uniform fuel.
     The second function of predrying is reduction of the mois-
ture content.  This increases both the thermal efficiency and the
steam-generating capacity of the boiler.  Figure 2-6 illustrates
the effect of fuel moisture on steam production.  The drier fuel
can be ignited more readily, since the energy needed to evaporate
water can go instead to volatilization of combustibles.  The
boiler responds more rapidly with drier fuel.  Elimination of
moisture from the flue gas reduces both the gas volume and the
corresponding gas velocities.  Thus, smaller fans can be used,
and particulate carryover is reduced.
     Fuel moisture may be controlled by several methods:
     1.   Vibrate loose water off the fuel on a shaker screen.
     2.   Press out water mechanically.
     3.   Control the processes that generate the fuel to limit
          water addition.
     4.   Drive off moisture by heating the fuel in dryers.
     Removal of water by vibration may be effective when the
moisture content exceeds 55 percent.  If the process that gen-
erates the wood adds large quantities of moisture  (for example,
hydraulic debarking), vibration can be an inexpensive and low-
maintenance approach to control of surface moisture.
     Presses can remove only limited amounts of moisture.  With
most hogged fuel, pressing can reduce moisture  levels to  50 to  55
percent.   Disposal of the bark pressate water is a serious
problem,  however; and the method is therefore not practical for
                               18

-------
                                 FURNACE BLACKS  OUT
                              ~68% MOISTURE
                            LIMITS OF COMBUSTION
                     20               40

                     FUEL MOISTURE, % (wet basis)
                                          60
Figure  2-6
Effect  of fuel moisture on  steam production.

                19
                                                               12

-------
removing moisture from salt-soaked hogged fuel.  Control of water
additions to fuel in production processes is usually difficult.
For example, most plants cannot replace hydraulic debarkers with
mechanical systems, and wood product plants have little control
over the transport and storage of logs in salt water.
     Heating the fuel can reduce its moisture content.  Moisture
levels in a range from 25 to 35 percent are usually adequate for
good combustion.  At levels below 20 percent, the dry fines can
cause significant dust problems.  Heating-type dryers, however,
can generate pollutants of three types:  if the wood fuel is
overheated  (above 300°F), the volatile organic material will
evaporate and leave the dryer with the exhaust gas stream, which
may condense in the atmosphere to form a visible plume; the dry
fines^may create dust; and, where the dryer is fired by a sep-
arate combustion system, the products from that system may
become pollutants.
     Three systems are currently being considered for drying fuel
outside the furnace-boiler system:  the hot hog, hot conveyor,
and the rotary dryer.    These systems can be operated with
separate burners  (fired with sanderdust or other fines) or by
directing boiler flue gases from the stack to the fuel dryer.
Use of stack gases puts the drying system in series with the
boiler.  Thus a fuel dryer breakdown interrupts the feeding of
dry fuel to the boiler, and a boiler breakdown shuts down the
fuel dryer.  These and many other factors must be considered with
respect to external fuel drying.  Table 2-5 summarizes advantages
and disadvantages of the three major external drying systems.

COMBUSTION OF HOGGED FUEL
     Because of the variable properties of saltwater-borne bark
and wood residue, designing a furnace that will properly consume
fuel to generate heat for the boiler with minimum particulate
carryover is a difficult task.  Some understanding of salt carry-
over in the flue gas is helpful in design of new boilers.  Boubel
                                20

-------
TABLE 2-5.   ADVANTAGES AND DISADVANTAGES OF THREE MAJOR
                EXTERNAL DRYING SYSTEMS14
                         Hot hog

 Advantages:   Exposes more surface by grinding to dry
              quickly;  can use boiler stack gas or separate
              heat source that burns wood fines

 Disadvantages:   High energy requirements kW,  (hp) ;
                 high maintenance costs;  limited moisture
                 reduction

                     Hot conveyor

 Advantages:   Can utilize boiler  stack gases

 Disadvantages:   Low gas temperatures;
                 low moisture release;
                 high maintenance costs;
                 low capacities

                     Rotary drum  drying

 Advantages:   Can accept high inlet gas temperatures;
              can dry large quantities of high-moisture
              material;  low energy requirements (kW,  hp) ;
              low maintenance costs; high retention time

 Disadvantages:   Requires space for installation
                          21

-------
and Junge  provide an explanation of physical reactions of salt
in boilers.  They suggest that when the fuel burns, a portion of
the entrained salt, being noncombustible, leaves the boiler as
particula.te matter in the flue gas.  The remainder is deposited
in the boilers.  Although the mechanisms by which salt is con-
verted to small aerosols in the boiler are not clearly defined,
the physical characteristics of salt provide a guide to possible
reactions.  The major portion of the salt is sodium chloride  (60
to 70%) and magnesium chloride  (15%).
     Since the combustion zone temperatures are typically above
1093°C  (2000°F), sodium chloride is almost completely vaporized
in the combustion zone.  Magnesium chloride, having very similar
melting and boiling points, is also vaporized.  Condensation of
the vaporized salts occurs as the gases are cooled in the con-
vection passes of the boiler.  It is hypothesized that rapid
cooling or quenching of the vapors causes formation of very small
particles.  Slow cooling, conversely, leads to formation of
larger particles.  Since the rate of gas cooling is not constant
and uniform throughout the boiler, the particles occur in dif-
ferent sizes.  Temperatures of the gases leaving the combustion
zone generally drops from the peak value to 316°C  (600°F) in
about 2 seconds.  This quenching is rapid enough to create many
fine particles of salt in the flue gases.
     Salt-laden hogged fuels are burned in the same furnaces
that burn other wood fuels.  The design of such furnaces must be
flexible enough to handle the fuel, with nonuniform moisture
content, and still follow the steam load demand on the boiler.
The furnace- may be separate from the boiler or integral with  it.
If it is separate, the firing is outside the boiler; and the  hot
gases, which are probably still burning, are directed from the
furnace to the boiler.  If the furnace is integral with the
boiler, the fuel is burned in the boiler, which is surrounded by
a heat transfer surface.  Both types are in use in the United
States today.
                               22

-------
     Detailed descriptions of various furnace types are given in
Appendix B.  The following discussion from Corder   briefly
summarizes these furnace types:  Dutch oven, spreader stoker,
inclined grate, suspension firing, cyclone, and direct firing.
Dutch Oven
     This is a two-stage furnace, consisting of (1) a Dutch oven,
in which moisture is evaporated and fuel gasified, and (2) a
secondary furnace, in which combustion is completed.  Fuel is fed
by gravity through an opening in the Dutch oven and forms a
conical fuel pile.  The Dutch oven has been widely used in the
past, but most new installations are using other systems.
Spreader Stoker
     The spreader stoker is used on many new installations.
Pneumatic or mechanical spreaders introduce fuel above the grates.
Some of the fuel burns in suspension; and the rest falls on the
grates, where burning is completed.
     A modified form of the stoker furnace is the inclined-grate
type, shown in Figure 2-7, in which the fuel is introduced in a
continuous ribbon at the top of the grate.  Moisture is removed
in the upper section, and burning is completed in the lower
section.  Ash is removed from the lowest section of the grate.
Suspension Firing
     This recent method for firing hogged fuel resembles a
pulverized-coal-fired system.  Hogged fuel of very small size
is blown into the furnace alone or in combination with natural
gas or oil.
Cyclone Furnace
     Firing of a cyclone furnace can be either horizontal or
vertical.   The horizontal type is a version of the Babcock and
Wilcox coal-fired cyclone furnace, and in fact it requires coal
as the primary fuel for combustion with hogged fuel.  The coal
ash provides a slag coating to ensure proper burning of hogged

                               23

-------
                                          APPROXIMATE CONTOUR
                                           OF WOOD REFUSE
                                                BED
                                            N WATER COOLED
                                              INCLINED GRATE
                                  LATERAL
                                  ZONING
                                   WAIL
 WUILLOTINE TYPE
ASH REMOVAL DOORS
                                                SHIELD
                                              FOR PROTECTIOI
                                               Of OPERATOR
                                              •HUE KMOVIN
                                                 ASM
Figure 2-7.    Inclined-grate stoker  furnace at a  paper
            and  pulp mill  in  British Columbia.15
                                 24

-------
fuel, which cannot exceed 30 percent of the furnace heat input
and must pass through a screen of 19-mm (3/4-in.) mesh.
     The vertical cyclone furnace was developed in the sixties in
Scandinavia especially for hogged fuel.  It is a refractory-lined
cylinder, in which an underfeed stoker pushes fuel up through the
bottom grate to form a conical pile.  Primary ,air under high
pressure enters the furnace tangentially from below to provide
cyclonic action (see Figure 2-8).
Direct Firing
     Direct-fired systems use hot gases from hogged fuel burning
as a supply of high-temperature gases for veneer dryers, lumber
dry kilns, and dryers for wood and bark particles.  The Energex
system fires finely divided wood or bark fuel in a cyclonic
burner.  Hot gases from this system can be used in a rotary drum
dryer for hogged fuel.  A pile burning furnace has reportedly
been used as a heat source for a veneer dryer,   and suspension
burning of undried bark in a cylindrical annular combustion
chamber on a laboratory basis has also been reported.

PARTICLE SIZE DISTRIBUTION OF SALT EMISSIONS
     The size distribution of the salt particles strongly affects
the removal efficiency of control equipment.  Particulate emis-
sions from boilers burning hogged fuel with no salt can be
handled adequately by conventional control devices.  With salt-
laden hogged fuel, however, the size range of the particulate
emissions is much smaller; and the size fractions below 1 ym
typically are composed of more than 50 percent salt.
     In an analysis of the effect of salt (NaCl) on overall mean
particle size in emissions from hogged fuel boilers, the salt
fraction reduced the overall mean particle size by a factor of
approximately 10 (see Figure 2-9).18  AS the figure shows, the
mean particle size of the nonsalt fraction is about 17.5 ym.
The mean particle size of the salt fraction is about 0.23 ym.
The overall mean particle size, about 2.15 ym, indicates the

                               25

-------
                                        PRIMARY
                                          AIR
Figure 2-8.  Bark-fired  cyclone  type  furnace.
                      26

-------
                                                       O ASH ONLY
                                                        SALT AND ASH
                                                       LH SALT ONLY
                COMPOSITE SIZE DISTRIBUTION
       0.1
                    10     ?0    30   40  50   60   70   80  85

                        PERCENT LESS THAN STATED PARTICLE SIZE
        Figure 2-9.   Particle  size distribution  at Weyerhauscr
Co.,  north  Bend, Oregon,  plant  (composite  from  11 tests).

                                    27

-------
extent of size reduction caused by salt in the emissions.  Some
data show that the overall mean particle size of the salt-laden
particulate can be as low as 0.2 ym.
     Comparison of particle size data from the Weyerhauser North
Bend plant with data from Crown Zellerbach's Port Townsend plant
and data from British Columbia verifies the reduction in mean
particle size due to salt particles.  Table 2-6 summarizes parti-
cle size data from various salt-laden hogged fuel boilers.  All
of these impactor measurements were taken at the outlet of a
multiclone or cinder collector.  The reduction in the overall
mean particle size can result in dust even finer than the North
Bend composite shown in Figure 2-9, depending on the size dis-
tribution of the nonsalt fraction of the dust.

EFFECTS OF SALT-LADEN HOGGED FUEL BOILER EMISSIONS ON AMBIENT
AIR QUALITY
     An important consideration in assessment of salt emissions
from hogged fuel boilers is the effect of these emissions on
ambient air quality in the vicinity of the boilers.  This section
summarizes data on ambient air quality near three facilities that
fire salt-laden hogged fuel:   (1) Simpson Timber Company's
Shelton plant, which is controlled by a fabric filter,  (2) Crown
Zellerbach's Port Townsend plant, which is controlled by a
venturi scrubber, and (3) Weyerhauser Company's North Bend plant,
which is controlled only by multiclones.  Federal and state
standards for suspended ambient particulate matter and  standards
of Washington and Oregon are shown in Table 2-7.
                               28

-------
     TABLE 2-6.   PARTICLE SIZE DATA FROM HOGGED FUEL BOILERS
                  WITH EXCESSIVE SALT EMISSIONS3
Source
Weyerhauser
North Bend,
Oreg .
Crown Zellerbach
Port Townsend,
Wash.
Simpson Timber Co.
Shelton, Wash.

St. Regis Paper Co
Tacoma, Wash.
Plant not
specified,
British Columbia
Salt in
fuel, %
0.1 - 0.9.
(0.48 avg.)

0.3 - 1.8


1.6 - 2.15
(dry basis)

Not
given
Low salt

High salt
Particle size dist.
Mean particle
size (x) , ym
h
2.2°

0.35 - ~3.0


85 to 95%
less than
0 . 4 ym
Salt 0.25,
overall "18
3.2

2.0
Geometric
deviation, o-

= 10

7.4 - -10


NA


NA

i—i
— /

= 39
Reference

18, 19

4,c


20, d


21

5


  Measurements are after the rnulticlone or cinder collector
  unless otherwise noted.
  Composite of 11 tests.
  Personal communication with Mr. Alan Rosenfeld, Crown  Zellerbach,
  Environmental Services.
  Personal communication with Mr. Robert Hoit, Simpson Timber  Co.
NA - Not available.
                               29

-------
to

o
                  80
                  70
                  60
                  50
                01
                n
               o 40
                  30
                  20
                  10
                                   WASHINGTON STATE AMBIENT AIR QUALITY STANDARD
                       A = SHELTON CITY

                       B = SHELTON PUD BUILDING


                       ,   ,  • J  .   .  .  I   .
                                                               i   i  i
                                                                                   i   i  i
97 1
1972
1973
                                                      1974-H-" — 1975 -*H — 1976  >|<  1977— H
                     Figure  2-10.  Suspended particulate data,  Shelton,  Washington,

                                   (12-month moving  geometric means).22

-------
       TABLE  2-7.  PARTICULATE AMBIENT AIR QUALITY  STANDARDS
                              (yg/m3)
Averaging
(time)
Annual
geometric
mean
24 hours
1 month
Federal
Primary
(health)
75


260

Secondary
(welfare)
60


150

Washington
60


150

Oregon
60


150
100
Simpson Timber Company, Shelton, Washington
     Air monitoring has been done in Shelton since 1965, when  the
Washington State Department of Health established particle fallout
stations within the city limits.  Air monitoring by the Olympic
Air Pollution Control Association (OAPCA) began in 1969.  The
measurements of suspended particulate are taken with high-volume
(HV) air samplers.  Values measured over a period of 5 years at
two sites in the Shelton city limits are summarized in Table 2-8
and shown in Figure 2-10.   These sites are located northwest of
the Simpson Timber Company's plant,  where fabric filters were
installed on the salt-laden hogged fuel boilers in 1976.  As
these data show, the particulate air quality has been in com-
pliance with standards on  an annual basis since 1970.  The
installation of the baghouses on Simpson's hogged fuel boilers in
1976 apparently did not cause an appreciable improvement in the
ambient air quality in the Shelton area.  The data may have been
affected,  however, by relocation of the ambient air monitor to
its present site in the same year.
                              31

-------
       TABLE 2-8.  SUMMARY OF SUSPENDED PARTICULATE DATA,
         SHELTON, WASHINGTON22  (yg/m3 except as noted)

Sampler
location
Shelton
City
Hall



Shelton-PUD

Year
1970
1971
1972
1973
1974
1975
1976

Monthly Avg .
Low
23
23
19
23
24
20
19
High
108
92
104
68
79
65
72

24 -hr
max.
194
132
155
103
171
99
132

No. >
150
3
0
1
0
1
0
0

No. >
260
0
0
0
0
0
0
0
Annual
geometric
mean
47
42
43
38
40
33
34
Crown Zellerbach, Port Townsend, Washington
     OAPCA began monitoring suspended particulate and particle
fallout at the high school in Port Townsend in 1970.  In 1973
the HV monitor was moved to its present location at the Port
Townsend Fire Station.  Both of these sites are located northwest
of the Crown Zellerbach mill and its hogged fuel boiler.  The
mill is the major source of air pollution in the Port Townsend
area.
     Measurements of suspended particulate are summarized in
Table 2-9 and shown graphically in Figure 2-11.  The data show
that both monitors have been in compliance with Federal and state
ambient air regulations since 1971.
     Crown Zellerbach installed a new hogged fuel boiler and
venturi scrubber at the Port Townsend mill in September 1977.
Because ambient air data are not available for 1977, it is not
known whether the better control of the salt-laden particulate
has yielded a measurable reduction in suspended particulate.
Weyerhauser, North Bend, Oregon
     Boubel and Junge  have summarized ambient air data in the
vicinity of Weyerhauser's North Bend plant.  The Oregon Depart-
ment of Environmental Quality has operated a total suspended
particulate monitor on the roof of the Coos Bay City Hall since
                               32

-------
TABLE 2-9.  SUMMARY OF TOTAL  SUSPENDED PARTICULATE DATA
         OBTAINED NEAR CROWN-ZELLERBACH BOILER22
                 (yg/m3 except  as  noted)
Sampler
location
Port Townsend
High School


Port Townsend
Fire Station


Period
of
record
1-12/70
1-12/71
1-12/72
1- 6/73
6-12/73
1-12/74
1-12/75
1-12/76
Monthly avg .
Low
16
15
17
17
20
25
High
47
51
48
39
38
61
24 ! 55
22
46
24-h
max .
70
80
90
67
47
170
88
68
No. >
150
0
0
0
0
0
1
0
0
No. >
260
0
0
0
0
0
0
0
0
Annual
geometric
mean
25
25
28
24
28
35
34
31
                          33

-------
   80
   70
   60
no
 E
 en

 ^ 50
 z
 o
                         WASHINGTON  STATE AIR QUALITY STANDARD
 a:
   40
   30
   20
   10
    0.
            A = PORT TOWNSEND HIGH SCHOOL
            B = PORT TOWNSEND FIRE STATION
                                                                  I	,
1971  »l<  1972   »h  1973  *\«  1974  >h  1975  H<   1976
                                                                      1977
      Figure  2-11.   Suspended particulate  data, Port  Townsend,  Washington,
                       (12-month moving geometric means).  ^

-------
1970.  This site is approximately 3 miles south of the Weyer-
hauser North Bend stack.   Figure 2-12 summarizes the geometric
means of measurements for each month in 1976 and 1977, and Table
2-10 summarizes yearly data since 1970.
     The total suspended  particulate for the Coos Bay sampling
station has been in compliance with Federal and state ambient air
regulations since 1970.   Weyerhauser's three hogged fuel boilers
at the North Bend plant are equipped with multiclones for par-
ticulate control.
General
     Since 1970, monitors near the hogged fuel boilers at Shelton,
Port Townsend, and North  Bend have shown compliance with an'nual
suspended particulate standards, and only Shelton, in 1970 were
there at violations of the 24-hour maximum average standard.
     Not enough data are  yet available after installation of
secondary collectors at Port Townsend in 1977 and Shelton in 1976
to determine whether air  quality has improved at the local
monitoring sites.
     The Council of Forest Industries of British Columbia
offers the following pertinent comment:
          The dispersion  of salt from a hogged fuel boiler
     depends on meteorological conditions in the vicinity of a
     mill.  With a high stack and reasonable wind velocities, the
     stack plume should be dispersed rapidly, and the salt should
     only add fractionally to that already in the air (coastal
     area).  Locations with low wind velocity and inversion
     conditions, may cause the plume to linger in the vicinity of
     the mill and significantly increase ambient background salt
     levels.

EFFECTS OF SALT-LADEN PARTICULATE ON CORROSION RATES
     Data on the effects  of airborne salt on corrosion rates are
              23
minimal.  Egan   states that such effects are very difficult to
determine.  Tsang and Stubbs,24 reporting on a study of the ef-
fects of salt emissions from a hogged fuel boiler in a coastal
                               35

-------
                    TABLE  2-10,
SUMMARY OF TOTAL  SUSPENDED  PARTICULATE DATA,
  COOS  BAY SAMPLING STATION7
Station number
and location
Coos Bay -
1th and Central
(City Hall)
0607101






Year
1970

1971
1972
1973
1974
1975
1976
1977
No. of
samples
89

49
81
56
52
59
54
56
Days >
150
-I

1
0
1
0
0
0
1
260
0

0
0
0
0
0
0
0
Particulate concentration, ug/m
Annual
geometric
mean
51.7

53.6
44.9
50.4
47.9
37.1
40.6
43.3
24-h avg .
Maximum
152

185
108
164
127
95
110
220
2nd highest
137

137
103
123
111
93
110
130
U)
(Ft

-------
    80
    70
     60
     50
 en

 3.
 o

 t—
 «=c
     40
 3   3(
        JFMAMJJASONDJFMAMJJASOND

        •	1976	4*	1977	H
Figure 2-12.  Suspended particulate data, Coos Bay,  Oregon,

                 (monthly  geometric means).
                            37

-------
environment, state that although the sodium chloride levels in
the surrounding environment increased measurably, the mill under
study did not cause chloride corrosion to increase beyond natural
background levels in the 1973 to 1975 study period.  Apparently
the salt deposition rate, an important factor in salt related
corrosion, was not high enough to cause an increase in the cor-
rosion rate.
                               38

-------
                   REFERENCES  FOR SECTION 2


1.   Boubel,  R.W.   Control  of  Particulate Emissions from Wood-
     Fired Boilers.   EPA 340/1-77-026.   1977.

2.   Suprenant,  N.,  et al.   Preliminary  Emissions Assessment of
     Conventional  Stationary Combustion  Systems.   Prepared for
     U.S. EPA by GCA Corporation,  Bedford, Mass.   January 1976.

3.   Whitman, J.E.,  and H.  Burkitt.   Compliance Alternatives for
     Stack Emissions from the  Hog Fuel Boilers  at North Bend,
     Oregon.   Prepared for  Weyerhauser Co.   December 1977.

4.   Cupp, S.J.   Operating  Experience with a Boiler Firing Salt-
     water Borne Hogged Fuel.   Crown Zellerbach Corporation, Port
     Townsend, Washington.   1978.

5.   Council  of  the Forest  Industries of British  Columbia.  The
     Basis for Requesting a Variance on  Marine  Salts as Polluting
     Particulates  in Stack  Gases from Hog Fuel  Fired Boilers.
     Submitted to  the Pollution Control  Branch, Department of
     Lands, Forests, and Water Resources, Government of British
     Columbia, Victoria, B.C.   September 1974.

6.   MacLean, H.,  and B.F.  MacDonald. Salt  Distribution in Sea
     Water Transported Logs.  Information Report  VP-X-45.

7.   Boubel,  R.W.,  and D.C. Junge.   The  Impact  of Salt Emissions
     from Weyerhauser Co. Wood Fired Boilers, North Bend, Oregon.
     Prepared for  Weyerhauser  Co.   March 1978.

8.   Leman, M.J.  Special Environmental  Problems  Originated by
     Burning  Bark  from Saltwater-Borne Logs. Proceedings of a
     Conference  on Wood and Bark Residues for Energy, Oregon
     State University.   February 1975.

9.   Henriksen,  J.S.  A Discussion of Contemporary Canadian Wood
     Waste Fired Boiler Systems.   MacMillan  Bloedel, Ltd.,
     Vancouver,  British Columbia.

10.   Brenton, M.D.   Fuel Preparation for Hog Fuel Boiler:  The
     System,  the Problems,  the Benefits.  A  Report of Weyerhauser
     Co.
                              39

-------
11.  Junge, D.C.  Boilers Fired with Wood and Bark Residues.
     Oregon State University, Forest Research Laboratory,  Research
     Bulletin 77, Corvallis, Oregon.  November 1975.

12.  Johnson, R.C.  Some Aspects of Wood Waste Preparation for
     Use as a Fuel.  Tappi,  58(7):  102-106, 1975.

13.  Corder, S.E.  Wood and Bark as Fuel.  Oregon State Univer-
     sity, Forest Research Laboratory, Research Bulletin 14,
     Corvallis, Oregon.  August 1973.

14.  Porter, S.M., and R.W.  Robinson.  Waste Fuel Preparation
     System.  Steams-Rogers, Inc., Denver.  Presented at  Energy
     and the Wood Products Industry Proceedings, sponsored by
     Forest Products Research Society, Atlanta.  November  15-17,
     1976.

15.  Corder, S.E.  Properties and Uses of Bark as an  Energy
     Source.  Oregon State University, Forest Research Labora-
     tory, Research Paper 3, Corvallis, Oregon.  April 1976.

16.  Deardorff, D.  Direct Combustion Systems.  In:   Wood  Residue
     as an Energy Source, A.B. Brauner, ed.  Proceedings P-
     75-13, Forest Products Research Society, Madison, Wisconsin,
     1975.  pp. 89-91.

17.  Jasper, M., and P. Koch.  Suspension Burning of  Green Bark
     to Direct-Fire High Temperature Kilns for Southern Pine
     Lumber.  In:  Wood Residue as an Energy Source,  A.B.  Brauner,
     ed.  Proceedings P-75-13, Forest Products Research Society,
     Madison, Wisconsin, 1975.  pp. 70-72.

18.  Whyte, J.E.  Hog Fuel Boiler Emissions - Particle Size Via
     Cascade Impactor Tests.  Weyerhauser Co., North  Bend,
     Oregon.  Technical Report No. 046-4504.  August  1976.

19.  Whyte, J.E.  Hog Fuel Boilers Particulate Emissions Compli-
     ance Tests, Without Ash Reinjection.  Weyerhauser Co.,
     North Bend, Oregon.  Technical Report No. 046-4210.  July
     1977.

20.  Ersnst,•C.F., and P.A.  Hamlin.  Evaluation of the Perfor-
     mance of the Ceilcote Ionizing Wet Scrubber on a Hogged  Fuel
     Boiler Burning Hogged Fuel'With a High NaCl Content.   ITT
     Rayonier, Inc., Olympic Research Division, Shelton, Wash-
     ington.  September 1976.

21.  Nelson, P.A.  Bimodal Nature of Aerosol Size Distribution of
     Industrial Effluents.  Washington Department of  Ecology,
     Olympia, Washington.  Presented at a Meeting of  the Air
     Pollution Control Association, Spokane, Washington.  November
     15, 1977.


                               40

-------
22.   Olympic Air Pollution  Control  Authority.   The  Olympic Air
     Pollution Control  Authority, as  a  Composition  of  Clallam,
     Jefferson,  Mason,  Grays  Harbor,  Pacific,  and Thurston Coun-
     ties.   Olympia,  Washington, August 1977.

23.   Egan,  F.J.   Effect of  Environmental  Factors on the  Corrosion
     of Steels.   Australian Corrosion Engineering,  15(6),  July
     1971.

24.   Tsang,  K.K.,  and K.P.  Stubbs.  Assessment  of the  Impact  of
     Sodium Chloride  Emissions  from a Coastal Pulp  Mill.   Pro-
     ceedings of the  69th Annual Meeting  of the Air Pollution
     Control Association, Portland, Oregon.  1976.
                             41

-------
                            SECTION 3
          CONTROL TECHNOLOGY TO REDUCE SALT EMISSIONS
                    FROM HOGGED FUEL BOILERS
     Among the several preventive and remedial ways of reducing
salt-laden particulate emissions from hogged fuel boilers, the
best preventive method is simply not to transport and store wood
residue and bark in saltwater.  Most often, however, this is not
possible.  Other preventive measures involve preparing the fuel
in such a way as to maximize combustion efficiency and minimize
stack emissions.  Remedial solutions are to improve an existing
control system or to install a more efficient secondary control
device.
     This section discusses the presently available technology
for reducing salt emissions from hogged fuel boilers:  fuel
handling and pretreatment, combustion modifications, and use of
conventional and novel particulate control devices.

FUEL HANDLING AND PRETREATMENT
Type and Duration of Storage
     Water storage of logs as flat rafts or small-log bundles, as
shown earlier in Figure 2-3, can affect salt emissions.  In flat
rafting of logs about 45 to 50 percent of each log remains out of
water whereas in bundled logs only 10 to 20 percent of the logs
are out of water.  As a result the salt content of flat-rafted
logs is lower than that of bundled logs, and the resultant salt
emissions from the hogged fuel are lower also.
     Stora.ge time of logs in saltwater also governs the salt
particulate emissions.  Reducing the storage time from 3 to 4
months to about 5 days usually yields a significant decrease  in
salt emissions from the stack.   Figure 3-1 illustrates the
                               42

-------
   o>
5  !•  6
a:

i— a


\ 1   2

o c
,
(X) O   o
   oj  U


cc

CD i—
                                   TYPICAL CHLORINE

                                        RANGE
            10     20     30    40     50

                  TIME STORED IN BAY,  days
60
70
    Figure  3-1.  Relation of  time in  saltwater  to

             absorption of water and salt.1
                        43

-------
relation of time in  saltwater to absorption of water and salt.
Another study  confirms this effect, indicating that approxi-
mately half of the total salt absorbed at 6 months is absorbed in
the first. 2 or 3 weeks of contact with seawater.
Fuel Pretreatment
Bark Pressing—
     A well-engineered pressing operation reduces the moisture
content of saltwater-soaked bark to about 50 percent wet basis.
The presses also remove substantial quantities of salt, and thus
reduce plume opacity and salt particulate emissions when the bark
is burned.  The bark pressate water, however, is a "high-strength1
wastewater that requires primary and secondary treatment plus
deepwater outfall.  Thus, although bark pressing does present an
opportunity to substantially reduce opacity and particulate
emissions by lowering salt content, this treatment creates a
wastewater disposal problem.
Methods of Fuel Preparation--
     As discussed earlier, boiler operators try to minimize the
amounts of dirt and moisture in fuel and the size of fuel fed to
the boiler.  Rinsing bark with freshwater both reduces the salt
content and cleans the fuel.  This treatment prevents accumula-
tion of excessive dirt in the boiler and reduces carryover of
salt in the flue gas.  Fuel size can be reduced by hogging and by
hammermilling.
     Moisture content of the saltwater-borne hogged fuel con-
tributes significantly to the presence of combustible particulate
                          4
matter in stack emissions.   Experiments have been done with
fuels having different moisture contents.  In one study, the
opacity of stack emissions increased by 1/2 to 1 Ringelmann
number when the moisture content increased from 57.4 to 61 per-
cent.   The wetter the fuel, the more difficult it is to hold
excess air at the desired low levels.  Figure 3-2 shows why
furnaces will black out or flame out with fuel having high
moisture content.  Temperatures shown are calculated theoretical
                              44

-------
      3,000
uj 2,000
a.
Z
LJ

Z
o

I/I

1C

o 1,000
    Z
    x
                      SOX MC
                       FUEL
                     sox EXCESS
                       AIR
                     12 SX
                           OBYWOOD  ^
                            FUEL    T
                          SOX EXCESS/
                             AIR ^/
                           IZ iX C02
                         60% MC
                          FUEL
                60t MC   IOOX EXCESS
                FUEL       AIR
              ISOX EXCESS   »4X CO,
                 AIR
                    Z
               IGNITION
             TEMPEATURE
Figure  3-2.  Maximum combustion  temperatures
   of a mixture  of Douglas fir and western
hemlock fuel under four  sets of  conditions.
                       45

-------
maximums.  Actual furnace temperatures are several hundred
degrees lower, depending on waterwalls and other furnace design
characteristics.
     Volatile gases represent 60 percent or more of the heating
value of wood fuel.  When furnace temperature drops below the
ignition temperature of the volatile gas, the furnace blacks out
and steam production is reduced drastically.  Unburned hydro-
carbons greatly increase the opacity of stack gas.  The water-
vapor contents measured in flue gases from hogged fuel boilers
range from about 6 to 32 percent by volume.  In burning hogged
fuel with an average moisture content of 45 to 50 percent by
weight, the water vapor content of the flue gas in the stack is
about 20 percent by volume.   This value varies with moisture
content of the fuel, relative humidity of the air, and the per-
centage of excess air.  Higher water vapor content in the flue
gas causes a white plume that does not violate opacity standards
but does contribute to an increase in opacity of stack emissions.
     The different systems of predrying to reduce the combustible
matter in the stack gas are discussed in more detail in Section 2.

COMBUSTION MODIFICATIONS
     Combustion modifications can maximize combustion of hogged
fuel in the firing zone and thus attenuate opacity.  The salt
in the hogged fuel, which contributes to a major portion of the
opacity, is noncombustible particulate matter.  Hence an increase
in rate of combustion will not reduce the carryover of salt
particulates in the flue gas, but can lead to a significant
reduction of other combustible particulate matter, which contri-
butes to opacity and stack emissions.
     Operating parameters that affect particle entrainment and
combustion include initial fuel moisture content, furnace excess
air levels,  and steam generation rate.   Investigation of the
influence of these parameters on carryover mechanisms could lead
                               46

-------
to changes in plant operation that would significantly reduce the
emission of combustible particulate matter.   Combustion model
studies and furnace entrainment models of hogged fuel (without
salt) show that an increase in initial fuel  moisture reduces the
burning rate and increases emissions in the  flue gas.   An in-
crease in excess air has the same effect but also reduces opacity,
Increasing the steam generation rate increases both burning rate
and furnace gas velocity.
     A recent experimental study  of the problems associated with
combustion of wood residue fuels illustrates that (1) combustion
on the grate can be controlled to reduce particulate emissions,
(2) increasing the fuel bed depth reduces particulate emissions,
an effect that is more pronounced at higher  combustion rates, and
(3) in a spreader-stoker boiler with a bed of wood fuel thick
enough to consume all of the oxygen in the underfire air, the
flow rate of underfire air controls the rate of combustion.
     With regard to control of noncombustible salt particles in
flue gas, recall the hypothesis presented pn page 22.  Addi-
tional work is needed on ways to slow the quenching effect on
vaporized salt leaving the combustion zone.   This would generate
larger particles of salt that would be collected more easily by a
conventional control system than are submicron salt particles.
     Combustion control modifications to hog fuel boilers during
combination firing with auxiliary fuels such as oil, natural gas,
coal, or solid waste would reduce their consumption as well as
carryover of combustible particulate in flue gas.
S^tack Diameter
     A consideration that is not a control technique but affects
visible opacity of particulate emissions is  the stack diameter.
The larger the diameter of the stack, the higher the apparent
opacity, even though actual smoke density and emissions are
constant (see Figure 3-3).  This effect can  be significant at
large steam plants or where several boilers  discharge into a
                               47

-------
  100 i—   5
   80
S  60
s_
CJ
Q.
O
O-
o
   40
   20
    OL_   o
                    5        10       15      20


                  CHIMNEY INSIDE DIAMETER, feet
Figure  3-3.   Increase in apparent  density of smoke
          with increasing  chimney diameter

             (actual  density is constant).
                       48

-------
common chimney.  Smaller stack diameters that yield lower visible
opacity readings should be considered in the design of new plants,

CONVENTIONAL PARTICULATE CONTROL DEVICES
     Four types of conventional control devices are used to re-
duce particulate emissions from hogged fuel boilers:  mechanical
collectors, electrostatic precipitators (ESP's), wet scrubbers,
and fabric filters.  Only the latter three can remove submicron
salt particles.  Mechanical collectors, nonetheless, are used on
salt-laden hogged fuel boilers and ESP's are not.
     The following sections on each control device summarize
basic design parameters, applicability for controlling salt
emissions, performance on hogged fuel boilers (salt-free and
salt-laden fuel), operation and maintenance problems, and dis-
posal techniques for collecting salt/particulates.  The best
available method of collecting salt/particulates is also dis-
cussed.  Case histories of a fabric filter and venturi scrubber
operating on salt-laden hogged fuel boilers are presented in
Appendix C.
Mechanical Collectors
General System Characteristics--
     Particulate emissions from hogged fuel boilers have tradi-
tionally been controlled by mechanical collectors.  Although
mechanical collectors generally do not meet applicable emission
regulations when controlling salt-laden particulate, discussion
of system characteristics and design philosophy is included in
this report because of the widespread use  of mechanical collec-
tors,  and their importance as part of a total system of parti-
culate control.  Figure 3-4 shows a multiple cyclone (multiclone)
collector, and Figure 3-5 illustrates gas  flow through a single
cyclone.  These collectors are usually designed for dust loadings
of 2 to 11 g/m  (1 to 5 gr/scf).    Particle size distribution of
ash from combustion of nonsalt bark is normally 30 to 40 percent
                              49

-------
         GAS
       OUTLET
 GAS
INLET
                          PARTICLE
                          DISCHARGE
Figure  3-4.   Simplified  diagram of  a  multiple cyclone.
                            50

-------
INLET
                              GAS EXHAUST
                                  PARTICLE
                                  SEPARATION
                                    ZONE
                                PARTICLE
                          ,o  /  DISENGAGING
                                  ZONE

                        PARTICLE
                         OUTLET
   Figure  3-5.
Cyclone  collector for particles
  in  flue gases.5
                          51

-------
less than 10 micrometers.  Pressure drop across the mechanical
collector ranges from 3.8 to 6.4 cm (1.5 to 2.5 in. water) for
                           g
best collection efficiency.
     With ash from salt-free bark, mechanical collectors can
sometimes achieve collection efficiencies of 85 to 90 percent
with outlet loadings of 172 to 258 kg/J  (0.4 to 0.6 lb/106 Btu).
A collection efficiency of 30 to 70 percent is a more realistic
estimate.  Even under optimum conditions opacity may be a prob-
lem, especially with stoker-fired boilers, which respond slowly
to upset conditions.  Opacity can be limited to 20 percent during
normal operation, but may exceed 20 percent during boiler upsets.
     The collection efficiency of a mechanical collector can be
increased by adding a second stage, which is usually furnished
with higher-efficiency vanes to collect the fine dust not caught
                   9
in the first stage.   This second stage may add 5 percent to the
collection efficiency.
     Another method is to add a small multitubular collector,
designed to remove 10 to 20 percent of the gas flow from near the
                                                9
top of the hopper housing of the main collector.   This collector
removes fine and leafy char dust, which would normally be en-
trained and contribute greatly to stack opacity.  Collection
                                         9
efficiency is improved by 2 to 3 percent.
     With salt-laden particulate, the collection efficiency of
the mechanical collector is considerably lower because of the
fineness of the salt particles; overall efficiencies of 30 to 50
percent would be expected.  Test data from Weyerhauser's North
Bend plant confirm this observation.
     Mechanical collectors would be less efficient on stoker
boilers than on the Dutch-oven type because the stoker releases
finer particulate.   The Council of Forest Industries of British
Columbia (CFIBC)   believes that the efficiency of mechanical
collectors has been overstated and that they are not effective at
particle sizes below 30 micrometers.
                               52

-------
Design Factors--
     The most important design parameters that affect cyclone
efficiency are pressure drop,  particle size distribution, inlet
gas velocity, inlet dust loading,  cyclone body, and diameter
dimension ratios.  '
     Pressure drop—The pressure drop across a cyclone varies as
the square of the  gas volume and is directly proportional to the
density of the dust-laden gas.  The total pressure in a cyclone
consists of separate losses in the inlet flue, the cyclone body,
and the outlet duct.
     Flow rates must be kept within the design range.  If flow
rates are too low, the centrifugal force is not great enough to
separate the particles from the carrier gas.  If flow rates are
too high, energy is wasted in pressure drop across the unit and
                                                 12
the return vortex  configuration may be disrupted.
     Particle size—The size and density of particles control
their settling velocity.  It is the specific gravity of the
particle, not the  bulk density that is important.   Small par-
ticles with low specific gravity and thus low settling velocity
may not be able to reach the cyclone walls during  the brief time
the gas is in the  cyclone.
     Figure 3-6 illustrates a typical efficiency curve as a
function of particle size for both a conventional  cyclone and a
multiclone collector.  With a single cyclone, collection effi-
ciency falls off much more rapidly as particle size decreases,
than with a multiclone collector.   Because the diameter of the
multiple cyclones  is much smaller  than that of the single large
cyclone, the collection efficiency is greater, especially for
fine particles.
     Cyclone dimensions—A cyclone of higher efficiency and
higher pressure drop could be  designed by (1) increasing the
length of the cyclone, (2) decreasing the inlet width, or  (3)
increasing the ratio of body diameter to outlet diameter, while
                              53

-------
                      SINGLE LARGE
                        CYCLONE
              MULTIPLE  SMALL
                CYCLONES
                 20         40
               PARTICLE  SIZE, ym
Figure  3-6.   Relation  of particle  size to
    collection efficiency of cyclones.
                 54

-------
at the same time providing a smaller body diameter.  An increase
in the length of the cyclone body provides a longer residence
time for gas in the cyclone and therefore more revolutions.  An
increase in body length minimizes the loss of efficiency due to
reentraimrient in the ascending vortex.   Increasing the ratio of
body diameter to the gas outlet diameter effects an increase in
efficiency up to ratios of about 3 to 4, with relatively little
gains above that.  As the ratio increases, pressure drop also
increases.  Tables 3-1 and 3-2 summarize the effects of dimen-
sional changes in a cyclone collector.
Operation and Maintenance Problems - Cyclones--
     Operational and maintenance problems associated with cyclone
systems removing particulates from flue gas of hogged fuel boil-
ers are principally due to plugging, leakage of air into the
system, and changes in gas flow to the  cyclone.
     Plugging of the cyclone tube(s) occurs often, especially
when salt condenses on the collector surfaces.  Erosion of the
interior of the cyclone creates surfaces on which particles may
become anchored.  Particles continue to collect at this site
until the tube becomes fully plugged.  Plugging can also occur
when the collection hoppers are overfilled.  The particulate
material must be removed from the hopper at the same rate that it
enters or before the particulates can back up into the tube.
Plugging can also be caused by caking of the particles on the
tube walls because of particle size, adhesive properties of the
particles, and condensed moisture in the system.
     Cyclones can be operated as push-through or pull-through
systems.  Most hogged fuel boilers use  a pull-through system to
reduce abrasion on the induced-draft fan.  Since this system
operates essentially under vacuum conditions, there must be no
leakage of air into the system.  Leakage can cause reentrainment
of particles, loss of removal efficiency and possibly fire.
     The removal efficiency of cyclones is sensitive to changes
in flow rates, which can be significant in the hogged fuel
                               55

-------
TABLE 3-1.  EFFECTS OF CYCLONE DIMENSIONS  ON  PERFORMANCE AND COST
                                                                  13

Increase cyclone size
Lengthen cylinder
Increase inlet area,
maintain volume
Increase inlet area,
maintain velocity
Lengthen cone
Increase size of cone
opening
Decrease size of cone
opening
Lengthen clean gas out-
let pipe internally

Increase clean gas out-
let pipe diameter
Performance
Pressure loss
Down
Slightly lower

Down

Up
Slightly lower

Slightly lower

Slightly higher

Up


Down
Efficiency
Down
Up

Down

Down
Up

Up or down

Up or down

Up and/or
down

Down
Cost
Up
Up

No change

Down
Up

No change

No change

Up


Up
                               56

-------
         TABLE 3-2.  EFFECTS OF PHYSICAL PROPERTIES  AND
                PROCESS VARIABLES ON EFFICIENCY13

Gas change
Increase velocity


Increase density
Increase viscosity
Increase temperature
(maintain velocity)
Dust change
Increase specific
gravity
Increase particle size
Increase loadings
Pressure
loss
Up


Up
Neg .
Down

Neg .
Neg .
Neg .
Efficiency
Up


Neg .
Down
Down

Up
Up
Up
Cost
Initial cost
down
Operating cost
up
Slightly
higher
Neg .
Neg .

Neg .
Neg .
Neg .
Neg. = Negligible.
                               57

-------
boilers because steam demands are often highly variable.  This,
combined with the small particle size typical of salt-laden
emissions can cause drastic reduction of removal efficiency.
Electrostatic Precipitators (ESP's)
General System Characteristics--
     Although ESP's are used widely in controlling particulate
emissions from combustion sources, they are rarely used on
boilers fired with hogged fuel.  Several successful installations
on boilers that burn salt-free wood have been put into operation
in the last 5 or 6 years.  Figure 3-7 illustrates a typical ESP.
     From the precipitation standpoint, ash from hogged fuel
boilers poses two potential problems that must be accounted for
in design:    (1) since it is carbonaceous, the ash has a lower
resistivity than coal ash and tends to lose its charge quickly,
and  (2) the ash is prone to reentrainment because of its low
density, coupled with flake-like particle shape.  When the ash is
salt-laden, the submicron salt particles require a conservative
design, i.e., a larger collecting surface.
     Performance of ESP's on boilers burning salt-free hogged
fuel has reportedly been excellent.  Where an ESP was retrofitted
downstream of a cyclone at a paper mill, the average outlet
loading was 0.04 g/m  (0.018 gr/acf); guarantee was 0.06 g/m
(0.025 gr/acf). 14
     The only reference to performance of an ESP on a boiler
burning salt-laden hogged fuel comes from the Council of Forest
Industries of British Columbia.    Test data from a pilot unit
installed at the Victoria Sawmill Division of B.C. Forest Pro-
ducts, Ltd., showed that performance varied from excellent  [0.05
g/m3  (0.02 gr/dscf)]  to poor [0.7 g/m3  (0.3 gr/dscf)].4  The
program showed that the ESP could reduce salt fume to an accept-
able level when it was maintained and operated within precise
limits.  When control of operation was not virtually perfect,
however, the performance deteriorated markedly.
     No full-scale ESP's have yet been installed on boilers
                               58

-------
                                                     TRANSFORMER-RECTIFIER
      TOP END
       FRAMES    |

                c
   HIGH VOLTAGE   N
    CONDUCTOR
 HIGH TENSION
SUPPORT INSULATORS
       PERFORATED
      DISTRIBUTION
         PLATES
       BOTTOM END
           FRAMES
UPPER H.T.  HANGER ASSEMBLY
(HANGER AND HANGER FRAME)
        UPPER H.T. WIRE
         SUPPORT FRAME
                                                                       GROUND SWITCH BOX
                                                                       ON TRANSFORMER
                                                                            DISCHARGE ELECTRODE
                                                                            VIBRATOR
                                                                                COLLECTING ELECTRODE
                                                                                M.I.G.I. RAPPERS
                                                                                 TOP HOUSING
                                                                                 HOT ROOF
                                                                                 ACCESS  DOOR

                                                                                 HOT ROOF
                                                                                 SIDE
                                                                                 FRAMES
                                                                            ELECTRODE

                                                                            ACCESS  DOOR
                                                                            BETWEEN
                                                                            COLLECTING PLATE
                                                                            SECTIONS
                                                                             PRECIPITATOR
                                                                             BASE PLATE

                                                                             SLIDE  PLATE
                                                                             PACKAGE

                                                                             SUPPORT  STRUCTURE
                                                                             CAP PLATE
                                    HORIZONTAL
                                   BRACING STRUT
                                            STEADYING BARS
                                                                     LOWER H.T.
                                                                  STEADYING FRAME
                                                                     COLLECTING
                                                                     ELECTRODES
      Figure  3-7.   Typical  electrostatic  precipitator  with  top  housing
                         (courtesy  of  Research  Cottrell,  Inc.).
                                                 59

-------
 burning  salt-laden  hogged  fuel.  Weyerhauser Co.15 has  reported
 that  ESP manufacturers  will  guarantee  outlet grain loading but
 not visible  opacity in  these applications.  Communication with
 a major  ESP  manufacturer indicates  that  the refusal to  guarantee
 visible  opacity  is  because salt content  varies with season and
 the salt particles  are  difficult to  control; in-stack opacity,
 however, would be guaranteed.  This  contact indicated that his
 company  would guarantee visible opacity  for an application with
 salt-free  hogged fuel.
 Design Factors--
      The three most important parameters that affect the design
 and subsequent performance of an ESP are (1) the gas volumetric
 throughput in m /min (acfm),  the total plate collection area in
 22                                22
 m   (ft ),  and power density  in watts/m   (watts/ft ) of collecting
 area.
      With  knowledge of  the total gas throughput, which  is dic-
 tated by the boiler firing rate, and the specific collection area
           23          2
 (SCA) in m /m  per  s (ft /1000 acfm), one can determine the total
 area  required for compliance with an emission standard.  Equa-
 tions for  determining the  required SCA are given below  in English
 units:
         SCA =  16.67 In2(l-n)
                     Wk
         n  =    1  - c  /c.
                      o  i
      Where
      n   =  Overall  mass collection efficiency, percent
      w,   =  Modified migration velocity,  ft/s
      c   =  Allowable outlet grain loading gr/dscf
      c.   =  Inlet grain  loading, gr/dscf
     Migration velocity—The modified migration velocity is a
function of electrical  energization of the precipitator and of
gas properties.   It  is  often conveniently linked with resistivity
level.  A typical migration velocity range in boiler applications
                               60

-------
with salt-free wood fuel is 0.6 to 1.5 cm/s (0.2 to 0.5 ft/s).
With salt-laden fuel,  the migration velocity would be in the
lower portion of the range.
     A digression is needed here to clarify the usage of w
(modified migration velocity)  in contrast to the effective
migration velocity w,  which is used in the conventional Deutsch-
Anderson efficiency equation.     The effective migration velocity
w is a function of several factors, including precipitator
length, overall mass collection efficiency, and gas velocity.
The variation in w within a given precipitator is caused by
changing particle size distribution as precipitation proceeds in
the direction of gas flow.
     The modified migration velocity,  w ,  as presented by Matts
             18
and Ohnfeldt,   can be treated essentially as a constant for any
application.  It is, of course, strongly dependent upon the inlet
particle size distribution.
     Power input—The  third basic design parameter is power
density required to establish the optimum voltage-current char-
acteristics of the corona, given the dust entering the precip-
itator.  Power density is a function of electrical resistivity,
particle size characteristics and distribution, gas loading and
composition, gas temperature,  and gas pressure.  It is often
conveniently linked with resistivity,  such that for a moderate
                 9
resistivity of 10  ohm-cm, the value will be approximately 27
       2              2
watts/m  (2.5 watts/ft ).  For boilers burning salt-free fuel
the range of power input is 11 to 22 watts/m   (1 to 2 watts/ft2).
This power density range would not change significantly with
salt-laden fuel.
     The selection of  power density is often conveniently based
on resistivity of the  dust.  Table 3-3 illustrates a general cor-
relation between power density and dust resistivity; this rela-
tionship would also apply to fly ash from coal firing.
                               61

-------
                TABLE 3-3.  DESIGN POWER DENSITY
Resistivity, ohm-cm
lO4"7
lO7'8
io9-10
10+11
io+12
>io+13
Power density,
watts/m2 (watts/ft2)
of collecting plate
43 (4.0)
32 (3.0)
27 (2.5)
22 (2.0)
16 (1.5)
< 11 ( < 1.0)
     Field voltages and current densities for salt-free hogged
fuel boilers range from 40 to 45 kV and 0.22 to 0.65 mA/m  (0.02
             2
to 0.06 mA/ft ), respectively.  These values are not constant for
each point in the precipitator.  At the inlet section where the
dust loading is greatest, the voltage-current characteristics
differ significantly from those at the outlet.
     Although it appears that resistivity plays a significant
role in selection of w  and power density, there is no precise or
                      .K
universally applied method for predicting resistivity on the
basis of the material entering the furnace.
Operation and Maintenance Problems - ESP's--
     Only a few ESP's are applied to hogged fuel boilers, and
none have been installed on boilers burning salt-laden hogged
fuel.  Data on operating problems are sparse.  One possible
problem is the possibility of fire from "char" on precipitator
walls (buildings) and especially in the hoppers.  The fire hazard
can be minimized by installing trough-type hoppers to remove dust
continuously.  Elimination of in-leakage will decrease the
availability of air for combustion.
     Reentrainment is also a problem with bark ash because of its
low resistivity.
                               62

-------
     Specific design parameters for application of an ESP to
hogged fuel firing are presented in Table 3-4.
Net Scrubbers
General System Characteristics--
     Wet scrubbers are applied to a number of hogged fuel boilers
to control particulate emissions.  Most are installed downstream
of a multiple cyclone collector.  In salt-free applications under
normal operating conditions the outlet loadings range from 0.05
to 0.14 g/m3 (0.02 to 0.06 gr/scf)  at pressure drops of 15 to 38
                       19
cm (6 to 15 in.) water.
     Only one boiler fired with salt-laden hogged fuel is currently
equipped with a venturi scrubber, Crown Zellerbach, Port Townsend
plant; this scrubber is installed downstream of a multiclone.
The reported outlet loadings are 0.16 to 41 g/m   (0.07 to 0.18
gr/scf), depending on the salt content of the fuel.  A salt
concentration greater than 1 percent will limit the performance
of this venturi scrubber, which operates at a pressure drop of 38
to 51 cm  (15 to 20 in.)  water.  The salt content of particulate
emissions from the scrubber is reported to be 50 to 70 percent or
more.  Opacity of the stack effluent is estimated at 35 percent.
This scrubber would require a pressure drop greater than 51 cm
(20 in.)  water to achieve the roughly 80 percent efficiency
needed to meet the 0.23 g/m   (0.10 gr/dscf) particulate emission
regulation, assumming a typical inlet concentration of 1.14 g/m
(0.50 gr/dscf), and greater than 1 percent salt in the fuel.
Design Factors--
     This discussion is limited to venturi scrubbers since they
can remove greater amounts of fine particles than other types of
scrubbers when pressure drops are high enough.  This is a prime
consideration in removal of submicron salt particles from hogged
fuel boilers.
     In conventional terminology, the venturi scrubber is cate-
gorized as a gas atomized spray scrubber.  A flooded disc type of
venturi scrubber is illustrated in Figure  3-8.  The collection
process relies mainly on the acceleration  of the  gas stream  to
                               63

-------
       TABLE 3-4.  DESIGN PARAMETERS AND DESIGN CATEGORIES
                   FOR ELECTROSTATIC PRECIPITATORS16
Dust composition

NaCl      :
C  (Char)
Sand
Fly ash

Precipitator capacity

No. of precipitators
No. of chambers  (units)/precipitator
No. of ducts/chamber  (unit)
Duct spacing
Plate height
Treatment length
Section lengths and total number of each  (per precipitator)
Collecting area
No. of electrical sections parallel to gas flow  (per precipitator)
No. of electrical sections across gas flow (per precipitator)
No; of hoppers parallel to gas flow (per precipitator)
No. of hoppers across flow (per precipitator)

Rapping, electrodes, etc.

Type discharge electrode
Meter (feet) discharge electrode/vibrator or rapper
Type discharge electrode vibrator or rapper
Type collecting electrode
Square meter (square feet) collecting electrode/rapper
Type collecting electrode rapper

Electrical energization   (of each electrical section)
       2          2
Watts/m   (watts/ft ) of collecting electrode
Square meter (square feet) of collecting electrode/T-R
Mode (switching)
Corona kilovolts   ~            _
Milliamperes/1000 m   (mA/1000 ft ) of collecting electrode
Milliamperes/T-R set

Performance-related parameters

Gas flow                      Overall mass collection efficiency
Gas temperature               Fractional mass collection efficiency
Gas (treatment) velocity      Inlet grain loading
SCA                           Outlet grain loading
                                64

-------
                     GAS  FLOW OUT
GAS FLOW IN
Figure 3-8.  Research-Cottrell flooded
        disc venturi  scrubber.

                 65

-------
provide impaction and intimate contact between the particulates
and fine liquid droplets generated as a result of gas atomization.
This is a high-energy-consuming device designed for substantial
particulate removal.  A pressure drop of 51 cm (20 in.), or
greater is required for salt-laden particulate removal, based on
current operating data.  The collection efficiency of venturi
scrubbers increases with increasing pressure drop and liquid-to-
gas (L/G) circulation rate.  There is, however, an optimum L/G
ratio above which additional liquid is not effective at a given
pressure drop.  The pressure drop can be increased by increasing
the gas velocity.
     In the design of venturi scrubbers, the follow key parameters
affect particulate collection:
     (a)  Gas velocities and gas flow rates
     (b)  Particle size distribution
     (c)  Pressure drop
     (d)  Liquid-to-gas ratio.
     The following information is also required to justify the
choice of such equipment.
     (a)  Gas handling capacity per module
     (b)  Total number of modules required
     (c)  Capital investment
     (d)  Annual costs
     (e)  Water requirement; water recirculation
     (f)  Availability of the equipment or necessary downtime
     (g)  Indication of fractional collection efficiency of the
          device
     (h)  Total power consumption as a fraction of the generated
          power.
     Velocity/gas flow rate--Sizing a venturi scrubber is often
based on the velocity and flow rate of the inlet gas.  Usually,
the inlet gas velocity is about 18 m/s  (60 ft/s), and the inlet
flow rate'depends on the boiler size.  Typical scrubber diameters
are under 3 m (10 ft).  If the gas flow rate is too high for one
scrubber, several scrubbers should be used.
                               66

-------
     Liquid-to-gas (L/G)  ratio—The L/G ratio typically ranges
from 0.13 to 2 m3/1000 m  (1 to 15 gal/1000 acf)  and is basically
a function of inlet gas temperature,  inlet solids content, and
                             14
method of water introduction.     At higher inlet gas tempera-
tures, evaporation of the scrubbing liquor may occur at the point
of contact.  Where the inlet dust loading is heavy, the L/G ratio
should be increased to minimized buildup of solids and plugging
of drains. Although pressure drop across the venturi is essential-
ly independent of a specific design,  the less efficient methods
of water introduction will require additional scrubbing liquor to
meet efficiency requirements.
     Pressure drop—The scrubber design for a given application
requires careful selection of  the throat velocity and L/G ratio
to achieve the maximum collection efficiency for the energy
spent.  The energy spent is often indicated by the gas pressure
drop across the scrubber, which ranges typically from 25 to 51 cm
(10 to 20 in.) of water for venturi scrubbers on hogged fuel
boilers.    To achieve a given pressure drop, numerous relation-
ships between throat velocity  and L/G ratio can be used.   Only
one set of conditions will yield the  maximum efficiency for the
energy spent.  That one set of conditions is the only one that
creates maximum droplet surface with  minimum L/G ratio during
atomization.
     Particle size distribution—The  particle size distribution
in the inlet gas stream is a key factor in scrubber design.
However, the particle size distribution in effluent from a hogged
fuel boiler often varies with the percentage of salt in the fuel
and with boiler operating conditions.  Although the efficiency of
particulate collection increases with increasing pressure drop,
the inlet particle size distribution  will determine the gas and
liquid velocities required to  achieve the desired overall mass
collection efficiency-
     Data on fractional collection efficiency for  submicron
particles is particularly difficult to obtain.  When one  is
                               67

-------
speaking of greater collection efficiencies, it should be clear
that this refers to fractional collection efficiencies in the
submicron particle size range.
     Materials selection—Firing of salt-laden fuel with supple-
mental oil-firing can cause corrosion by abrasive solids.  Suit-
able construction materials such as stainless steel must be
selected.  Where abrasion-resistance requirements exceed the
limits of stainless steel, fiberglass-reinforced plastic (FRP, or
polyester) may be used.  Abrasion-resistant liners are also
needed to withstand high temperature.
     Mist eliminator—Mist eliminators are needed to control
undesirable emissions of liquid droplets from scrubbers caused by
atomization and carryover of liquid.  Because of the solids in
the scrubber liquor, the entrainment of water droplets can cause
operating problems and liquid losses.  Suspended or dissolved
particles can cause buildup of solids, and suspended solids can
cause erosion.  Among the many problems caused by buildup is the
increase in pressure drop.
Operation and Maintenance Problems - Venturi Scrubbers--
     Common maintenance problems for venturi scrubbers include
plugging of lines, nozzles, and pumps and erosion/corrosion and
scaling of internal components.  In operation of Georgia-Pacific
                                                     18
Corp. wet scrubbers on salt-free hogged fuel boilers,   the pro-
blems are similar to those in other applications.  High suspended
solids content due to inadequately sized clarifers has led to
deposition in horizontal transfer lines, drain weirs, and quies-
cent areas of the scrubber.  The abrasive solids have accelerated
the wear of pump impellers and spray nozzles.
     Evaporation plus leaching of mineral salts  (calcium, iron,
and other compounds) from entrained particulate has caused build-
up of dissolved solids in the recycled scrubber water, with
resultant scaling and sludge problems.
     Unexplainable changes in pH have caused corrosion of mild
steel.
                               68

-------
     Operators of some Georgia Pacific plants have also cited
foam generation as an operational problem, especially with venturi
scrubbers.
     Where such problems are not caused by inadequate design,
they have been overcome by changes in operation and by maintain-
ing the quality of water used in the.scrubber within narrow
limits.
     Crown Zellerbach's Port Townsend venturi scrubber is the
only one installed on a boiler burning salt-laden hogged fuel.
Plant engineers characterize the operation of this scrubber as
satisfactory.   Some initial problems such as erosion/corrosion of
the distribution header, erosion of the fiberglass separator, and
vibration of the separator assembly were quickly resolved.  The
biggest operating problem at this installation occurs when salt
content of the fuel exceeds 1 percent, causing reduced efficiency
(=70%)  of the  scrubber.
Fabric Filters
General System Characteristics--
     Fabric filters are used on two hogged fuel boiler instal-
lations:  Simpson Timber Co. in Shelton, Washington, and Long
Lake Lumber Co. in Spokane, Washington.  A third baghouse system
has been purchased but not yet installed at Georgia-Pacific's
Bellingham, Washington, mill.  The Simpson and Georgia-Pacific
mills use logs that have been stored in seawater.
     Plant engineers estimate the overall mass efficiency of the
Simpson Timber Co. baghouses at 90 to 95 percent,  stating that a
number of bags are always broken, due to improper installation,
faulty construction, etc.
     Outlet mass loadings were less than 0.05 g/m   (0.02 gr/scf),
and opacity did not exceed 3 percent  (mostly only heat waves were
visible).   The Long Lake facility reports an overall mass effi-
ciency of'99 percent.
     Although  no fractional efficiency data are available, the
dramatic reduction in opacity with installation of the baghouses
indicates that they are efficient collectors of submicron salt
                               69

-------
particles.  Fractional efficiency data from tests of baghouses in
other industries support this view.
Design Factors—
     Fabric filters are basically simple.  The removal of parti-
culates from gases is accomplished by forcing the gases to flow
through the fabric, which removes the particulates by one or more
of the following mechanisms:
      (1)  Inertial impaction
      (2)  Diffusion to the surface of an obstacle by Brownian
          diffusion
      (3)  Direct interception because of finite particle size
      (4)  Sedimentation
      (5)  Electrostatic phenomena.
     Parameters important to fabric filtration system design
include air-to-cloth ratio, pressure drop, cleaning mode and
frequency of cleaning, composition and weave of fabric, degree of
sectionalization, and gas cooling.  Baghouses are relatively
insensitive to process variables such as chemical composition of
the gas (providing the correct bag fabric is chosen), particle
size, electrical resistivity, etc; thus there tends to be very
little substantial design difference among baghouses with the
same cleaning mechanism, regardless of the application or manu-
facturer.   Differences that are noted generally relate to main-
tenance (e.g., number of bag rows accessible from an interior
walkway, method of bag cuff attachment to cell plate).  Table 3-5
summarizes design parameters for the baghouses on two existing
and one proposed salt-laden hogged fuel boiler.  These parameters
are briefly discussed below.
     Air-to-cloth ratio/pressure drop--Air-to-cloth ratio  (A/C)
is the ratio of volume of gas filtered  (m /s or acfm) to the
                           2      2
available filtering area (m  or ft ).  The A/C ratio affects the
pressure drop across the filter, i.e., a higher ratio yields a
higher pressure drop.
     The ratio of air to cloth becomes an economic  tradeoff in
baghouse design because the higher the A/C ratio, the smaller the
                               70'

-------
            TABLE 3-5.  DESIGN PARAMETERS FOR BAGHOUSES

                     ON HOGGED FUEL BOILERS

Volume flowrate, acfm
Inlet gas temperature,
Op
A/C ratio, acfm/ft
Bag cleaning method
Pressure drop, in. HO
Bag fabric
Precollector
Material handling
system
Simpson
Timber
130,000
500
4.5
Pulse jet
9-9.5
Teflon- coated
fiberglass
Mechanical
cinder
collector
Screw
conveyor
Long Lake
Lumber
25,000
400
4.0
Pulse jet
5.8-6.8
Nomex
None
Screw
conveyor
Georgia-
Pacific
180,000
440
4.0
Pulse jet
a
Teflon- coated
fiberglass
None
Screw
conveyor
Metric conversions:  acfm x 0.028 = m /min

                            2           ^
                     acfm/ft  x 5.09 = m /s per m


                     (°F - 32J/1. 8 = °C


                     in. H20 x 2. 54 = cm HO
  Collector not yet installed.
                               71

-------
amount of filter fabric required.  Offsetting this advantage,
however, are the requirements for a larger fan and a cleaning
mechanism of higher energy, leading to a reduction in bag life.
Equipment suppliers determine A/C ratio by considering the char-
acteristics of the dust to be collected, the type of fabric, the
cleaning mechanism, and the specified operating pressure drop.
     The size range of a significant portion of the particulate
emitted from boilers burning salt-soaked hogged fuel is from less
than 1 to 10 ym.  Thus enough of the larger particles are present
to allow use of a fairly high A/C ratio, limited by the amount of
fines also in the mix.  Operation at higher ratios may lead to
excessive pressure drops, as at Simpson Timber Company's baghouse
installation.  The pressure drop is 8 cm (3 in.) over design,
probably because the particle size distribution is skewed toward
the submicron range.  Operating pressure drops for hogged fuel
boiler baghouses range from 15 to 30 cm  (6 to 12 in.)  water, and
it is preferable to operate at the lower end of this range.
     Fabric selection--Selection of a suitable fabric is a func-
tion of temperature, abrasion resistance, acid resistance, and
cost.  The main function of the fabric is to provide a rigid
filtering medium for formation of the initial dust cake, which
then acts as a fine filter that can achieve high efficiencies
(>99%) in collection of submicron particulates.
     Bag fabrics are woven or felted.  Felt is a genuine filter
medium and is more efficient than woven fabrics in collection of
the dust; it is also more expensive.  Bags made of fiberglass
coated with Teflon or silicon graphite, woven Teflon, or filtered
Teflon may be used in the high temperature range  (150° to 260°C,
(300° to 500°F).  Fabrics for high temperature operation are
costly.   Table 3-6 lists some of the fabrics available for low
and high temperature operation and the characteristics of each.
Normal life expectancy of bag fabric is 1 to 3 years in conven-
tional operations.
     The two Simpson Timber units contain Teflon B-coated fiber-
glass bags for operation at 260°C (500°F).  The Long Lake baghouse

                               72

-------
                  TABLE  3-6.   CHARACTERISTICS  OF  VARIOUS  BAGHOUSE  FABRICS
                                                                                         22
Fiber
Cotton
Wool
Nylond
Orlond
Dacron
Polypropylene
Nomex
Fiberglass
T«flond
Operat inq
exposure ,
0 p
Long
180
200
200
240
275
200
425
550
450
Short
225
250
250
275
325
250
500
600
500
Supports
combus-
tion
Yes
No
Yes
Yes
Yes
Yes
No
Yes
No
Air
perme-
abi 1 i tya
10-20
20-60
15-30
20-45
10-60
7-30
25-54
10-70
15-65
Composition
CP 1 1 ulose
Protein
Pol yamide
Polyacrylonitrilo
Abra-
s i on'3
G
r,
E
G
Polyester E
Olef in
E
Polyamide E
Glass P-F
Polyfluoroethylene
F
Resistance to
Minera 1
acidsb
P
F
P
G
G
E
F
E
E
Organic
acids0
G
F
F
G
G
E
E
E
E
Alkalib
G
F
G
F
G
E
G
P
E
Cost
rankc
1
7
2
3
4
6
8
5
9
--J
OJ
          Metric conversions:   ("F  -32)/1.8  =  °C
                                ft /min  per  ft   x  5.09  =  m  /s  per  m
          a ft3/min per  ft2  at  0.5  in.  W.G.
            p = poor, F  =  fair,  G =  good, E  =  excellent.
          c Cost rank on a scale of  1  (low)  to 9  (high).
            DuPont registered trademark.

-------
is equipped with Nomex bags and operates at about 210°C  (410°F).
The new unit at Georgia-Pacific will use Teflon-coated fiberglass
for operation at 227°C  (440°F).
     Cleaning mechanisms--The most common methods of dislodging
the built-up dust layer are mechanical shaking, reverse air, and
pulse jet.
     During shake cleaning, the pressure to the bags is reduced
by turning off the fan and a gentle motion is applied to the top
bag attachment (see Figure 3-9).  When bags are made of fiber-
glass, the cleaning mechanism must be very gentle to prevent
damage.  Shaking with a period of 50 cycles per minute and about
                                           23
5 percent of the bag length is recommended.
     Reverse air and shaking are equally effective, but pressure
drop across the clean filter may increase to the point that both
methods are required to bring the pressure drop down to an
acceptable level.
     The pulse jet cleaning mechanism is relatively new.  A
sudden blast of compressed air is injected through a nozzle
installed at the top of each bag (see Figure 3-10).  The bag
expands to its maximum diameter at the top, and the expansion
travels down the bag, throwing the dust from the outside of the
bag.  This type of cleaning mechanism is more energy-intensive
than the shaker or reverse air mechanisms and usually entails
higher A/C ratios and pressure drops.
     Pulse jet is the preferred cleaning mechanism for all three
hogged fuel boiler baghouses listed in Table 3-5.  It allows for
maximum filtering time with minimum interruption of the filtering
flow and allows for a slightly smaller baghouse.
Operation and Maintenance Problems - Fabric Filters--
     Since only two U.S. hogged fuel boiler installations use
fabric filters to control particulate emissions, information on
maintenance problems and practices is very limited.  Simpson
Timber Co., which operates a fabric filter on a boiler fired with
salt-laden hogged fuel, provided some maintenance data.
                               74

-------
SHAKER
MOTOR
               REVERSE AIR AND CLEAN AIR PLENUM
                                                         LEAN AIR
     Figure  3-9.   Reverse air  or  shaker  type,
                           75

-------
DIRTY AIR
                                                      CLEAN AIR
                                                     COMPRESSED AIR
             Figure 3-10.   Pulse jet type
                           76

-------
     The primary problem with operation and maintenance of the
two collectors at Simpson Timber centers on the collection hop-
pers.  The very light collected dust has caused plugging.  Even
though the screw conveyor can adequately handle the volume of
material emptying from the hoppers,  the material tends to bridge.
Much of the dust is submicron NaCl particles.   To prevent plug-
ging, Simpson Timber is trying out a hopper vibrator system that
is actuated at the end of each cleaning cycle.   Although this
system appears to be relieving the plugging, personnel also feel
the need for a good method of sensing when the  hoppers begin to
plug.  They have not yet devised a reliable mechanism.
     Simpson also has installed a bypass chute  on the baghouse
hopper for use during cleanout.  The screw conveyors could not
handle the large volume of material released when a plugged
hopper was dislodged, so the chute was provided to relieve the
extra load.
     Except for some operational problems during the first few
months of operation, the two baghouses have run satisfactorily
for over 2 years.  Routine maintenance is, of course, performed.
Every 3 to 6 months, when the system is off line, the fan housings
and  impellers are cleaned by blowing them out with compressed
air.  Vibration detectors have been installed on the fan to warn
of potential imbalances.  This system allows maintenance personnel
to clean the fans before a scheduled inspection if needed.
     Bag life is reported to average around 15 months.   The usual
cause of bag failure is abrasion against the support cage.
Ultimate Disposal of Salt-laden Ash Collected by Conventional
Control Equipment
     Although particulate control devices remove particulate  from
stack gases, the material that is removed must be disposed  of
by one of  several methods referred to  as ultimate disposal.   The
waste products  are basically of two types,  a dry ash  and a  wet
slurry.  The dry ash comes from cyclones, baghouses,  and ESP's,
and  the wet slurry from wet  scrubbers  and ESP's.  Many techniques
                               77

-------
are available for ultimate disposal of such wastes, but the
presence of salt in the waste material' (as much as 90%) compli-
cates disposal and limits the choice of ultimate disposal methods.
     Where the ash contains little or no salt, it can be disposed
of by several methods.  The volume of the ash can be reduced by
reinjecting it from a cyclone collector into the boiler for a
                      24
more complete burning.    The most common method is to place the
ash in a sanitary landfill.  Under a properly controlled landfill
scheme this can be an effective method of disposal.  When a high
proportion of the ash is salt, however, the landfill operators
must maintain strict control to prevent leaching of the salt into
local ground and surface waters.
     Several important factors should be considered in selection
and design of a landfill.  First, upon preliminary selection of
the landfill site, the local geology and groundwater conditions
must be studied thoroughly as a basis for selection of techniques
and materials that will ensure isolation of the landfill and its
leachate from the local environment.  Where the groundwater table
is high, for example, it may be necessary to lower it locally by
dewatering techniques.  Where the soil is highly permeable (e.g.,
sand or loam), an impermeable barrier must be constructed to
ensure against groundwater contamination.  This is usually done
by compacting a clay layer (about 5 feet thick) that allows
movement of water through it, but at a controlled rate.  The clay
also filters out the leachate as it passes through.  Other types
of barriers used with varying degrees of success are asphalt and
concrete, which are subject to differential settling and cracking,
and plastic, which is subject to puncture and is not entirely
resistant.  Ideal sites are those where the soil is mainly clay
or where the underlying strata are impervious.
     The market for the resale of dry ash is increasing, par-
ticularly for use in road construction and in the asphalt and
concrete industry.  Fly ash is being used to stabilize roadbeds
(as a sub-base material), as an additive in asphalt, and as an
                                78

-------
aggregate for concrete when converted to pellets.   Again, however,
the salt fraction would severely limit any usefulness of the bark
ash for such purposes.
     Similarly,  although some ash is now being applied to the
soil as fertilizer,  ash that contains salt cannot be used for
this purpose.
     Methods for disposal of wet slurries containing fly ash and
dissolved gases  vary widely depending upon the contents of the
slurry.  When slurry is stored in lagoons the solids fraction
often will settle out, given enough time.  The supernatant is
then drawn for reuse in the slurry system, for further treatment,
or for disposal  to the environment.  When the slurry contains
salt or other potential pollutants, the lagoon can be properly
designed and constructed to ensure isolation of slurry from the
environment.  Sludge from the lagoons can be pumped out and sent
for further treatment, landfilled, or incinerated.  Because
solids content of the sludge from lagoons and clarifiers is
usually under 10 percent, additional treatment is usually needed
to make it disposable.  The physical and chemical sludge treat-
ment methods are similar to those used in treating municipal
sludges, that is, placement in drying beds,  centrifugation,
pressing, vacuum filtration, and processing for fertilizer.
     Ultimate disposal of the dry cake is similar to that for dry
ash.  Again it is emphasized that where the ash or sludge cake
contains salt or other potential pollutants, the disposal method
must ensure protection of the environment.
     In installations currently firing salt-laden hogged fuel the
primary methods  of ash disposal are discharging to municipal
treatment facilities, landfilling, or reinjection of the ash into
the boiler.  Ash from the fabric filters at the Simpson Timber
plant in Shelton is mixed with the continuous boiler blowdown and
sent to the municipal sewage treatment plant at a rate of about
20 gal/min.  Because the suspended solids in this slurry account
for approximately two-thirds of the total suspended solids entering
                               79

-------
the  treatment  plant,  Simpson  Timber must pay  a  surcharge.   Requests
by Simpson  Timber  for approval  to dump  the ash  into the bay were
rejected  because even though  the salt would have no detrimental
effects,  the ash is considered  a pollutant.   Landfilling of the
material  was also  disapproved because of potential leaching of
the  salt  into  ground  water.
     Crown  Zellerbach's Port  Townsend Plant treats the slurry
from the  scrubbing system along with the process wastewater
stream.   This  treatment system  consists of a  primary clarifier
and  aerated lagoons.   Sludge  from the clarifier is sent to  a
lagoon for  aeration and settling.  The clarifier supernatant and
the  lagoon  effluent are then  discharged to the environment.  Ash
from the  mechanical collectors  and the  sludge from the aeration
lagoon are  then disposed of in  a landfill.
     Weyerhauser's North Bend plant reinjects fly ash in essen-
tially a  closed system.  Diverting the ash from reinjection was
      25
tried,    but yielded  no significant reduction in emissions  and
was  discontinued.
     In summary, there is no  standard method  for ultimate dis-
posal of  fly ash with high salt content.  The salt content  pre-
cludes reuse of the fly ash in many applications, and the salt is
too  impure  to  be sold or returned to the sea.  Although landfill-
ing  is the method  most often  used for disposal, not all landfills
can  accept the ash because of severe handling and/or leaching
problems.  Ash in  slurry form,  however, can be mixed and treated
with other process wastestreams or with municipal wastewater.
The  ash and salt are  thus diluted to levels at which the material
can be handled or  treated efficiently and economically.  Wet
treatment therefore seems to  be the more effective means for
ultimate disposal  of  ash.
Best Available Conventional System for Controlling Salt
Emissions from Hogged Fuel Boilers
     Operating data on boilers  fired with salt-laden hogged fuel
indicate that the  best available conventional control system is
a mechanical collector followed by a fabric filter.  A mechanical
                               80

-------
collector is needed to remove cinders that would increase the
chance of fire in the fabric filter.   Table 3-7 gives available
performance data for conventional secondary collectors on salt
particulate emissions.
     Simpson Timber Company's fabric  filter installation at
Shelton has operated successfully for over 2 years in compliance
with both grain loading and opacity regulations.  The dramatic
decrease in opacity is evidence of the ability of the fabric
filter to capture submicron particles of salt.  This ability to
maintain high levels of collection efficiency in the particle
size range from 0.1 to 10 micrometers has been demonstrated in
fabric filter applications in other industries.  Some maintenance
problems were encountered but plant officials do not regard them
as excessive.
     Use of a venturi scrubber on the salt-laden hogged fuel
boiler at Crown-Zellerbach1s Port Townsend mill has been partial-
ly successful in that operations comply with the applicable
emission regulation when the salt content of the fuel is below 1
percent.  At higher salt levels the scrubber cannot remove enough
of the additional fine salt particles to achieve compliance.
More efficient removal of submicron salt particles would require
additional pressure drop over the present maximum of 51 cm  (20
in.) water.  Opacity has not been measured but is estimated by
plant officials at 35 percent.  Plant engineers report that the
Port Townsend scrubber has not required excessive maintenance
although trouble areas may become apparent with more operating
time.
     Electrostatic precipitators have not been applied to boilers
burning salt-laden hogged fuel, probably chiefly because ESP
manufacturers will guarantee outlet grain loading but not visible
opacity.  For compliance with visible opacity regulations when
the boiler fuel is salt-laden, an ESP may be too large to be
economically competitive with a fabric filter or wet scrubber.
Sizing of the ESP would have to accommodate the "worst case" in
terms  of salt content in the fuel.

-------
                       TABLE  3-7.    PERFORMANCE OF  CONVENTIONAL  SECONDARY COLLECTORS
                              ON  SALT-LADEN  PARTICIPATE  FROM  HOGGED  FUEL BOILERS

Source
Simpson Timber Co.
Shelton, Wash.
Crown Zellerbach
Port Townsend, Wash.


B.C. Forest Products
Victoria, British
Columbia

Control
device
Fabric
filter
Von tun
scrubber


Pilot
ESP


Date
of
test (s)
4/76
2/78

5/78 and
7/78
9/75 to
1/76

Inlet
loading ,
g/smJ
(gr/scf )
NA
NA

1.28
(0.56)
0.53-2.04
(0.23-0.89)

Outlet
loading ,
g/sm3
(gr/scf )
n . 02 and n. 01
(0. 01 and 0. 04)
0.16
(0.07)
0.40
(0.17)
0.11-0.46
(0.05-0.20)


Efficiency,
*
90-95
NA

"69b

43-95



Visible
opacity,
H
• 5
NA

= 35

NA


Salt
content
in fuel,
%
NA
0.4

>1

NA


Salt
content
in flue
gas, %
~70
= 50

= 70

=35-81



Reference (s)
26 , a
20

c

11


00
NJ
              Personal communication with Mr.  Robert lloit, Simpson Timber Co.,  September 197B.

              Approximate efficiency; inlet sample taken on 7/78; outlet sample taken on 5/78.

            c Personal communication with Mr.  Alan Rosenfeld,  Crown Zellerbach  Environmental Services, August 1978.

            NA -  Not available.

-------
     Appendix C presents two case histories of conventional
secondary collectors applied to salt-laden hogged fuel boilers.
(Simpson Timber Co.  in Shelton and Crown Zellerbach in Port
Townsend).

NOVEL FINE PARTICULATE CONTROL DEVICES
     Experience at plants with salt particulate emissions shows
that, except for fabric filters,  the conventional control devices
cannot consistently reduce emissions enough for compliance with
particulate and stack opacity regulations.  Thus, there is a need
for more effective control, especially in the submicron particle
size range.  This section evaluates some novel or promising
particulate control devices now being tested for control of salt
emissions from hogged fuel boilers.  Detailed descriptions of the
operation,  performance, and costs of each applicable novel
device are presented in Appendix D.
Classification of Novel Control Devices
     Three categories of wet scrubbing   are considered promising
for control of the submicron salt particles from hogged fuel
boiler emissions:  foam scrubbing, flux force condensation, and
electrostatic scrubbing.  Of these, electrostatic scrubbing
appears to offer the greatest potential for full-scale applica-
tion to salt-laden hogged fuel boilers.
     Dry scrubbers also are considered here as a novel device,
since there have been two pilot tests and one full-scale test
on three salt-laden hogged fuel boilers.
                2 8
Foam Scrubbing--
     In foam scrubbers the foam is generated by forcing aerosol
gas through a screen sprayed with a surfactant liquid.  Particle
collection is believed to take place mainly by diffusion and
sedimentation, mechanisms that are predictable and rather well
understood.  Application of this method has shown that the
initial cost is higher than that of the most expensive conven-
tional method, reliability is undetermined, and particulate
collection efficiency is not high.
                               83

-------
Flux-force/Condensation—
     Flux-force/condensation effects accompany the condensation
of water vapor from the gas and are generally caused by contact-
ing hot, humid gas with colder liquid or by injecting steam into
saturated gas.  The transfer of water vapor toward the cold
liquid surface sweeps particles with it and is referred to as
diffusiophoresis.  Heat transfer from the gas to the liquid also
causes particle movement toward the cold liquid, called ther-
mophoresis.  Condensation of water on the suspended particles
causes an increase in mass  (particle plus condensate), referred
to as particle growth.  Collecting the particles by inertial
                                                          29
impaction is easier after they have grown by condensation.
     A pilot study of this method showed that initial costs and
                              29
operating costs are very high.    The water vapor plume contri-
butes to higher opacity, which is undesirable.
Electrostatic Scrubbing--
     This device embodies the principles of an electrostatic
precipitator and a scrubber.  The basic idea is to augment the
collection processes associated with spray scrubbing and elec-
trical collection forces.  It involves the use of electrostatical-
ly charged water droplets or charged pollutant particles or
both.  '    The scrubber may be of the spray or venturi type.
Particulate collection efficiencies higher than 99 percent have
                                31 32
been achieved with some devices.  '    The initial costs are
somewhat higher than those associated with conventional methods,
but lower than those of the foam and force-flux/condensation
scrubbers.  Operating costs are reported to be much lower than
those of conventional methods.  '  '    It appears that devices
based on electrostatic scrubbing would provide good control over
opacity of stack emissions and would be more effective than the
other novel devices in controlling particulate emissions.
Further testing is required to evaluate electrostatic scrubbing
for control of salt-laden particulate emissions from hogged fuel
boilers.
                               84

-------
     Table 3-8 summarizes characteristics of five novel control
devices that operate on a principle of electrostatic augmentation.
Appendix D presents more detailed descriptions of these devices.
Dry Scrubbers--
     The dry scrubber is a recently developed system that uses a
moving bed of granular material (media) instead of water droplets
to capture particulates.  The dirty media is shaken at the bottom
of the unit, and particulates fall into a storage bin.  The
cleaned media is then conveyed back to the top of the unit.  The
following advantages are claimed for the unit.
     1    It requires no water supply.
     2.   The particulate is removed dry.
     3.   Because there is no corrosion potential, mild steel can
          be used.
     4.   The unit can be small and lightweight because of high-
          velocity throughput.
     This device can be effective when temperature and natural
collection are well controlled.  Officials of the Simpson Timber
plant at Shelton rejected the dry scrubber in favor of a baghouse
because the scrubber would not eliminate the stack plume.  A full-
scale installation on a 3alt-laden hogged fuel boiler at Port
Gamble, Washington, was shut down because of scrubber operating
problems.  These problems reportedly concern cake buildup at the
discharge of the moving bed and blinding of the screen that
contains the moving bed.
     On a more recent full-scale test on a salt installation in
Canada, Conbustion Power Company (CPC) has solved this problem by
providing a steeper angle for the discharge cone.  However,
opacity it, still being exceeded.
     CPC has decided that in the future an electrostatic cage
will be included in each dry scrubber in an attempt to improve
fine particulate collection efficiency.  The dry scrubber will be
marketed as the "electroscrubber".  Some existing plants, such as
the salt installation in Canada will be retrofitted with the
electrostatic cage.

                                85

-------
              TABLE 3-8.  NOVEL FINE  PARTICLATE  CONTROL DEVICES APPLIED TO
                         BOILERS BURNING  SALT-LADEN HOGGED FUEL3
Manufacturer/
unit
Ceilcote Co./
ionizing wet
scrubber
TRW Inc./
charged droplet
scrubber
Pollution Control
Systems, Inc. ,/
UW electrostatic
scrubber
Union Carbide
Bendix Division/
APS electrotube
Air Pollution
Systems/electro-
static scrubber
Commercially
available
Yes


Yes


Yes



Yes


a


Base
equipment
Wet
scrubber

Wet
scrubber

Wet
scrubber


Wetted-
wall pipe
precipitator
Venturi
scrubber

Particles
charged
Yes


No


Yes



Yes


Yes


Water
droplets
charged
No


Yes


Yes



No


No


References
36, 39, 40


27, 32, 38


30, 31



32, 37


32


00
    a All devices have undergone pilot-scale  field  tests  on  boilers  burning salt-laden hogged
      fuel; information from some of the  tests  is sparse.

-------
                    REFERENCES  FOR SECTION  3


1.    Leman,  M.J.   Special  Environmental  Problems  Originated by
     Burning Bark from Saltwater-borne Logs,  Proceedings of Con-
     ference on Wood and Bark Residues for  Energy,  Oregon State
     University.   February 1975.

2.    MacLean,  H., and B.F.  MacDonald.  Salt Distribution in Sea-
     Water Transported Logs.  Information Report  VP-X-45.

3.    Leman,  M.J.   Air Pollution Abatement Applied to  a Boiler
     Plant Firing Saltwater-soaked Hogged Fuel.   Simpson Timber
     Company,  Shelton, Washington.

4.    Johnson,  R.C.   Some Aspects  of Wood Waste  Preparation for
     Use as Fuel.  Proceedings  of Conference  on Wood  and Bark
     Residues for Energy,  Oregon  State University,  February 1975.

5.    Boubel, R.W.  Control of Particulate Emissions from Wood-
     fired Boilers.   PEDCo Environmental, Inc., EPA Contract No.
     68-02-1375,  1977.

6.    Adams,  T.N.   Particle Burnout in Hog Fuel  Boiler Furnace
     Environments.   Tappi,  60(2).   February 1977.  p. 123.

7.    Tuttle, K.L.,  and D.C.  Junge.   Combustion  Mechanism in Wood
     Fired Boilers  JAPCA  28(7) July 1978.  p.  677.

8.    Adams,  T.N.   Combustion Control Modifications to Hog Fuel
     Boilers to Reduce Oil Consumption During Combination Firing.
     Tappi.   60(4)  April  1977.  p.  126.

9.    Horzella,  T.I., and H.R. Newton.  Controlling Air Pollution
     from Hogged  Fuel Boilers.  Pulp and Paper.  February 1974.
     pp. 71-75.

10.   Whitman,  J.L.,  and H.  Burkitt.  Compliance Alternatives for
     Stack Emissions from  the Hog Fuel Boilers  at North Bend,
     Oregon.  Prepared for Weyerhauser Company, Corporate Engi-
     neering.   December 1977.

11.   Council of Forest Industries of British  Columbia.  The Basis
     for Requesting a Variance  on Marine Salts  as Polluting Par-
     ticulates in Stack Gases from Hog Fuel Fired Boilers.
     September 1974 .

                               87

-------
12.  Cheremisinoff, P.N., and R.A. Young,  Air Pollution Control
     and Design Handbook, Part I, Chapter 10.  1977.

13.  Stern, A.C.  Air Pollution, Third Ed. Vol. IV, Ch. 3.  1972.
     p. 106.

14.  Bump> R.L.  Handling Ash from Bark-Fired Boilers.  Power.
     February 1977.

15.  See R.C. North Bend Hog Fuel Boiler Emission Collection
     Options.  Weyerhauser Co.  Environmental Design Section.
     April 28, 1978.

16.  PEDCo Environmental, Inc.  Operation and Maintenance of
     Particulate Control Devices on Kraft Pulp Mill and Crushed
     Stone Processes.  EPA Contract No. 68-02-2105.  Draft
     Report.  August 1978.

17.  Deutsch, W.  Ann der Physik 68:335.  1972.

18.  Matts, S., and P.O. Ohnfeldt.  Efficient Gas Cleaning with
     SF Electrostatic Precipitators.

19.  Mick, A.H.  Wood Waste Fired Boilers:  Wet Scrubber Tech-
     nology.  Georgia Pacific Corporation.  Presented at the 69th
     Annual Meeting of the Air Pollution Control Association.
     Portland, Oregon.  June 27-July 1, 1976.

20.  Cupp, S.J.  Operating Experience with a Boiler Firing Salt
     Water Borne Hogged Fuel.  Crown Zellerbach Corporation.
     1978.

21.  The Mcllvaine Wet Scrubber Manual, Volume I, Chapter III.

22.  Gorman, P.E., A.E. Vandergrift,  and L.J. Shannon.  Fabric
     Filters in Gas Cleaning for Air Quality Control.  Marchello,
     J.M., and J.J. Kelly (Eds.).  Marcel Dekker, Inc., New York.
     1975.

23.  Danielson, J.A.  Air Pollution Engineering Manual.  Air
     Pollution Control District of Los Angeles and U.S. Environ-
     mental Protection Agency, AP-40.  May 1973.

24.  Walker, R.T., and G.J.  Jones.  Recovery and Disposal of Fly
     Ash and Grate Ash from Hog Fuel Fired Boilers.  MacMillan
     Bloedel, Ltd., Technical Report 220.  Pulp and Paper Maga-
     zine of Canada, 74(6),  June 1973.

25.  Whyte, J.E.  North Bend, Oregon Hog Fuel Boiler Emissions -
     Particle Size Via Cascade Impact or Tests.  Weyerhauser Co.
     Research and Development.  Technical Report No. 046-4504-76-14
     August 1976.


                               88

-------
26.   CH2M Hill,  Inc.   A Study of  Particulate  Emission  Discharges
     from the Boiler  Baghouse Installation.   Prepared  for Simpson
     Timber Co.,  Shelton,  Washington.   April  1976.

27.   Fine Particulate Scrubbing-New Problems  and Solutions.   EPA-
     600/2-77-193.  September 1977.

28.   Ramsey, G.H.   Evaluation of  Foam  Scrubbing as  a Method  for
     Collecting  Fine  Particulate.   EPA-600/2-77-197.   September
     1977.

29.   Calvert, S.,  and S.  Gandhi.   Fine Particle Collection by a
     Flux-Force  Condensation Scrubber.  EPA-600/2-77-238 .
     December 1977.

30.   University  of Washington Electrostatic Droplet Scrubber.
     In Evaluation of Eight Novel Fine Particle Collection
     Devices. EPA-600/2-76-035.   February 1978.

31.   Pilat, M.J.,  G.A. Raemhild,  and D.C.  Harmon.   Fine Particle
     Control with UW  Electrostatic Scrubber.   EPA 600/2-77-193.
     September 1977.   pp.  303-318.

32.   Keller, F.R.   Fluidized Bed  Combustion Systems for Energy
     Recovery from Forest Products Industry Wastes. Presented at
     Meeting of  Forest Products Research Society,  Denver,
     Colorado.  September 1975.

33.   Deardorff,  D.  Wet Wood Waste as  a Viable Fuel Supply.
     Presented at meeting of Forest Products  Research  Society,
     Denver, Colorado.  September 1975.

34.   Jasper, M.,  and  P- Koch.  Suspension Burning of Green Bark
     to Direct-Fired  High-Temperature  Kilns  for Southern Pine
     Lumber.  Presented at meeting of  Forest  Products  Research
     Society, Denver, Colorado.   September 1975.

35.   Personal communication:  Mr.  M. Guidon.   Georgia  Pacific
     Corp.  Bellingham, Washington.  July 17, 1978.

36.   Ernst, C.F.,  and P.A. Hamlin.  Evaluation of the  Performance
     of the Ceilcote  Ionizing Wet Scrubber on a Hogged Fuel
     Boiler Burning Fuel with a High NaCl Content.   ITT Rayonier
     Inc., Olympic Research Division,  Shelton, Washington.
     September 1976.

37.   Guidon, M.W.   Pilot Studies  for Particulate Control of  Hog
     Fuel' Boilers Fired with Salt-Water Stored Logs.  Georgia
     Pacific Corp., Bellingham, Washington.   November  1977.

38.   TRW Inc.  TRW Charged Droplet Scrubber Applications Exper-
     ience in Paper and Pulp Industry.  Redondo Beach, Cali-
     fornia.  1977.

                              89

-------
39.  Ceilcote Co.  Ionized Wet Scrubber - New Principle for Air
     Pollution Control.  Technical Bulletin No.  1250,  1255. July
     1976.

40.  Particle Charging Aids Wet Scrubber's Submicron Efficiency.
     Chemical Engineering, July 21, 1975.
                               90

-------
                            SECTION 4
      COST ASSESSMENT OF CONTROL TECHNOLOGY FOR SALT-LADEN
         PARTICULATE EMISSIONS FROM HOGGED FUEL BOILERS

     A review of the literature and contact with wood products
companies indicate that little information is available regarding
costs of conventional control equipment applied to salt parti-
culate emissions from hogged fuel boilers.  Only one recent
comparative cost study of different control devices has been
reported.  Other available data consist of scattered cost esti-
mates of only one type of control device rather than comparative
costs.

COMPARISON OF ESTIMATED AND ACTUAL COSTS
     Weyerhauser Corp. has estimated capital and annual operating
costs for ESP's, baghouses, and venturi scrubbers, as applied to
their salt-laden hogged fuel boilers in North Bend, Oregon.
Table 4-1 summarizes these costs, both with and without combus-
tion modifications to the boiler.  Capital costs are in 1978
dollars.  Operating costs do not include ash disposal.
     With boiler modifications, which result in a smaller gas
volume  (3136 m /min, 112,000 acfm), the ESP is the most expensive,
followed closely by the fabric filter and then the venturi scrub-
ber.  Without boiler modifications and with a gas flow of 5040
m /min  (180,000 acfm), the fabric filter is the most expensive,
followed by the ESP and venturi scrubber, neither of which in-
volves as drastic an increase in cost as does the fabric filter.
     Cost data are available from Simpson Timber Company's hogged
fuel boilers in Shelton, Washington, where fabric filters were
installed to process a total of 6440 m3/min  (230,000 acfm) of
salt-laden particulate emissions.  The capital cost in 1976 was
                               91

-------
              TABLE  4-1.   CAPITAL AND ANNUAL OPERATING COST ESTIMATES  FOR

                 SECONDARY COLLECTORS AT WEYERHAUSER CO. NORTH BEND  PLANT1

                                     (in 1978 dollars)
With boiler modifications:
Total direct cost
Contingency and engineering
Total construction cost
Total operating and
maintenance cost/year
Without boiler modifications:
Total direct cost
Contingency and engineering
Total construction cost.
Total operating and
maintenance cost/year
Baghouse
$1,142,000
171,000
$1,313,000
$ 116,400

$1,629,000
244,000
$1,873,000
$ 190,400
Electrostatic
precipitator
$1,163,000
239,000
$1,402,000
$ 108,700

$1,322,000
271,000
$1,593,000
$ 141,100
High-energy
wet scrubber
$ 863,000
207,000
$1,070,000
$ 207,600

$ 979,000
235,000
$1,214,000
$ 279,000
vo
to

-------
$1.9 million,  or $295 per m /min ($8.26 per acfm).   Escalated to
1978 dollars at 7.5 percent per year,  this cost is  $2.2 million
or $341 per m3/min ($9.56 per acfm).   This compares favorably
with Weyerhauser1s baghouse cost estimate of $372 per m /min
($10.41 per acfm)  for gas flow of 5040 m /min (180,000 acfm).
     The venturi scrubber installed in 1977 on the  hogged fuel
boiler with gas flow of 4872 m3/min (174,000 acfm)  at Crown
Zellerbach's Port Townsend mill cost approximately  $900,000
installed, or $185 per m /min ($5.17 per acfm)  in 1977.   Esca-
lated to 1978 at 7.5 percent per year, this cost is $968,000 or
$199 per m3/min ($5.56 per acfm).  This is approximately 20
percent lower than Weyerhauser's estimate of $241 per m /min
($6.74 per acfm) for a high-energy venturi scrubber.  Weyerhauser,
however, has specified a pressure drop of around 76 cm (30 in.)
water, whereas the Crown-Zellerbach unit operates at a maximum of
51 cm  (20 in.) water.
     The Council of the Forest Industry of British  Columbia pre-
                                                            4
sented a cost comparison of various control devices in 1974.
These costs, which are low because they are in 1974 dollars, show
that the ESP and baghouse are most expensive and are approximately
equal in capital cost at $179 to 214 per m /min  ($5 to $6 per
acfm), followed by the high-energy venturi scrubber, at $107 to
$143 per m /min ($3 to $4 per acfm).
     Table 4-2 summarizes cost information on conventional secon-
dary control devices.

FINANCIAL HARDSHIP
     It is not easy to determine whether financial  hardship  is
created by forcing a company to install secondary collectors to
control salt emissions from hogged fuel boilers.  Obviously,
companies that are operating profitably can more readily afford
to install expensive secondary collectors.  Marginally profitable
installations, in addition to lacking funds, may be located  in
small communities where a major portion of the population depends
on the mill for employment.  If a company decides to  close  such

                               93

-------
                    TABLE  4-2.
SUMMARY  OF  COST  INFORMATION  ON  SECONDARY  COLLECTORS
APPLIED  TO  SALT-LADEN  HOGGED FUEL  BOILERS
vo
Source
Weyerhauser Co.
Northbend, Oreg.






Crown Zellerbach
Port Townsend, Wash.
Council of Forest
Industries of
British Columbia



Simpson Timber Co.
Shelton, Wash.
Type
Fabric
filter

ESP

Venturi
scrubber

Venturi
scrubber
Wet
scrubber

ESP
Fabric
filter
Fabric
filter
Control devic
tf f iciency ,
%
NAb

K
NA

NAb


60-80

NAC


NAC
NAC

99* (design) ,
90-95(test)
e
Gas flow,
acfm
180,000
(112,000)

180,000
(112,000)
180,000
(112,000

174,000

NA


NA
NA

230,000

Capital
$ 106
1.873
(1.313)

1.593
(1.402)
1.214
(1.070)

0.900

NA


NA
NA

1.9

$/acfm
10.41
(11.72)

8.85
(12.52)
6.74
(9.55)

5. 1

3-4


5-6
5

8.26

Annual
$ 106
0. 190
(0.116)

0.141
(0.109)
0.279
(0.208)

NA

NA


NA
NA

0.075

$/acfm
1.06
(1.04)

0.78
(0.97)
1.55
(1 .85)

NA

0.33-0.43


0.23-0.34
NA

0.33

Comments
Numbers in parentheses
assume boiler modifi-
cations are made; costs
are estimates; 1978
dollars



Costs are actual; 1977
dollars
Costs are estimates;
1974 dollars




Capital costs are
actual; 1976 dollars
                Metric conversions:  acfm x 0.28  = m /min
                                   S/acfm x 35.7 = S/m  per min
                  Operating and maintenance costs only.
                  Particulate emission regulation is 0.46 g/sm  (0.20 gr/scf);  efficiency
                  is  not specified.
                  Particulate emission regulation is 0.23 g/sm  (0.10 gr/scf):  efficiency
                  is  not specified.
                NA -  Not available.

-------
a mill, the effect on the community could be detrimental.
     When an agency is considering enforcement action against a
mill that violates regulations of opacity or particulate emissions
because of salt in the hogged fuel, there should be a detailed
analysis of the retrofit difficulty and cost of installing
secondary collectors.  The financial status of the company should
be evaluated in terms of the economic feasibility of installing
control equipment.  If emissions from such a plant are excessive
and cause a violation of ambient air standards in the community,
the company may have to consider closing the plant if they cannot
afford to purchase the control equipment.  If, however, emissions
are close to complying with the particulate emission regulation
and do not cause a violation of ambient air standards, the
enforcing agency may consider some compromise, such as improve-
ment of the existing system or treatment of only a portion of the
flue gas.
     In summary, it is reasonable to expect industry to control
emissions that have a detrimental environmental impact.  Some
consideration, however, should be given to the severity of these
impacts and the economic capability of each company.  This should
be done on a case by case basis.
                              95

-------
                    REFERENCES FOR SECTION 4


1.    See,  R.C.  North Bend Hog Fuel Boiler Emission Collection
     Options.   Weyerhauser Corp., Tacoma, Washington.  April 28,
     1978.

2.    Hoit,  R.S.  Baghouse Filters on Hog Fuel Power Boilers.  Pre
     sented to Air Pollution Control Association, Pacific North-
     west International Section, Seattle, Washington.  April 12,
     1978.

3.    Personal communication:  Mr. J. Walther, Crown Zellerbach
     Environmental Services.  August 24, 1978.

4.    Council of the Forest Industry of British Columbia.  The
     Basis  for Requesting a Variance on Marine Salts as Polluting
     Particulates in Stack Gases from Hog Fuel Boilers, Presented
     to the Pollution Control Branch, Department of Lands,
     Forests and Water Resources, Government of British Columbia,
     Victoria, B.C.  September 16, 1974.
                               96

-------
                                       APPENDIX A
                            SALT-LADEN HOGGED FUEL BOILERS  IN
                             WASHINGTON, OREGON, AND ALASKA
   TABLE A-l.  INSTALLATION LIST OF SALT-LADEN HOGGED  FUEL BOILERS IN WASHINGTON  STATEC
Description of plant
Industry name
Add ress
NEDS number
Puget Sound Plyvood
Tacoma
2 )0 tn»t F Street
1560-0007-01
(706) 627-4111
North Pacific Plyvood
Tfl r oma
1549 Dock
1560-0001-01
(206) 572-4304
Publishers Forest
Product *
Anocor t es
1940-0007-01
(206) 293-2191



Georgia-Pacific
Box 1236
Bel 1 Ingham
2400-0004-01
(206) 733-4410
Mt. Baker Plyvood
Box 997
Be 11 Ingham
2400-0002-01
(206) 733-3960
Simpson Tlmbe- Co.
Waterfront
Shelton
1220-0002-03
(206) 426-3381
steam plant 1
1200-0002-04
steam plant 52


Boiler
capac 1 ty ,
10 Btu/h, or
Ib steao/h
90 x 10* Btu/h
or 90,000 Ib/h
steaBb


40 x 106 Btu/h
or 40,000 Ib/h
steam"


Boiler 1:
61 x 10 Btu/h
ur 63.000 Ib/h
steam"
Boiler 2:
32 x 106 Btu/h
or 32,000 Ib/h
steamb
60 x 106 Btu/h
or 60,000 Ib/h
steamb


50 x 106 Btu/h
or 50,000 Ib/h
steamb


1 at 90,000
180 x 10*
Btu/h



4 at 27,500
60 > 10*
Btu/h
5 boilers
Fuel
rh.ir.ict rr 1 st 1 cs
Fuel type(s) /firing rote(s)
Bark.
Ib/h
22,550




No
data



2,000



1,750



13.333




825




40 , 000





5,000



Oll,c
gal/h
liA




NA




No data



No data



400




NA




NA





NA



Coal ,
Ib/h
NA




NA




NA



NA



NA




NA




NA





NA



Cas,
ftVh
NA




NA




NA



NA



NA




18.750C
natural
gaa


NA





NA



Salt
f rac t ton
of
fuel,
t
No data




Varies
out tO
source


0.34-1.3



No data



No data




Variable,
itu data



No data





No data



Fails
Plowrate ,
scfm
98,000




56,600




69,100



50,000



106,000




50 , 000




55,200





88,300




Parti
load
Inle"t
No data




No data




No data



No data



No data




No data




Nc data





Nc data



culate
Ings,
/scf
Outlet
No data




No data




No data



No data



No data




No data




0.01





U.04



f ract Ion
of
part Iculates ,
I
Inlet
No data




No data




No data



No data



No data




No data




No data





No data



Outlet
70




-v.50




•v-50



50



No data




No data




70





70



Control method/equipment characteristics
Control method
No data




Gravity collector
Multlclone
Cluder collector


Multlclone/
relnjectlon


No control device



Centrifugal
collector
baghouse being
Installed

No data available




Cinder collector
Baghouse (1971)




Cinder collector



Control
device
removal
efficiency,
I
No data




<80




No data







No data
95-99-f
design


No data




994
design




99f
design


Remarks
274 tona/yr of partlculate
emitted; (NEDS)



3 tona/yr of partlculate
emitted (NEDS); by '79 fuel
will be out of water


297 tons/yr of partlculate
emitted. (NEDS)


Variance for 2 yrs - 40*
opacity; possible baghouse.


682 tons/yr of partlculate
emitted; (NEDS).



Ill tona/yr of partlculate
emitted; (NEDS) boiler deslgr
to burn hog fuel, residual
oil, natural gas, and liquid
petroleum gas.
Plune opacity less than
51; 2 additional
boilers bypass bagnouse
and are reserved for oil
firing.





(continued)

-------
                       TABLE A-l    (continued)
Description of plant



Industry name
Address
NEDS number

Crown-Zellerbach
Port Townsend
0900-0001-05
(206) 385-1616


Weyerhauser
Raymond
1480-0004-04
(206) 942-2442
Scott Paper Co.
2600 Federal Avenue
Everett
2000-0002-01
(206) 259-7333
Pope & Talbot
Port Gamble
1020-0002-02
St. Regis Paper
1220 St. Paul
Tacoma
1560-0006-01
(L'Ob) 572-8300

Bolle,r
capac Ity ,
10 Btu/h or
Ib steam/h
Boiler 10
Stoker
Wood- JOO, 000
Wood/oil-
250,000
Oil-250.000
Ib steam/h
No data



60 x 106 Btu/h
or 60,000 Ib/h
steam





70 x 10* Btu/h
or 70,000 Ib/h
steam"


Fuel characteristics



Fuel typc(s)/f Irlng rate(s)
Bark,*-'
Ib/h

51,250





No
data


15,000







20,000




Oil ,c
g.< 1 /h

1,424





NA



No data







2-3 bbl
to 20
bbl


Coal,
Ib/h

NA





.4A



«







NA




r;.is ,
flVh

NA





NA



NA







NA




S.llt
f rac 1 1 on
of
fuel.
Z

0.76-1.6





No data



No data







Highly
variable
depending
on source

Fralsitons rnor.ir tprlst Irs




Flow rate.
scfm

74,000





No data



74,800







75,000






Partlculatc
loadings ,
gr/scf
Inlet

1.52
0.56




No data



No data







No data




Outlet

0.56
0.17




No data



No data







.10-. 15




Salt
fraction
of
participates,
t
Inlet

No data.





No data



No data







No data




Outlet


50-70




No data



No data







70




Control method/equipment characteristics





Control method

1) Multlclone
2) Venturl scrubber




No data



Gravity collector







1) Ducon multlclone
2) Low 6P scrubber



Control
device
removal
efficiency
I
actual
1) 58.5
2) 69.5
88 , 8 over-
all


No data



80-95







No data








Remarks

Efficiency estimates from
6/78 testa; high salt In
fuel. 1131 tons/yr of
partlculate emitted (NEDS)


171 tona/yr of partlculate
emitted (NEDS)


286 tons/yr of partlculate
emitted; baghouse to be
Installed (NEDS)


Boiler shutdown; scrubber 'did
not control salt.

400 tons/yr of partlculate
emitted. (NEDS)



CO
       •  Majority of data  from National Emission Data System (NEDS) Information supplied by Washington Department of Ecology.
         Based on 1000/lb  steam.
         Based on hourly maximum  design rate.
       NA - Not applicable.

-------
                  TABLE  A-2.    INSTALLATION  LIST  OF  SALT-LADEN  HOGGED  FUEL  BOILER  IN  OREGON*
Description of plant
Industry nane
Address
Georgia Pacific
(,oos Bay
Weyerhauaer
U.S. Highway 101
North Bend
Boiler
capacity,
106 Btu/h or
Ih ateaai/n
Boiler 1: ib hr
Design: 100, OOC
Actual: 75.OOC
Boiler 2:- It, '.i
Fuel t ha'-flr t pr Is 1 1 r s
Fuel type(s)/f Irlng rate(s)
Bark,
Ib/h
Varl-
ableb
Design: 100, OOu
Actual: 70,00(1
Boiler 1: •»< nriSS.OOO
Design: 70.0WJ
Actual: *0,00 * t :
Control
device
removal
ef f Irtpncv,
Z
No d
-------
                TABLE  A-3.   INSTALLATION LIST FOR SALT-LADEN  HOGGED FUEL BOILERS  IN ALASKA'




Industry name
Address
NEDS number
Ketchlkan Co.
Power boilers 1
and 2
Alaska Limber 4
Pulp Co.
No. 1 pover boiler
No. 2 power boiler
Alaska Wood Product,
Alaska Pulp Co.
Vangell Limber Plant,
Alaska Pulp Co.

Boiler
capacity.
10* Btu/h or
Ib steam/h
336 x 106 Btu/t
(240,000 Ib
steam/h)


161- x 10 Btu/t
165 x 106 Btu/t
36 x 106 Btu/l

42 x 10* Btu/t





Fuel type(s)/flrlng rate(s)
Bark,
Ib/h
801


401



991

99Z

Oil,
gal/h
20*


601



11

11

Coal.
Ib/h











Cas,
ft'/h











Salt
, f me t ton
of
fuel.
*











Fmtssions rli.ir.ir f or f s t 1 1 ",




Flow rate
ScflK
277,000




95,000
76,300
42,000

49,000



Participate
loading;! (
gr/scf
Inlet
No data










Out tot
0.5-0.7




0.15

0.17

0.13

" S.iTt
frnrt Ion
of
particulars,
7.
Inlet











Outlet
86*


6M



80

70







Control method
Hultlclone
New mechanical collec-
tor to he Installed


Mnltlclone and
venturl scrubber
Cyclone system

Cyclone system

Control
dev Ice
removal
ef f i clency.
7.
No data




85Z
85*
85*
90* design
75*
90* design




Remarks
After Installation of a new
mechanical collector, the
estimated opacity will be 30*.


Plume appears black with
30* opacl ty.
Plume shows 40* opacity.

Plume shows 30* opacity.

o
o
     Data provided by Alaska Department
                            of Environmental Conservation.

-------
                           APPENDIX B
              TYPES OF HOGGED FUEL BURNING FURNACES1*

     A furnace or furnace-boiler system fired with hogged fuel
incorporates several subsystems:  the primary fuel,  combustion
air, ash handling,  instrumentation,  and auxiliary fuel systems.
A brief discussion  of these subsystems is followed by descrip-
tions of the principal types of  furnaces that are fired with
hogged fuel and the major drying systems used in hogged fuel
operations.
     The system by  which the hogged fuel is introduced to the
furnace must be capable of delivering it at variable rates.   It
must be reliable and easily maintained.  Both cost and energy
requirements must be considered  in fuel system design.
     The air system supplies air for combustion and possibly for
cooling of grates or refractories.  The air system must accommo-
date variations in  fuel flow and maintain efficient combustion.
If the system operates by natural draft, the stack must be prop-
erly designed.  Most modern plants do not use natural draft
systems but instead rely on fans to maintain air flow.  The fans
may be driven by electric motors or steam turbines.   The total
air system includes grates, ductwork, dampers, and controls and
may also incorporate an air heater.
     The ash handling system must be sized for the dirtiest
possible fuel, that is, for fuel with the maximum expected ash
content.  Not all of the ash contained in the fuel drops through
the grate to the ash pit.  Some  is carried through the boiler
with the combustion gases, where it may accumulate in "dead"
spaces.  If it does not remain in the boiler, it enters the stack
as fly ash_,  which either is removed from the flue gases by pol-
lution control devices or is emitted from the stack with the gas.
If it is removed, final disposal of the fly ash must  be provided
for.
*Most of this appendix abstracted from  reference  1.
                                101

-------
     Instrumentation and control systems enable the operator to
fire the furnace for maximum efficiency with minimum pollution.
Particulate matter is generated in the furnace and carried
through the boiler.  Although the monitoring and control systems
are expensive, they are needed to indicate current operating
conditions.
     An auxiliary fuel system, which carries the load when wood
fuel is not available, must come on line rapidly and efficiently.
It also must function well with the air supply system and the
instruments and controls.

DUTCH OVEN
     The Dutch oven was the standard design used for wood firing
before World War II.  Because they are relatively small, steam
plants often operate several Dutch ovens in parallel to provide
the desired capacity.  Figure B-l is a cross-section of a Dutch
oven, which is primarily a large, rectangular box, lined on the
sides and top with firebrick  (refractory).  Heat is stored in the
refractory and radiates to a conical fuel pit in the center of
the furnace.  The heat aids in driving moisture from the fuel and
evaporating the organic materials.  The refractory may be water-
cooled to minimize damage to the furnace by high temperatures.
     The fuel pile rests on a grate, through which underfire air
is fed.  Overfire air is introduced around the sides of the fuel
pile.  By design, combustion in a Dutch oven or primary furnace
is incomplete.  Combustion products pass between a bridge wall
and a drop-nose arch into a secondary furnace chamber, where
combustion is completed before gases enter the heat exchange
section.
     This furnace design incorporates a large mass of refractory,
which helps to maintain uniform temperatures in the furnace
region.  This tends to stabilize combustion rates, but also
causes a slow response to fluctuating demands for steam.  The
Dutch oven system works well if it is not fired at high combustion
                                102

-------
                                                         ,- FUEL IN
TO CINDER
COLLECTORS,
AIR HEATER,
AND STACK
 AUX. FUEL
   BURNER
 (IF USED)
                                  UNDERFIRE
                                  -AIR  IN
           Figure  B-l.
Dutch  oven furnace and  boiler.

-------
rates and if the steam load is fairly constant.  With this design,
however, .the underfire airflow rate is dependent upon height and
density of the fuel pile on the grates.  When the fuel pile is
wet and deep, the underfire airflow is low and the fire may be
deficient'in oxygen.  As the pile burns down, the pressure drop
through the pile decreases and the flow rate increases, causing
an excess of air in the furnace.  With fluctuating steam loads,
the result is a continuous change from insufficient air to
excess air.  Because of this feature, together with slow response,
high cost of construction, and high cost of refractory maintenance,
Dutch ovens are being phased out.
     In a well-designed Dutch oven a grate approximately 9 feet
on each side is close to the economical limit of area that can be
supplied with fuel from one opening.  The feed opening is located
so that the conical pile thins to a feathered edge at the furnace
front and reaches a depth of 12 inches at the bridgewall.  With
empirical factors, the known slope of the pile, and the clearance
between apex and arch, it is possible to determine required
height of arch above the grate.  The maximum size of the furnace
unit or cell is thus well-defined and standardized.  The dimen-
sions most frequently used for Dutch oven grates are 8 feet wide
by 9-1/2 feet long and 9 feet wide by 11 feet long.
     Dutch ovens are usually designed with gravity systems that
feed fuel from an overhead conveyor.  Airflow may rely on natural
draft or fans.  Heated forced-draft air is sometimes used, but
most designs rely entirely on the mass of the refractory to dry
the fuel.
     Ash removal is a major problem because not all the ash drops
through the grates to the ash pit.  Provision must be made for
shutting down the furnace periodically to rake the ash from the
grates.   When several Dutch ovens are operating in parallel, one
may be inoperative for cleaning.
     Auxiliary fuel usually is not fired into the Dutch oven but
rather into the secondary chamber below the boiler.  Combustion
           ^
Engineering^ reports high maintenance costs because of the

                               104

-------
tendency of the refractory surfaces to flux when oil is burned in
combination with wood;  continuous use of auxiliary fuel for Dutch
ovens is not recommended.
     Most Dutch ovens at lumber  mills are of the flat-grate type
shown in Figure B-l.   A sloping-grate furnace is used at some
paper mills that burn wet bark.   The fuel enters the front end of
the furnace across its full width and travels down the sloping
grate as it moves through the furnace.  The upper front section
of the grate, which forms the primary drying zone, consists of a
refractory hearth set at an angle of approximately 50 degrees.  A
regulating gate controls fuel-bed thickness at the point of
entrance.
     The middle section is composed of stationary grate bars set
at an angle of 45 degrees and provided with horizontal spaces to
admit air.  The lower section of the grate is set at slightly
less than 45 degrees  and may be  provided with fuel-pushers that
can be operated as required.   Horizontal dump plates extend from
the end of the grate  to the bridge wall.  Progressive feeding of
the fuel from point of entrance  to the dump is achieved by grate
slope.  As the fuel dries, it slips more readily and the lesser
slope in the second section serves as a retardant.  The slope of
the third section prevents the formation of an excessively thick
fuel bed at the bridge wall end  of the furnace.   A portion of the
combustion air is supplied through the two lower grate sections,
and the remainder through tuyere openings in the front of the
bridge wall.  The face of the bridge wall is sloped to cause gas
from the lower end of the fuel bed to sweep over and mix with
gases coming from the drying section of the furnace.
     The fuel bed of  the sloping-grate furnace is comparatively
thin so that, with relatively low undergrate pressures, air can
be distributed through the bed to provide uniform combustion
throughout.  For good operation, however, the fuel should be
quite uniform in size;  otherwise streaks or pockets of greater
density than adjacent areas may lead to formation of blowholes  in
the thin portions of  the bed.  The rate of combustion can be

                              105

-------
increased more rapidly, in relation to the draft, than in flat-
grate furnaces, although the latter can carry much higher over-
loads.  By carefully controlling the rate of feed and using zoned
air supply, the operator can obtain complete combustion with
lower draf-t velocities and less excess air than in operation of
flat-grate furnaces.  Because of this responsiveness, the inclined
grate lends itself to the use of automatic combustion controls.
     Another type of furnace that operates on the same principle
as the Dutch oven is the Dietrich cell.  Figure B-2 shows a
single Dietrich cell under a small, horizontal-return-tube  (HRT)
boiler.  The cell acts to gasify the fuel, and the burning gases
then enter the boiler.  The operational constraints on the
Dietrich cell are the same as those on a Dutch oven.  For both,
the maximum turndown is 3/1.  Control is difficult with rapidly
varying steam loads.  Refractory maintenance is expensive and
time consuming.  The ashes must be raked by hand, and disposal is
usually by means of a wheelbarrow to an open outside pile.

SPREADER STOKER
     Since World War II nearly all of the wood-fired boilers
constructed in the United States have been spreader stokers.  The
design earlier proved satisfactory for coal firing, and many of
the early units were only slightly modified to fire wood residue
or bark.  Some of the more recent units have been specifically
designed for wood firing.  The spreader stoker is an example of
an integral furnace-boiler system.  The fuel is burned in the
base of a water-wall boiler unit rather than in a refractory
chamber.  Figure B-3 illustrates a spreader stoker at the Eugene
Water and Electric Board, Eugene, Oregon,  (EWEB) power plant.
Figure B-4 shows a typical small package spreader stoker, which
can be sent to a plant in modules and erected rapidly.  Several
features distinguish the spreader stoker from the Dutch oven.
                                106

-------
 STACK
                                  •- STEAM TO KILN
                                      FUEL
                                  SCREW  CONVEYER






                           ASH DOOR
                   AIR PLENUM
Figure B-2.   Pile burning:   Dietrich cell,
                     107

-------
                                                        BOILER
        SMOKE
        INDICATOR
        INDUCED
        DRAFT
O
00
        MECHANICAL
        DUST
        COLLECTOR
                                                                              COAL
                                                                                           D
GE-





RS
)
/
/

r
-~ i
•
i
i
i
i
i
i
i
• WOOD
f REFUSE
y STORAG
f











	 V
                                                                                       FEEDER
                                                                                  PNEUMATIC
                                                                                  STOKER
                                                                                        VARIABLE
                                                                                        SPEED FEED,
                                                                                        DRIVE	
                  Figure B-3.   Spreader-stoker-fired  steam generator EWEB  -  Number 3.3

-------
          STEAM OUT c
                                               STAC*
Figure B-4.  Small spreader-stoker  furnace.
                     109

-------
     1.   The fuel is dried by hot forced-draft air rather than
          by radiant energy from a large mass of refractory -
          This is accomplished by passing the flue gases through
          a gas-to-gas heat exchanger before exhausting them to
          the stack.  The forced-draft fan takes in ambient air
          and blows it through the heat exchanger, where it is
          heated to approximately 204°C (400°F) before going to
          •the furnace.  This hot air is forced through the thin
          bed of fuel on the grates to dry the fuel.

     2.   Fuel is fed to a spreader stoker from an overhead
          conveyor, usually through a variable-speed auger
          metering system, to the spreader located at the front
          of the boiler.  The spreader may be a mechanical
          "paddle wheel" type, which knocks the hogged fuel into
          the furnace, or a pneumatic type, which uses air
          pressure to blow the fuel across the grates.

     Figure B-5 shows the pneumatic stoker installed at the EWEB
plant.

     The spreader-stoker system may use a traveling grate, a dump

grate, or a fixed grate.  The traveling grate moves from the rear

of the furnace toward the front.  The larger pieces of fuel are

thrown to the rear of the furnace and therefore remain on the

grate longer to burn.  The ashes on a traveling grate system are

dumped at the front of the furnace.

     3.   Because the spreader stoker is an integral furnace-
          boiler system, it is substantially smaller than a Dutch
          oven of the same output.  Because of the smaller size
          and lighter weight  (no refractory), small units can be
          transported by truck or rail.

     4.   Spreader stokers respond rapidly to load changes.  The
          thin fuel bed and lack of refractory contribute to a
          low "thermal inertia."  This rapid response can be
          detrimental, however, because only a brief failure of
          the fuel system causes the fire to be extinguished.
          Turndown ratios of 4:1 are quoted for spreader stokers.
FUEL CELL
     A fuel cell is a suspension burning system that burns  small-

size, dry fuel supported by air rather than by grates.  The fuel

particles, mixed with combustion air, completely fill the com-

bustion chamber.  This feature differs from fluidized-bed
                                110

-------
DEFLECTOR PLATE
                                   51 cm (20 in.)
                                    AIR PRESSURE
  Figure  B-5.
Pneumatic  stoker - No.  2  boiler
  at EWEB  plant.3
                          111

-------
combustion, wherein fuel particles remain in the "bed" even
though supported by air.  Sanderdust usually is burned in a fuel
cell.  With adequate size reduction, wood and bark residues also
can be burned in suspension.  The advantages of suspension
burning include low capital costs for combustion equipment
because no grates are required and ease of operation because
grate cleaning is eliminated.  The ash goes into suspension as
particulate matter in the exhaust stream or falls to the furnace
bottom for removal.  Rapid changes in rate of combustion are
possible.
     Figure B-6 is a fuel cell of this type.  Figure B-7 shows
the same fuel cell installed to supply heat to a boiler.
     Suspension burning has disadvantages, however.  Because most
of the ash escapes with the exhaust gases, control of fly ash may
be difficult.  For this reason some suspension units are designed
to "slag" or melt the ash in the combustion chamber and thus
reduce the amount of ash entrained in the exhaust-gas stream.
Temperature control in the combustion chamber is critical.  If
the ash-fusion temperature is exceeded, the ash may form large
pieces, which can plug or damage the system.  Fuel preparation
must be thorough to provide sizes small enough for suspension
burning.   Moisture content also must be controlled within rea-
sonable limits,  a requirement that can be costly with systems
burning wood and bark.  With sanderdust fuel, no further pro-
cessing is needed.  Residence time is critical  (as in any com-
bustion system).   Suspension burning inherently provides short
residence.  At high combustion rates, the residence time may be
insufficient for the process to go to completion.
     The capacity of fuel cells is limited; therefore, as more
energy is needed, more fuel cells are added.  As fuel-drying
systems are perfected, it is probable that more fuel cells will
be used,  even on larger boilers.  Figure B-8 shows the complete
system requirement for use of wet wood residue and bark as a fuel
for a large suspension burning system.  Fuel cells are particu-
larly hard on refractory because of the high temperatures.

                              112

-------
                    «fB*C'0°!^^  f
«rv*A«'C«
  AID Pl(»U**
Figure B-6.   The Energex cyclonic burner  (fuel cell)
            WOOD f UEl FROM
            (SiOCEX METERING BIS
    Figure B-7.   An Energex-fired package  boiler.
                            113

-------
                                               COARSE
                                                FUEL
                                               STORAGE
                                              r 351 M.C.
121°C +  (250°F
                                                           D
                                                           FLUE GAS
                                                           CLEANING
                                                    PRIHARY
                                                    AIR FAN
                                                    FINE,  DRY WOOD
                                                    101 M.C.
                                           )SUSPENSION  BURNERS
                                           JOIL/6AS STANDBY
Figure B-8.   Large  suspension  burning system,
                             114

-------
FLUIDIZED-BED COMBUSTION
     One of the newer systems developed to burn solid fuels is
the fluidized-bed combustion furnace.   The system can burn high-
moisture fuels and can react to changes in steam demand more
rapidly than some of the other systems.  Fluidized-bed combustion
of cellulose materials was originally  developed to incinerate
wastes from pulp and paper mills having moisture contents up to
67 percent.
     The fluidized-bed system incorporates a large mass of finely
ground inert material (like sand) ,  which provides a very large
exposed surface area.  The inert material is contained in a
vessel, through which air is passed upward so that the bed
becomes "fluidized"; it resembles a boiling liquid that keeps the
particles in a state of constant agitation.  The bed is preheated
to about 760°C (1400°F).  When a finely divided solid fuel is
introduced, the hot inert mass provides sufficient energy and
radiating surface to "flash"-evaporate the fuel moisture and
gasify the volatile component of the fuel.  The remaining fixed
carbon in the fuel is oxidized as it moves through the fluidized
bed.  The process generates little or  no flame but rather a
glowing bed.  Combustion is rapid,  and the fluidized bed proper
contains no unburned organic material.  Particulate emissions are
therefore minimal.
     The fluidized bed may be used as  a hot gas generator for a
separate boiler,  or heat may be transferred directly from the bed
to the steam by placing bundles of tubes in contact with the
inert material of the bed.
     In a 1975 presentation, Keller4 described application of
the fluidized-bed system to steam plants using wood residue fuels
and indicated plans by Energy Products of Idaho to have 10
fluidized-bed units in operation by September of that year.  This
development has not proceeded on schedule.
                               115

-------
DIRECT FIRING
     Within the past 5 years, installations have been made in the
United States in which the hot gases from burning bark (and wood)
are used directly for heat.  Applications involving direct firing
of wood and bark include veneer dryers, drying kilns for lumber,
and dryers for wood and bark particles.
     Deardorff  describes a pile-burning, hogged-fuel-fired
furnace that supplies heat directly to a veneer dryer.  Jasper
and Kock  report on a suspension burning system in which undried
bark is pulverized and burned in a cylindrical, annular combus-
tion chamber.  The system has been tested in the laboratory, and
the authors propose construction of a production model to be used
with a lumber dry kiln.
     Although direct-firing systems are not hogged-fuel-fired
boilers, they are included in this report for two reasons:   (1)
the problems involving fuel, control, and pollutant emissions are
similar to those with furnaces used in conjunction with steam-
producing boilers; and (2) direct-fired units may replace the
current wood waste boilers, since developmental work on direct
firing is progressing rapidly.

OPERATING VARIABLES
     Variables governing furnace operation are classified in
Table B-l as fuel-related, air-related, and operator-related; all
of these factors contribute to the overall efficiency of the
system.
                                116

-------
    TABLE B-l.  FACTORS AFFECTING THE COMBUSTION REACTION  IN
            BOILER INSTALLATIONS FIRED BY HOGGED FUEL6
Fuel-related factors
     Species
     Size
     Moisture content
     Ultimate analyses
     Proximate analyses
     Heating value
     Method of feeding fuel
     Distribution of fuel in furnace
     Variations in fuel feed rates
     Depth of fuel pile in furnace
     Separate firing practices
     Auxiliary fuel usage

Air-related factors
     Percent excess air
     Air temperature
     Ration of overfire air to underfire air
     Turbulence of air
     Flow relation between forced-draft and induced-draft
       systems

Other factors
     Cleanliness of the combustion system
     Basic furnace design
     Maintenance of components
     Steam generation rate
     Steam drum water level
                               117

-------
                    REFERENCES FOR APPENDIX B


1.    Boubel, R.W.  Control of Particulate Emissions from Wood-
     Fired Boilers.  Prepared for PEDCo Environmental,  Inc.   EPA
     340/1-77-026.  1977.

2.    De Lorenzi, 0. (Ed.)  Combustion Engineering.  First Edi-
     tion, Published by Combustion Engineering - Superheater,
     Inc.  New York.  1952.

3.    Brown, O.D.  Energy Generation from Wood Waste.   Paper
     Prepared for International District Heating Association,
     French Lick, Indiana.  June 20, 1973.

4.    Keller, F.R.  Fluidized Bed Combustion Systems for Energy
     Recovery from Forest Products Industry Wastes.  Presented at
     Meeting of Forest Products Research Society, Denver, Colo-
     rado.  September 1975.

5.    Deardorff, D.  Wet Wood Waste as a Viable Fuel Supply.
     Paper Presentation from the Forest Products Research Society
     Meeting, Denver, Colorado.  September 1975.

6.    Junge, D.C.  Boilers Fired with Wood and Bark Residues.
     Oregon State University, Forest Research Laboratory.  Bul-
     letin 17, Corvallis, Oregon.  November 1975.
                               118

-------
                           APPENDIX C
      CASE HISTORIES OF SECONDARY COLLECTORS FOR CONTROL OF
          SALT-LADEN PARTICULATE FROM HOGGED FUEL BOILERS
FABRIC, FILTER,  SIMPSON  TIMBER  COMPANY, SHELTON, WASHINGTON
The  Emission  Problem         •
      The  Shelton power  plant of .Simpson  Timber Company  consists
of seven  boilers.   Six  of  the  boilers are  rated at  12,480 kg/h
 (27,500 Ib/h) of steam  production  each.  Boiler 7 produces  about
20,420  kg/h  (45,000 Ib/h)  steam, with a  rated capacity  of 40,840
kg/h (90,000  Ib/h)  steam production.  Almost all fuel burned  in
the  boilers is  derived  from salt-water-borne logs.  '
      Historically,  visible emissions _were  4 to 5 Ringelmann No.
opacity and over 0.69 g/std m   (0.3 gr/sdcf). ' Attempts to  reduce
emissions by  combustion with large amounts of overfire  air  were
futile  because  about 70 percent of the flue gas is  salt particulate.
Large chimney diameter  also contributed  to "a high apparent  smoke
density.   Changes  in log handling  in  1973  substantially increased
                              4
visible emissions  from  stacks.  Flat  raft  storage was replaced by
bundled log storage in  an  effort to reduce the space occupied by
floating  logs and  to reduce the depositing of debris in the bay
by dumping of single logs.   ,
      Emission source tests clearly indicated that submicron salt
particles constituted a major  portion of particulate emissions.
Efforts were  made  to limit the log storage time in  saltwater  and
thereby reduce  the salt content of logs.  Limiting  the  log  stor-
age  time  reduced the salt  content  in  stack emissions somewhat,
but  even  with as little as 2 weeks storage (logistically the
shortest  possible  time) compliance with  the particulate emission
regulation was  not achieved.   The  use of mechanical bark presses
'to reduce'the salt and  moisture content  of fuel was rejected
because of the  serious  problem of  disposing of bark pressate
water.

                                119

-------
Selection of Particulate Control Device
     A preliminary study of wet scrubbers indicated that only
high-energy units would remove the submicron salt particulates to
the degree required for compliance.  Because of corrosion and
operational problems, the test facilities were not available
within the allowable time limit.3
     Then pilot tests were performed with a static bed dry
scrubber.  The visible salt plume from stack persisted, and
because of the limited operating experience with dry scrubbers on
boiler gases, the use of dry scrubbers was rejected.
     After pilot tests with fabric filters indicated that this
system could achieve compliance with emission regulations,
Simpson Timber Co. decided to install baghouses for particulate
control.
The Fabric Filter System
     Two baghouses manufactured by Standard Havens Company were
installed in 1976.  One unit handles 47.2 m /s  (100,000 acfm) of
flue gas from Boiler 7 while the other unit handles 61.4 m /s
(130,000 acfm) of flue gas from four smaller boilers, part of a
group of six, of which two were allowed to bypass the baghouse.
     The two baghouses are identical except for size, one having
six compartments or modules and the other eight.  Each module is
3.66 m (12 ft) square and contains 196 bags, 12.7 cm  (5 in.)
diameter by 4.30 m (14 ft) long.  The two systems use a total of
2744 bags.  Effective air-to-cloth ratio is 22.9 to 1 m /s per
 2                  2
m   (4.5 to 1 acfm/ft ).  Temperature of gas entering the  baghouse
ranges from 205° to 260°C  (400° to 500°F) and spikes to 290°C
(550°F).   The bags are Teflon-coated woven fiberglass and are
supported on 1.3 cm by 3.8 cm  (0.5 in. by 1.5 in.) galvanized
wire mesh cages suspended from a top tube sheet.
     Baghouse construction is basically mild steel because
operating temperatures are above the dew point  and the  salt
remains dry and solidified.  Heaters are installed in the reverse
air ducts to maintain heat during shutdown periods and  before

                                120

-------
startup.   The housings are insulated with 7.5 cm (3 in.) of
fiberglass and have a galvanized outer skin to prevent internal
condensation.  The structural frame is also galvanized for ease
of long-term maintenance.   Bags can be replaced with the facility
in full operation by isolating one module at a time with the
reverse air dampers.
     Total additional power required for the filter system in-
cluding compressors, agitators, conveyors drives, and new induced-
draft fans is about 900 watts  (1200 hp).   Both boiler systems are
equipped with Buell multiclone-type cinder collectors ahead of
the baghouses.  For the purpose of cleaning,  a dual system of
reverse air and pulse jet  is used, with pulse jet as the basic
mode.  In reverse air cleaning low-pressure air is supplied in
reverse direction through  the bags while the module is isolated
from the main stream by dampers.  In pulse jet cleaning high-
pressure air is injected for brief periods into the outlet of the
bags and no isolating dampers are needed.  Typically, a given bag
is pulsed every 15 to 20 minutes to hold the pressure drop across
the baghouse in the range  of 2.5 to 2.7 kPa (10 to 11 in. W.G.).
Air for bag pulsing is supplied at 690 kPa (100 psi) by two 75-
watt (100-hp) screw-type compressors.
     The particulates caught by the two baghouses average 1360 kg
(3000 Ib) per day with 70  percent salt.   These are disposed of
in the city sanitary sewer along with 38 to 76 liters/s  (10 to
20 gal/min) of blowdown water.
Operational Experience
     From a gas-cleaning standpoint the performance of these
filters to date has been successful.  The National Council for Air
and Stream Improvement (NCASI) emission tests conducted on the
smaller unit indicated outlet particulate concentrations below
0.046 g/std m  (0.02 gr/sdcf) corrected to 12 percent CO,,.4  The
larger unit tested by CH^M-Hill produced an outlet concentration
                3       z
of 0.092 g/std m  (0.-04 gr/sdcf) corrected to 12 percent CO
The opacity of stack emissions was 5 percent.
                                121

-------
     In the first year of operation, plugging of the tapered
hoppers beneath the bags caused a buildup to the point that the
lower ends of the bags were surrounded with hot salt and ash.
The heat and lack of ventilation apparently caused the bags to
break down and disintegrate.  In 2 years of operation 70 percent
of the bags have been replaced; of these, 300 were destroyed by
plugging that occurred on two occasions.  The onset of plugging
is hard to detect and harder to prevent.  Bag replacement has
become a continuing process.  Cages can be reused but about 20
percent are lost by normal wear and tear or by damage suffered in
the bag-changing process.  Bag replacement cost was budgeted at
$60,000 per year, out of $75,000 total annual operating costs.
     Among other problems, a leaky bag can often fill the cage
solidly from end to end with salt and ash, making it heavy and
awkward to handle.  The rotary air lock valves handling the bag-
house residue are not large enough to handle the amount of ash
broken loose during cleaning of a hopper.  A bypass chute was
installed to alleviate this situation.  Because the original
electric resistance heaters in the reverse air ducts failed in
the salty atmosphere, they were replaced with steam coils.  No
fire in the baghouse has occurred so far.
     A problem with new ID fans is caused by the tendency of the
forward-curved blades to catch a buildup of salt, causing an
imbalance when it is released from one or more blades.  Vibration
detectors were installed on the bearings to give an alarm when
this happens.  Several brief shutdowns for removal of this
deposit have been required.

VENTURI SCRUBBER, CROWN ZELLERBACH CORPORATION, PORT TOWNSEND,
WASHINGTON
The Emission Problem
     The C.rown Zellerbach Corporation's Port Townsend plant  re-
placed nine Dutch oven boilers with one 99,800 kg/h  (200,000
Ib/h) hogged fuel boiler in September 1977.  Under normal  operating
conditions this new boiler provides all power  for  the  400-tons/day

                                122

-------
kraft mill.  No. 6 fuel oil'is burned as auxiliary fuel.  The
boiler control system can respond automatically to changes in
steam demand by adjusting the fuel rate for either ^the hogged
fuel or the No. 6 fuel oil. •
     The hogged fuel consists of sawmill wastes purchased from
neighboring mills, of which approximately 90 percent is fir and
hemlock.  The moisture content ranges from 53 to 58 percent by
weight and the salt content ranges from 0.7 to 1.6 percent, all  •
dependent on source and season.  Sulfur content of the No. 6 fuel
oil is high, approximately 1.5 percent.  Sludge from the primary
clarifier with moisture content of 60 to 65 percent is also'
burned in-the boiler, but' this is considered to reduce boiler
efficiency.                            •
Pollution Control Equipment
     For control of particulate emissions Crown Zellerbach has
installed a 660-tube multiclone system manufactured by U.O.P. Air
Correction Division.  The pressure drop for the multiclone system
is 6.4 cm  (2.5 in.) water gauge.  Additional information with
respect to the multiclone system is not available.  Downstream of
the multiclone system, Crown Zellerbach has installed a .variable-
throat venturi scrubber manufactured by Western Precipitation.
Pressure drop ranges from 38 to 51 cm  (15 to 20 in.) of water.
The scrubber is designed to handle a total gas flow of 4925
m /min  (174,000 acfm) with an inlet temperature of 202°C  (395°F)
and an exit temperature of 66°C (150°F).                 .     ,
Operational Experience                                           :
     Total particulate emissions from the boiler in February 19786
were reported as being 0.23 g/m   (0.07 gr/dscf)..   This complies
with the current emission standard of 0.239 g/m   (0.10 gr/dscf)
set for this installation.  Salt content of the fuel was  0.4
percent, ajid salt: accounted for approximately 26 percent  of total
emissions.  More recent tests  (June 1978) with the salt content
of the fuel greater than 1 percent have shown outlet loadings of
0.32 to 0.41 g/m   (0.14 to 0.18 gr/dscf).7  In some of' these tests

                                123

-------
an attempt was made to induce nucleation of the particulate in
the stack by adding water to the venturi scrubber sump to cool
the gases.  Only -15 to -12°C (5 to 10°F) reductions in temperature
were achieved.  in these tests salt constituted 70 percent or
more of the outlet emissions.  Opacity of the plume was estimated
at 35 percent.
     Thus operation of the scrubber does not comply with the
particulate emission regulation when salt content of the fuel is
greater than 1 percent; pressure drops above 51 cm  (20 in.) water
would be required to achieve compliance.
     Detailed operation and maintenance procedures are not yet
well defined.  A few operating problems have been reported to
date.  Vibration that occurred during startup was eliminated by
additional reinforcement of the scrubber's structural steel.
Erosion and corrosion that occurred in the distribution header
are now prevented by enclosing the header and scrubber nipples.
Upon erosion of the fiberglass separator, it was replaced with a
stainless steel separator.
     No major operation and maintenance problems with respect to
the U.O.P. multiclone system have been reported.
Cost of Pollution Control Equipment
     The capital cost of the scrubber system, flange-to-flange,
is $500,000.  The ductwork, stack, and miscellaneous items
totaled $400,000, giving a total of $900,000 for the complete
scrubbing system.
     Operation and maintenance costs for the multiclone and  the
scrubber systems are not yet available.
                                124

-------
                  REFERENCES FOR APPENDIX C


1.  Leman, M.J.   Special Environmental Problems Originated by
    Burning Bark from Salt Waterborne Logs, Proceedings of a
    Conference on Wood and Bark Residues for Energy, Oregon
    State University.  Feburary 1975.

2.  Air Pollution Abatement Applied to a Boiler Plant Firing
    Salt Water Soaked Hogged Fuel, Simpson Timber Company, EPRS
    Proceedings No.  P-75-13, p. 60.  September 1976.

3.  Hoit, R.S.  Baghouse Filters on Hog Fuel Power Boilers,
    Presented at Spring Specialty Meeting of Air Pollution
    Control Association, April 12, 1978.

4.  Hood, K.T.  Emissions From a Baghouse Operated on a Bark
    Boiler Fired on Residues from Salt Waterborne Logs, NCASI
    Special Report.   May 1976.

5.  Cupp, S.J.  Operating Experience with a Boiler Firing Salt
    Water Borne Hogged Fuel.  Crov.-n Zellerbach Corporation.
    1978.

6.  Drabek, J.  Summary of Source Test - Crown Zellerbach.  Port
    Townsend plant - Venturi Scrubber.  State of Washington,
    Department of Ecology.   February 1978.

7.  Personal Communication - Crown Zellerbach Corporation.
    August 1978.
                              125

-------
                           APPENDIX D

       PERFORMANCE OF NOVEL CONTROL DEVICES APPLICABLE TO
                   SALT PARTICULATE EMISSIONS
IONIZING WET SCRUBBER
     The ionizing wet scrubber  (IWS) combines the principles of
electrostatic particle charging, image force attraction, inertial
impaction, and gas absorption to collect submicron solid par-
ticles.   It was developed by Ceilcote Company to remove fine
solid particulate down to 0.05  gm diameter.  The unit  requires
little energy and its collection efficiency is high  for both
submicron and larger particles.  With fiberglass-reinforced
polyester and thermoplastic materials throughout most  of the  IWS,
it is impervious to corrosive atmosphere according to  Ceilcote
Co.   Figure D-l depicts the ionizing wet  scrubber.
                             SPRAY HEADER
          ELECTRODE
            WIRtS
          Figure D-l.   Ionizing  wet  scrubber (IWS).
Operatiori
     Pollutant  particles  in  entering gas steam are first electro-
statically  charged  within an ionizer section that utilizes a
high-voltage  d.c. power source.   Discharge electrodes have
                                 126

-------
negative polarity, and wetted  "mini-plates" serve as grounded
electrodes.  The plates are  continuously flushed with water to
prevent an accumulation of solid  particles or residues.  Then the
gas stream enters a packed scrubber  section where particles are
removed either by inertial impaction or  by attraction of the
charged particles to a neutral  surface.   Particles of 3 to 5 i_im
and larger are collected through  inertial impaction within the
packed bed.  Smaller charged particles are removed by image force
attraction, as shown in Figure  D-2;  they are attracted to neutral
packing surfaces or scrubbing liquid droplets.   The collected
particles are removed continually  from the stream by a liquid
scrubbing medium, which flows vertically down through the packing.
Operating temperature range  is  93° to 121DC (200° to 250°F).

                          CHARGED PARTICLE

                           IMAGE FORCE
                           ATTRACTION
                             INDUCED SURFACE CHARGES
                           NEUTRAL  SURFACE

                           MIRROR IMAGE
                          OF CHARGED PARTICLE
          Fiqure D-2.
Collection Efficiency
Image force attraction.
     Ceilcote reports that collection efficiency  in the fine
particle range decreases only slightly as  the  particles become
smaller.  The collection efficiency characteristics of ESP's,
fabric filters,  and the IWS from Ceilcote  are  shown in Figure D-
3.
                                127

-------
               1001
                                    1 TWO-STAGE IKS
                                    2 CLOTH COLLECTORS
                                    3 ELECTROSTATIC PRECIPITATOR
                  0   0.2  0.4  0.6  0.8  1.0  1.2  1.4   1.6  1.8  2.0
                                PARTICLE SIZE, i»
           Figure D-3.  Collector efficiency versus  fine
                           particle size.1
     The  IWS is a  fractional  collector.  A single-stage IWS unit
removes  a fairly constant percentage of  incoming particles
regardless of particle size distribution of the total  loading.
To increase the collection efficieny,  the  IWS can be  used as a
multistage unit.    Figure D-4 presents  typical collection effi-
ciency curves for  single and  two-stage  systems.
                100
                90
                80
                70
                60
                50
                40
                30
                20
                10
'HO-STAGE IWS

OW-STAGE IWS

                 0.1     0.2  0.3   0.5     1.0    2.0
                               PARTICLE size, v*
       3.0
           6.0
               10.0
           Figure  D-4.  Collector efficiency of  typical
                         one-stage and  two-stage  IWS units.
                                   128

-------
 Energy  Consumption
      Operating  costs of  the  IWS  are  reportedly  very  low.   Pres-
 sure  drop  through a single-stage unit  is approximately  3.8 to 5.0
 cm  (1.5  to 2  in.) water.  Total  system energy usage  by  a  single-
 stage unit is approximately  52.7 to  65.9 watts  per cm /min (2.0
 to  2.5  bhp/1000 acfm) .   Energy usage by a two-s.tage  unit  is  105
 to  131  watts  per m /min  (4.0 to  5.0  bhp/1000 acfm).
                                                2
 Pilot Studies on Salt-laden Hogged Fuel Boilers
      In  mid-1976 a test  program was  established to evaluate  the
 Ceilcote IWS  in removing fine salt particles from flue  gas of  a
 typical  hogged fuel boiler firing fuel with high salt content.
 The pilot  tests were conducted jointly by eight companies  at
 Simpson  Timber Company.  It was concluded that the Ceilcote  IWS
 can achieve a typical compliance level of stack emissions  at
 0.229 g/sdm   (0.10 gr/sdcf)  adjusted to 12 percent C0_  and also
                                         2           <-•
 can achieve less than 20 percent opacity.    The operating  pressure
 drop  per stage of IWS was about 1.3 cm (0.5 in.) water.
 Emission source--
      The source of the salt fume was the No. 7 hogged fuel boiler
 of Simpson Timber Company.   The boiler burned only hogged  bark,
 sawdust, and shavings from logs transported and stored  in  salt
 water for  3 weeks to 6  months.   A slip stream of 28.3 to  85.0
 m /min  (1000 to 3000 acfm)  was  used with the pilot plant.  Par-
 ticulate in the flue gas at  this point measured 0.46 to 0.69
 g/sdr,   (0.2 to 0.3 gr/sdcf),  of which 79 percent was NaCl  parti-
 culate,  4 to 6 percent  was  C02, and 14 to 16 percent was  0 and
 negligible organic compounds.  Flue gas temperature ranged from
 260° to 315°C (500°  to  600°F).
Ceilcote IWS pilot plant--
     The Ceilcote IWS pilot  plant consisted of a quench section,
a prescrubber, and two  IWS  stages.   Each stage included an ion-
 izing  section followed  by a  cross-flow packed scrubber  similar  to
                                129

-------
the prescrubber.  The quench section was used to cool the inlet
gas from 204 to 206°C (400 to 475°F) to about 65°C (150°F) to
protect the polypropylene tellerette packing in the scrubbing
sections.  The prescrubber was installed to remove large parti-
cles.  The electrical potential in the ionizing section was main-
tained at 16 to 26 kV, and current ranged from 2 to 4 mA at high
particulate loadings to about 20 mA at low loadings.
Results--
     Table D-l summarizes the pilot test results.  Overall effi-
ciency of the pilot control system typically ranged from 99.5
percent at low gas flow rate to 94.6 percent at very high gas
flow rate.  The overall efficiency without electrostatic augmen-
tation was 63.8 percent, which showed that the IWS had a significant
effect on fine particulate collection.  Measurement of aero-
dynamic particle size distribution at inlet and outlet indicated
that IWS efficiency is nearly independent of particle size.
     No information is available on disposal of the wastewater
                       2
collected from the IWS.

CHARGED DROPLET SCRUBBER  (CDS)3'4'5
     The charged droplet scrubber  (CDS) was developed by TRW
Systems, Redondo Beach, California.  The operating principle is
similar to that of the IWS, but instead of charging the parti-
cles, the TRW scrubber charges the water droplets.  The CDS is
designed as standard modules that can be combined in parallel to
handle higher inlet gas flow rates.
Operation
     Figure D-5 illustrates the operation of the TRW module.  The
CDS produces a spray of electrically charged liquid droplets,
which are accelerated through the electrostatic  field between the
spray tube and collector plate.  The pollutant particles  in  the
entering gas are attracted to liquid droplets by means  of  direct
collisions or indirect charging encounters.  The particles are
                                130

-------
                   TABLE  D-l.
PILOT  TEST  RESULTS  OF  CEILCOTE  IWS  ON  SALT-LADEN
      HOCKED  FUEL  BOILER,   1976
Typical
to -st runs
1
2
3
4C
5d
Ga • f 1 ow r a te ,
in Vm in ( .i<~ f m)
34 .6 (1210)
49.1 (1740)
5B.5 (2066)
87.8 (31001
59.3 (2095)
Gas
velocity across
ionizinq plates,
-1.31 (4.1)
•1.89 (6.2)
- 1 . 29 (7.5)
•1.1 (10.2)
NAP
Pressure drop
per s t aqr ,
cm ( in . ) II 0
0 . 76 (0.1)
1.0 (0.4)
1.3 (0.5)
1 . 9 (0.7)
NAP
I n 1 r t
partlculfltp
concon t r,i t i on ,
q/sdm1 (qr/^dcf)
O.B37 ( n . t f . 6 0 )
0. 556 (0.24 )0)
0 . 490 (0 . 214fl)
0. 78 1 (0.1414)
0.423 (O.IRr.O)
P ri r t i c u 1 .1 1 e
conron t r a t inn
brtwren Rtaqfs
q/Rdm1 (qr/sdcf)
0.0254 (0.0111)
0 . 0799 (0.0149)
0.077? (0.03)7)
0.1117 (0.0488)
0. 2400 (0 . 1050) f
Ou tint
part i c u 1 a t o
< - o n r r n t r A \ l o n ,
q/sclm1 (qr/silrf)
0 .004H (0. 002 1 )
0.0151 (0 . 0069)
0 . 0162 (0 . 0071 )
0.0419 (0.0183)
0 . 1 500 (0 . 0670)
O(>ar i t y
a ' | ,1 i n r. t
blue. r.\ -j ,
H
' 10
' 1 0
10
NAP
NA^
  No data available on water requirements.
  No information available for correction  to  12*  CO  .
C Flow rate hiqher than test ranqe.
  Hiqh voltage turned off  across  ionizinq  section.
  No t avaliable.
  Particulate concentration  after  prescrubber.

-------
                   ELECTRODE
              (HIGH POS.  POTENTIAL)
       CHARGED
       DROPLET
        SPRAY
       COLLECTOR
         PLATE
     NEG.  POTENTIAL)
                                     WATER/DUST
                                      SLURRY
                                     CARRY-OFF
                                SCRUBBED-GAS
                                 DISCHARGE
                                TO ATMOSPHERE
                  -.? \~~i--Y.
               .,v*.*.-v\i)'<"-i
  LIQUID
  INLET
      d.c.
     POWER
                        DUST-LADEN
                         GAS FLOW
Figure D-5.   Charged  droplet scrubber  operation.
                          132

-------
carried to the collector plate,  from which they are continuously
drained off with the liquid.
     The feed tube is held at a  high positive potential with
respect to collector plate.   The droplets theoretically have high
local charge densities without wasteful corona currents so that
the CDS can use smaller collector surfaces and operate at lower
power levels.  The slurry is  drained into a settling tank or
other processor, and liquid  is then clarified and recirculated
through the CDS.  The residual sludge can be recovered or dis-
carded .
Field Pilot Tests
     Field pilot tests have  been conducted for various instal-
lations.  Typical industries  in  which the CDS has been pilot-
tested are pulp and paper, utility (coal-fired),  rock products,
and metal foundry.  Overall  collection efficiencies of over 99
percent have been achieved for fine particulates.   The overall
efficiency can be increased  by adding collection  stages or
increasing the specific collecting area.
Recovery Boiler Sulfite Process--
     A pilot test was conducted  to control submicron particulate
emissions containing 70 percent  NaCl and  25 percent SC>  from a
recovery boiler in the sulfite process of a pulp  and paper indus-
try-   The mass mean particle  size was 0.5 ym, and inlet loading
to the CDS ranged from 0.229  to  0.687 g/sdm3 (0.1 and 0.3 gr/sdcf).
In all tests the discharge far exceeded the allowable maximum of
0.0916 g/sdm  (0.04 gr/sdcf). Efficiency was most sensitive to
specific collecting area  (SCA, plate collecting area per unit
value of gas).  With a SCA of 0.52 m2/am3 per min  (0.16 ft2/acfm),
collection efficiency was 95  percent.  The overall energy con-
sumption was about 18 watts  per  am /min gas flow  (0.5 watts/acfm).
Bark  Boiler--
     Though no field tests have  been performed on a boiler burning
salt-laden hogged fuel, pilot tests have been successful on a
                               133

-------
bark boiler  fired  with combinations of heavy  oil  and hogged
fuel.   The  flue gas contained ash, carbon, and  5 percent NaCl.
     A field pilot test is required to show the  effectiveness of
the CDS in collecting fine particulates  emitted  from burning of
salt-water-soaked  hogged fuel.  Table D-2  presents a comparison
by TRW of CDS performance with that of conventional control de-
vices.  TRW  says that low energy consumption  by  the CDS would
lead to low  operating costs.
UNIVERSITY  OF  WASHINGTON ELECTROSTATIC  SCRUBBER
                                                 3,7
     The  UW  electrostatic scrubber was  developed at the univer-
sity under the direction of Dr. Michael Pilat.   This device is
also based the principle of electrostatic  augmentation, but con-
figuration is  different.  The scrubber  is  electrostatically
augmented by charging of the liquid  droplets and the particulates
to opposite  polarities by inductive  charging and corona charging,
respectively.
Operation
     Figure  D-6 shows schematically  the UW electrostatic scrubber
operation.
          GAS INLET






.••V_.:-U--


Ik- t-





	 1~

1 	 1 	 1


                CORONA'
             (PARTICLE CHARGING)   CHARGED WATER SPRAYS
                         (COLLECTION OF CHARGED PARTICLES
                       BY OPPOSITELY CHARGtD WATER DROPLETS)
HIST ELIMINATOR
         Figure D-6.  UW electrostatic scrubber operation.
                                                            7
                                 134

-------
      TABLE  D-2.
CHARGE  DROPLET  SCRUBBER  PERFORMANCE COMPARISONS AT
   EQUIVALENT COLLECTING  EFFICIENCIES4
                             ELECTROSTATIC
                              PRECIPITATOR
                                             CHARGED-DROPLET
                                              SCRUBBER  (CDS)
          APPROXIMATE SIZE
           COMPARISONS:
                         BAGHOUSE
 Performance comparisons:

Power6 3
watts/m /min (watts/cfm)
Residence time
mJ/iWm1n (ft3/cfm)
Collecting area
m?/m3/m1n (ft?/cfm)
Pressure drop
cm (1n.) H20
Mechanical
cosa
10.6-28.2 (0.3-0.8)
0.03-0.04 (0.03-0.04)
0.20-0.65 (0.06-0.20)
1.3-1.8 (0.5-0.7)
No moving
parts
Baghouse
35.3-46.0 (1.0-1.3)
0.09-0.14 (0.09-0.14)
0.56-1.64 (0.17-0.50)
10.2-15.2 (4.0-6.0)
Frequent pulsing
or shaking
Electrostatic
precipi tator
10.6-28.2 (0.3-0.8)
0.12-0.20 (0.12-0.20)
0.65-2.00 (0.2-0.6)
0.5-1.3 (0.2-0.5)
Frequent
rapping
VentuMd
scrubber
141.0-424.0 (4.0-12.0)
0.01 (0.01)
Not applicable
51.0-152.0 (20.0-60.0)
High fan noise
steam plume
  Estimated for 10  - 20 cm collector spacing.
  Estimated at a1r-to-doth ratios of 2 and 6, pulse air cleaned.
c Estimated for 2-3 and 3-5 field sectlonaHzation.
  Estimated for 51  - 152 cm H,0 pressure drop.
e Includes process  gas fan pressure, power supplies, and auxiliaries.
  Includes demlsters, hoppers, gas distribution, and collecting  sections.

-------
TH
    Pollutant particles in the entering gas are electrostatically
charged (negative polarity) in the corona section.  From the
corona section the gases and charged particles flow into a
scrubber chamber, into which are sprayed electrostatically charged
water droplets  (positive polarity).  The gases and some entrained
water droplets flow out of the spray chamber into the mist elim-
inator, consisting of a positively charged corona section in
which the positively charged water droplets are removed from the
gas stream.
Collection Efficiency
     The fractional mass efficiency of the UW scrubber versus
aerodynamic diameter was determined using simultaneously operated
impactors.  Results are given in Figure D-7,3 which shows that
collection efficiency increases with increase in particle size
from 0.3 to 1.0  urn and then remains approximately constant.  Data
on the collection efficiency of the scrubber operated with and
without particle/droplet charging  show the effects of the charg-
ing on collection efficiency.  In  a pilot test on emissions
                        0.?  0.« 0.6 0.8 I. C   ?    «  4610
                              MIllCH DIMT1II. v*
         Figure D-7.   Particle  collection efficiency of
              electrostatic  spray  droplet scrubber
                  as  function of particle size.^
                                136

-------
from a pulverized-coal-fired boiler,  fractional collection effi
ciency for a 0.5 um particle was  99.1 percent.    The inlet gas
flow rate was 1.694 m3/h (997 acfm).   Particle  and droplet charg-
ing voltages were -65 kV and +20  kV,  respectively, and a water
to gas ratio was 0.78 m3/1000 am3 (5.82  gal/1000 acf).  The
   ^                                 3
outlet concentration was 0.0046 g/sdm  (0.002 gr/sdcf).
     The pilot tests indicate that the UW electrostatic  scrubber
can effectively collect fine particulates in the 0.3 to  1.0 um
size range when operated with water  usage rates of approximately
     _        ^                    p
0.8 in /1000 am  (6.0 gal/1000 acf).    Similar pilot tests of the
UW scrubber on an electric arc steel  furnace, a magnesium sulfite
recovery boiler (pulp and paper mill), and a hogged fuel (wood
waste) boiler  have also shown its effectiveness.
Energy Consumption
     The UW scrubber consumes power  in four portions of  the
equipment:  energizing of the pump to provide pressure drop
across the scrubber, energizing of the pump to  supply water
pressure for the spray nozzles, corona charging of the parti-
culate,  and inductive charging of the water droplets.  The
pressure drop across the scrubber is  very low,  about 1.3 cm  (0.5
in.) water.   Calculated total power  consumption for the UW
scrubber is 600 watts per 28.3 m  /min (0.8 hp/1000 acfm).
Wastewater Disposal
     Requirements for treatment  of the wastewater from a UW
electrostatic scrubber would be  the same as that for wastewater
from any scrubber used to clean  the same off-gas.  This electro-
static scrubber consumes more water than most scrubbers, and
therefore the cost of wastewater  treatment would be  somewhat
higher.
Field Pilot Test on Salt-laden Hogged Fuel Boiler
     A field pilot test was conducted recently with  a UW electro-
static scrubber to control salt particulate emissions from a
hoaged fuel boiler.  The test was conducted at a  pulp and  paper

                               137

-------
company in the northwestern United States by Pollution Control
                                                           9
Systems, Inc., a licensee of the UW electrostatic scrubber.
Operation within the compliance levels of 0.114 g/sdm  (0.05
gr/sdcf) particulate and 20 percent-opacity was achieved.  Two-
stage operation of the UW electrostatic scrubber appeared to be
more stable than one-stage operation.  The inlet gas was cooled
before entering the scrubber, and large particulates were col-
lected in the cooling section.  The remaining particulates were
negatively charged, and sprayed droplets were positively charged.
The water for generation of droplets was recycled with some make-
up supply.  The ratio of water to gas was approximately 2 m /1000
am   (15 gal/1000 acf).  The surface area of the charging section
was about half of that of an ESP of same capacity.  Additional
information concerning the pilot tests was not available.

A.P.S. ELECTRO-TUBE
     The electro-tube was developed by Air Pollution Systems,
Inc., and is licensed to Union Carbide Corporation.    The APS
electro-tube is a pipe-type electrostatic precipitator with a
central rod electrode and wetted wall collector.  Figure D-8
depicts the APS electro-tube.
Operation
     Pollutant particles in the entering gas are charged in a
high-energy field by a high-intensity ionizer.at the base of the
electrode.  The charged particles then migrate to the wetted wall
in the body of the device.  The initial saturation charge on the
particles, higher than that in a conventional ESP, facilitates
migration in the collecting electric field.  A laboratory pilot
test has been performed with a range of gas flow rates,  17.0 to
22.7 am /min  (600 to 800 acfm) with a test aerosol, titanium
dioxide  (TiOO.    The mass median aerodynamic diameter  of  the
dispersed 'aerosol was about 1.2 ymA.  Overall efficiency of  99.3
percent was obtained at a gas flow rate of 17 am /min.
                                138

-------
            FROM HIGH-
         VOLTAGE SOURCE
              OPTIONAL
              SECONDARY
              AND TERTIARY
              IONIZATION
              ZONES
AEROSOL
 INLET
                                                 AEROSOL OUTLET
                                                   TO BLOWER
HIGH-INTENSITY
IONIZER SECTION
                              TT
                                                    TANK
                    OUTLET
                  WATER DRAIN
       Figure  D-8.   A.P.S.  electro-tube.
                                              11
                           139

-------
                                                 T ")
      Pilot Test on Salt-laden Hogged Fuel Boiler
     Figure D-9 shows the pilot scale setup for the APS electro-
tube.  The electro-tube was tested at a gas flow rate of 23 m /min
(800 acfm), maximum per tube.  The wetted-wall anode is 3.05 m
long by 0;30 m diameter  (10 ft by 1 ft).  Since these tube
dimensions are the same for both the pilot and full-size instal-
lations, scaleup to a full-size unit should require simply adding
the number of tubes needed for a specific application.  A 5.1-cm-
diameter  (2-in.) driving field electrode was used for testing.
     The inlet gas flow rates ranged from 14 to 23 m^/min (500 to
800 acfm).  Analysis of total particulate showed 35 to 80 percent
NaCl; 30 to 35 percent of the total was less than 1.0 ym in
size.  Inlet gas temperatures ranged from 101° to 192°C (214° to
378°F).  The inlet particulate concentrations ranged from 0.39 to
2.13 g/sdm  (0.17 to 0.93 gr/sdcf) at 12 percent CO-.  Corrected
                                                      3
outlet concentrations ranged from 0.0043 to 0.10 g/sdm  (0.0019
to 0.044 gr/sdcf) at 12 percent C0?; opacity was near zero.
Actual outlet particulate concentrations ranged from 0.002 to
0.049 g/sdm  (0.00087 to 0.0214 gr/sdcf).  The overall collection
efficiency ranged from 95.6 to 99.0 percent.  The measured
pressure drop across the electro-tube ranged from 0.75 to 2.5 cm
(0.3 to 1.0 in.) water.  The water-to-gas ratio ranged from  0.18
to 0.80 m /1000 am   (1.33 to 6 gal/acf).  The only operating
problem was plugging of inadequate drain fittings.
Energy Consumption
     The electrical operating costs of  this device are only  a
fraction of what is required for conventional particulate col-
lection devices.  One of the reasons is the very low pressure
drop.
Wastewater Disposal
     Wastewater treatment requirements  are  the  same  as  those for
a conventional scrubber.
                                140

-------
INSULATOR COMPARTMENT
I 	
4 — i INSULATOR
ANODE
WASH WEIR-
\
^
/I ( <*
HIGH
POWER


I "—I 	 ^ >
' 8 in. dia. \
j < \
1 	 . ^\
VOLTAGE
SUPPLY

0-120 kV
15 mA
)


?-in. dia. DRIVING
FIELD ELECTRODE-
  DISCHARGE
  ELECTRODE <
  LOCATION   ^
     12-in.
   dia. ANODE —
VP
                                        AIR PURGE HEATER
                                            80 ft /min
                                               OUTLET
  1NLFT
  FRESH WATER
    \x
                                                 RECYCLE
                                                 PUMP
                      CONDENSATE  DRAIN
    Figure D-9.   A.P.S.  electro-tube  pilot.
                                                       12
                              141

-------
A.P.S. ELECTROSTATIC SCRUBBER
     The electrostatic scrubber developed by Air Pollution Sys-
tems is basically an electrostatic ionizer followed by a venturi
scrubber.    Figure D-10 shows a schematic diagram of this de-
vice.  An "electrode is placed upstream of the venturi to charge
the inlet particles, which then enter the venturi throat.  The
charged particles are attracted and collected by the highly
polarized water molecules.
     The charged particles are also collected on the wal.ls of the
ionizer section prior to the throat of the venturi.  A thin film
of water runs down the inclined surfaces to keep the walls clear
and prevent high-voltage arcing.  The particle-laden water drop-
lets are then collected by a cyclonic separator and sent into a
settling tank (clarifier), from which water can be recycled to
the scrubber system.
     Laboratory pilot tests have been conducted at inlet gas flow
                3                           3                11
rates of 21.0 am /min (740 acfm) and 22.7 am /min  (800 acfm).
The test aerosol was TiO_. with mass median aerodynamic diameter
of 1.0
Field Pilot Tests on Salt-laden Hogged Fuel Boiler
     A pilot test was conducted in 1977 at a Canadian company in
the Pacific Northwest on a hogged fuel boiler firing saltwater-
soaked logs.    No further details are available.

MATERIALS OF CONSTRUCTION
     Moist flue gas containing large amounts of salt particulate
is highly corrosive.  Materials of construction constitute a
significant part of the total capital cost of a novel device to
control salt emissions.  A good corrosion-resistant material of
construction such as Inconel 625 is very expensive.  The ionizing
wet scrubber and the charged droplet scrubber are made of fiber-
reinforced plastics.  A proper balance of costs, corrosion-
resistance, and strength is required in selecting materials for
devices handling corrosive gases.

                                142

-------
                                    TO INDUCED-
                                     DRAFT FAN
                                       CLEAN
                                        GAS
                                        OUT
                        WATER
                       TO WASH
                       IONIZER
                         WALL
                                     CYCLONE
                                    ENTRAPMENT
                                     SEPARATOR
IONIZER
SECTION
VENTURI
 SPRAY
          HIGH-
          VOLTAGE
          POWER
          SUPPLY
                      RECYCLE
                       PUMP
      Figure  D-10.   APS electrostatic  scrubber
                           143

-------
COMMERCIAL AVAILABILITY AND COSTS

     Table D-3 lists the manufacturers of the novel control de-
vices just described.
     TABLE D-3.
COMMERCIAL AVAILABILITY OF NOVEL CONTROL
           DEVICES.
       Control device
                   Manufacturer/address
      Ionizing wet
       scrubber
      Charged droplet
       scrubber
      UW electrostatic
       scrubber
      APS electro-tube
      APS electrostatic
       scrubber
              Ceilcote Co.
              144 Sheldon Road
              Berea, OH  44017

              TRW Systems, Inc.
              One Space Park
              Redondo Beach, CA
                                                  90278
              Pollution Control Systems Corp,
              300, Evergreen Bldg.
              Renton, WA  98055

              Union Carbide, Bendix Div.
              61 E Park Dr., Wood Road Ave.
              Tonewanda, NY  14150

              Air Pollution System, Inc.
              18642 68th Avenue, South
              Kent, WA  98031
Capital Costs

     No detailed cost data are available with which to determine

capital costs of these control devices.  Order-of-magnitude

comparisons can be made with conventional control devices.

Initial costs depend mainly on configuration of the charging

section, materials of construction, and collecting ...area.  The

latter depends upon volume of gas handled, size distribution of

particulates, salt content of the flue gas, and inlet gas temp-

erature.  The initial costs of an electrostatically augmented

scrubber are generally reported to be somewhat higher than cost

of a baghouse (higher reliability and no hazards are claimed,

however).  As an example, in 1978 the installed cost of a two-

stage ionizing wet scrubber with maximum capacity of 85,000 am /h
(50,000 acfm) gas is $350,000.14
                                144

-------
Operating Costs
     The operating costs of these devices are reported to be very
low relative to those of conventional control devices.  Overall
energy consumption by a single-stage unit ranges  from 27.0 to
52.7 watts per am /min (0.8 to 2.0 hp/1000 acfm).   The cost of
wastewater treatment would be same as with a  conventional scrub-
ber.  Hence overall cost of operation is  reported  to be less than
that of operating a conventional  wet scrubber or venturi scrubber.
                              145

-------
                   REFERENCES FOR APPENDIX D


1.   Ionized Wet Scrubber - New Principle for Air Pollution Con-
     trol, Ceilcote Co., Technical Bulletin No. 1250, 1255.  July
     1976.

2.   Ernst, C. F.,  and P. A. Hamlin.  Evaluation of the Perfor-
     mance of the Ceilcote Ionizing Wet Scrubber on a Hogged-Fuel
     Boiler Burning Fuel with a High NaCl Content.  ITT Rayonier
     Inc., Shelton, Washington.  September 20, 1976.

3.   University of Washington Electrostatic Droplet Scrubber,
     Evaluation of Eight Novel Fine Particle Collection Devices.
     EPA-600/2-76-035.  February 1976.

4.   TRW Charged Droplet Scrubber Applications Experience in Pulp
     and Paper Industry.  TRW, Inc., Redondo Beach, California.
     1977.

5.   Baker, W. , Hybrid Wet Scrubber's Pilot Tests Show High-
     Collection Efficiencies.

6.   Personal Communication:  Mr. Robert Hession, TRW Inc.,
     Redondo Beach, California.  July 18, 1978.

7.   Fine Particle Control with UW Electrostatic Scrubber.   EPA-
     600/2-77-193.  September 1977.

8.   Pilat, M. J.,  and D. F. Meyer.  University of Washington
     Electrostatic Spray Scrubber Evaluation.  EPA-600/2-76-100.
     April 1976.

9.   Personal Communication:  Mr. Arnthroeen, Pollution Control
     "System Corp.,  Seattle, Washington.

10.  Personal Communication:  Mr. Don Norman, Air Pollution
     Systems, Inc., Kent, Washington.  July 17, 1978.

11.  Evaluation of Four Novel Fine Particulate Collection Devices
     EPA-600/2-78-062.  March 1978.

12.  Guidon, M. W., Pilot Studies for Particulate Control of Hog
     Fuel Boilers Fired with 'Salt-Water Stored Logs.   Georgia-
     Pacific Corp., Bellingham, Washington.  November 1977.


                               146

-------
13.   Personal Communication:   Mr.  Mike. Guidon,  Georgia-Pacific
     Corp.,  Bellingham,  Washington.   July 17,  1978.

14.   Personal Communication:   Mr.  Roger Ruehl,  Air Pro,  Inc.,
     Rep.,  Ceilcote Co.,  Cincinnati,  Ohio.   July 20,  1978.
                             147

-------