oEPA
              United States
              Environmental Protection
              Agency
              Industrial Environmental Research
              Laboratory
              Research Triangle Park NC 27711
              Technology Transfer
Capsule Report

Paniculate Control by
Fabric Filtration on
Coal-Fired
Industrial Boilers
I


-------
Technology Transfer
EPA 625/2-79-021
 Capsule Report

 Particulate Control by
 Fabric Filtration on
 Coal-Fired   |
 Industrial Boilers
July 1979
This report was developed by the
Industrial Environmental Research Laboratory
Research Triangle Park NC 27711

-------
Top view of baghouse, Martinsville, Virginia

-------
1.   Introduction
Until recently, the accepted method
for collecting fly ash from fossil-
fuel-fired boilerjs has been by
electrostatic precipitator (ESP).
Because of uncertainties in service
life and the concern that fabric
plugging during| temperature
excursions below the dewpoint
might disrupt boiler operations,
fabric filters have had limited use
for the control of particulate
emissions from! industrial coal-fired
boilers.  However, recent
events—namely,, the conversion
of oil- and  gas-rired boilers to
coal and the pending promulgation
of more stringent particulate
emission reguletions—have led to
a renewed interest in the use of
baghouses for boiler particulate
control.

Currently about 40 industrial
coal-fired boilers are controlled by
fabric filtration systems. These
installations represent variations in
boiler size from 5 to 165 MW and
a range in  air flow from 10,000 to
700,000 actual |ft3/min (4.7 to
330 m3/s). Fifteen vendors are
                                   associated with
                                   installations.
               the 40
                                   There are several reasons for the
                                   increased use of fabric filters on
                                   industrial boilers. Industrial
                                   facilities usually purchase coal for
                                   the short term and thus may
                                   purchase from a different supplier
                                   every year. This! practice leads to
                                   wide variation in coal properties,
                                   the most important being sulfur
                                   content. A broad range of sulfur
                                   content can be tolerated more
                                   easily in a baghouse designed with
                                   the proper bag material than in
                                   an ESP, where collection would
                                   degrade with lower sulfur fuel.
                                   Moreover, many industrial plants
                                   are already using baghouses to
                                   control some other process stream,
                                   and familiarity with the equipment
                                   simplifies maintenance.
When ESP's are used, coal
composition—particularly sulfur
and alkali metal content—dictates
particle capture capability. The
greater relative availability of low
sulfur coals, especially in the
Western States, indicates that they
must constitute a major energy
source in the  future. Fly ash from
the combustion of low sulfur coals
(virtually all coal deposits west of
the Mississippi River) is particularly
difficult to capture by ESP because
this fly ash has a greater
resistivity. This characteristic has
necessitated the use of oversized
precipitators, hot-side precipitators,
or some form of flue gas
conditioning to meet emission
regulations. Also, low velocities are
needed at the inlet plenum to
distribute the  flue gas properly,
and these velocities can lead to
excessive fly ash buildup and
troublesome operation.

As a result of these factors, the
economics of fabric filters
compares favorably with that of
ESP's when the fuel source tends
to vary significantly or when low
sulfur fuels are used.

-------
2.   Theory
The basic design of a fabric
filtration  unit is straightforward.
It consists of high efficiency woven
or felted fabric, usually in the
form of tubes or bags that are
placed in a housing structure
called a baghouse. Distribution of
the dirty  gas by means of specially
designed entry and exit plenum
chambers provides equal flow
distribution through the filtration
medium.
Baghouses can be operated under
negative or positive pressure,
depending on the location of the
fan and the properties of the
aerosol  system. If the fan is
upstream of the baghouse but
downstream of the dust source, as
in Figure 1(a), the unit will be
under positive pressure and the
fan will be handling dirty gas. With
the fan  downstream of the
baghouse, as in Figure 1(b), the
bag compartments will be under
negative pressure and the fan will
be handling clean gas. Placement
of the fan usually depends on
several  site-specific factors,
including how the unit will be
checked for filter leaks. In recent
years, however, the trend has been
toward placing the fan
downstream of the baghouse.
                                   Induced draft fan^Martinsville, Virginia

-------
                                                                                                                  1
Particulate matter is initially
captured and retained on the fibers
of the cloth by means of
interception, impingement,
diffusion, gravitational settling,
and electrostatic attraction. Once
a mat or cake of. dust is
accumulated, further collection
is accomplished by sieving as well
as by the foregoing mechanisms.
The cloth then serves mainly as
a supporting structure for the dust
mat responsible for high collection
efficiency. Periodically the
accumulated dust is removed for
disposal. Some residual dust
remains and serves as an aid to
further filtering.
The air-to-clothj ratio is an
important criterion in specifying a
baghouse for any application. This
ratio is defined as that of the
actual gas volume flow rate to the
net fabric area;' it has the same
units (feet per minute or meters
per second) as the superficial or
face filtering velocity. Low air-to-
cloth ratios ten'd to increase
collection efficiency because there
is less chance of creating pinhole
leaks in the dust layer where
aerodynamic drag effects are
meaningful. The air-to-cloth ratio
depends on many factors—such as
bag  material, cleaning method, and
particle size and concentration
properties—and ranges from 2:1 to
10:1 and higher for industrial
boiler applications.
The most important components of
any fabric filtration system are
the bags or filtering elements. The
performance of these components
may be very difficult to predict.
Frequently, bag materials are
coated with various substances to
lengthen their service life and
enhance particle collection.
Examples are graphite finishes to
help reduce static charge buildup
and simultaneously to "lubricate"
the yarns to reduce abrasion, and
Teflon® and silicone treatments to
improve resistance to abrasion.
       Bottom
        ash
       Coal
       Boiler
                  heatetv
       Bottom
        ash
                  M
                                              Baghouse
 Figure 1.
 (a) Positive Pressure Baghouse and (b) Negative Pressure Baghouse
                                                   (b)

-------
Table 1 shows chemical,
mechanical, and temperature
resistance of some leading fabric
materials. The selection of the
proper bag material is most
important; bags are usually the
highest maintenance component of
any filter system, accounting for
up to 20 percent of the erected
cost with some fabrics. Depending
on application, bag life may range
from 18 to 36 months and display
typical replacement levels of 15 to
25 percent per year. Some
common causes for bag failure are:

• Blinding, when excessive dust is
  irreversibly retained within the
  fabric pores, making gas flow
  resistance prohibitively high
• Caking, when a solid mat of
  self-adhering fly ash is formed
  on the dirty, gas side of the bag,
  and cannot be removed by
  normal cleaning
• Burning,  abrasion, chemical
  attack, and routine aging,
  characterized by fiber degradation
  and yarn breakage

The method used to clean fabric
filters has an important bearing
on the choice of operating air-to-
cloth ratio and selection of bag
material. Fabric cleaning  may be
accomplished by mechanical
shaking, by reversing the direction
of the flow of air through the bags,
by directing a high velocity air jet
at the bags from a reciprocating
manifold, or by rapidly expanding
the bags by a pulse of compressed
air.
Cleaning by mechanical shaking is
accomplished by isolating one of
several bag compartments from
the air flow and vigorously shaking
the bags for about a minute to
dislodge the dust. For simplicity of
operation, the bags are usually
attached to a motor-driven
oscillating carriage. Because of
tensile stresses produced by this
approach, strong fabric material
must be used.
Table 1.
Evaluation of Leading Fabric Materials
&••:• 	 i' 	 '^ 	 : ":
- Cotton

||jy|onB 	
IDacron'' 	 	
in, .- • i
nOrlorv* 	 	
tCreslanc 	
pDynel^ 	
^Polypropylene 	 '1 LV.V .
[fefibrrt* " .'.."...".'. .'.
F 	 : 	 i 	 : 	 : 	 :
ipiltron0 	 	 	 	

,{; 	 ^ 	 ; 	 , 	
?Minimum continuous
Temperature3
180
200
	 ~'. 	 200 	
275
250
250
160
"7. 	 200 	
450
500
270
400

Acid
resistance
Poor
Very good
Fair
Good in most mineral acids;
partly dissolves in
'. concentrated H 2804
Good to excellent in mineral acids
Good in mineral acids
Little effect even in high
concentration
	 Excellent
Inert except to fluorine
Fair to good
Good to excellent
Fair
operating temperatures recommended by the Industrial Gas
Alkali
resistance
Very good
Poor
ExceJIent
Good in .weak .alkali; fair
in strong alkali
Fair to good in weak alkali
Gqod in, weak alkali
Little effect even in high
jconcefrtratiori
Excellent^,
Inert except to trifluonde,
JSbJoripe, and molten
alkaline metals
Fair to good
Good
Excellent at low
, temperature
i *K -- -AS ft i "• 1
Cleaning Institute
Flex
abrasion
Very good
Fair to good
Excellent
Very good
Good
Good to very good
Fair to good
, Excellent
Fair
Fair
Good to very good
Excellent

1
1
!,
-1
at,
ft
i

   Jupont registered trademark.
tfAmencan Cyanamid registered trademark.
j^Owens-Corning Fiberglas registered trademark.
f*W. W, Criswell Division of Wheelabrator-Frye, Inc., trade name.

-------
Reverse flow baghouses employ
an auxiliary fan that forces air
through the bags in a direction
opposite to that of filtration. This
procedure collapses the bag and
fractures the dustcake. The reverse
flow rate, which is ordinarily about
the same as the face velocity
during filtering, deflates the bag
and helps to dislodge the dust
cake from the fabric surface. It is
common practice to combine
shaking and reverse flow cleaning
in the same unit.

Reverse jet cleaning uses a
slotted blow ring that slowly
traverses each bag while directing
a high velocity air jet against the
downstream filter face. The blow
ring travels the length of the bag
on repeating or intermittent cycles
that are regulated by pressure or
timer controls. The jet of high
pressure air dislodges the dust
cake through combined actions of
bag flexure and aerodynamic
reentrainment.

Reverse pulse cleaning is the
newest and perhaps the most
efficient method to clean bags,
given the proper working
environment. A short pulse of
compressed air (about 0.1  second)
is injected into each bag through a
venturi, causing the bag to expand
while creating intense dust
separating forces. If the process
is carried out when a compartment
is off line, the purged dust can
settle into the hopper instead of
being reentrained in the gas
stream. Compartment isolation,
however, can p -esent two
problems: First, complete
settlement of the dust to the
hopper may lead to overcleaning,
resulting in unacceptably high
emissions. Second, compart-
mentalized operation of a pulse jet
system introduces additional
damper and control costs, and
increases the average operating
resistance and
maximum and
resistance.
Because of the
the range between
"ninimum
              unique
characteristics of combustion
processes, special precautions
must be taken lo avoid operational
difficulties. Dur ng boiler startup,
the air stream and all internal
surfaces of the baghouse must be
preheated to prevent sulfuric acid
and moisture from condensing on
fabric baghouse surfaces.
Condensation problems at any
time during boi
lead to serious
er operation can
chemical corrosion
and degradation of all materials of
construction, ireluding filtration
fabrics. Aside from damaging the
fabric irreversibly, interim plugging
of fabrics or bridging and sticking
of dust in the hoppers and
materials-hand ing systems may
also lead to rapid curtailment in
gas flow rates and to excessive
emissions.  Because most large
boilers are  equipped with a
preheater to improve combustion,
baghouse condensation problems
can be eliminaied with existing
hardware.
The baghouse and hopper
structure also should be insulated
to minimize heat losses. In some
instances, adequate insulation may
make preheating unnecessary.
Separate internal heaters may also
be required in the ash hoppers
to prevent dust caking and to
promote the flow of ash out of the
hopper. As an adjunct, or possibly
an alternative, to complete system
insulation and internal heating,
continuous air recirculation during
shutdowns may be used to control
the condensation problem.

In the event that an emergency
causes a rapid rise in fabric
resistance, provision should be
made for immediate bypassing of
the flue gas to the stack until the
boiler can be turned down or taken
off line, according to prescribed
operating procedures,  to avoid
hazardous and sloppy  combustion
in a reducing atmosphere. This
step will reduce the chance of
permanent damage to the fabric
during transient plugging.

-------
3.    Present Applications
 Fabric filter applications have
 shown substantial diversification
 in recent years, with the result
 that control of fly ash emissions
 from utility and industrial boilers is
 now commonplace.  Combustion
 and metallurgical applications will
 represent a major part of U.S.
 baghouse sales in the future.

 A survey of industrial boilers
 shows that operating data are
 currently available for 37 industrial
 boilers controlled by bag houses.
 These facilities include 22 stoker-
 fired coal boilers, 11 pulverized
 coal units,  2 oil-fired boilers, and
 2 boilers fired with hogged fuel
 (sawdust and wood bark). Another
 half dozen  or so facilities are now
 contemplating baghouse
 installations, but have not yet
 specified design and operating
 parameters.

 The current users represent 15
 different vendors; however, four
 vendors have supplied 24 of the
 37  operating systems. Boiler
 capacities range from 8,500 to
 1.4 million Ib/h (1.0 to 176  kg/s)
 steam, and flue gas flow rates vary
 from 3,500 to 730,000 actual
 ftVmin (1.7 to 345 mVs). Woven
 glass fabrics with various surface
 treatments to improve wear
 characteristics (Teflon®, silicone,
 or graphite) are used almost
 exclusively. Eighteen of  the
 systems went on line before 1976;
the remainder either started or
were planned to go on line
after 1976.
A survey of these coal-fired units
reveals that 9 of 11  pulverized coal
boilers  use reverse air cleaning,
and these 9 devices operate at an
average air-to-cloth ratio of
about 2.3:1.

In the case of stoker-fired units,
the choice of cleaning method is
equally divided between the
reverse air and pulse-jet approach.
The average air-to-cloth ratio for
reverse air systems is about 2.1:1
in contrast to 5:1 for those using
pulse-jet cleaning. Pulse-jet
cleaning with  industrial boilers
appears to be  relatively common
because of the availability of
commercial units in the small
boiler size ranges. In discussing
the variations  in cleaning methods
and operating  velocities for the
two coal-firing methods, stoker
and pulverized, knowledge of the
dust characteristics in the gas
stream  is essential. Available data
from several sources  indicate that
the mass median diameter of
uncontrolled fly ash emissions
from a pulverized coal boiler is
about 12 /urn, whereas a
stoker-fired system discharges

-------
much coarser particles, up to
about 50 nm. The main effect of
the smaller particle size generated
by pulverized coal firing is to
increase the pressure drop across
the fabric for a fixed mass  inlet
loading, roughly equal to 2 to 4
inches (5 to 10 cm) water for
air-to-cloth ratios ranging from
2:1 to 2.5:1.

Distinctive features of reverse air
and pulse-jet cleaning methods
determine which is more
appropriate for a given application.
Reverse air cleaning is ordinarily
used in conjunction with 10-oz/yd2
(0.3-kg/m2) woven  fabrics  that can
withstand tensile stressing at the
50- to  100-pound (23- to 45-kg)
level without failure. Filtration
velocities, however, are usually
restricted to less than 3 ft/min
(0.015 m/s) to prevent excessive
dust penetration. Because  of its
high cleaning efficiency, pulse-jet
cleaning should be  used only with
felted fabrics. The principal reason
is that the cleaned felt alone still
affords a significant dust retention
capability, much more so than
woven fabric cleaned to the same
level. Improved felt  performance
arises not only from the greater
fabric weight, ~16 versus
10 oz/yd2 (~0.5 versus 0.3 kg/m2),
but also from the high degree of
fiber dispersion characteristically
found in felted media.
   p  Jf\          v
mF"^ ,"      '     ~\
                                    Reverse airflow inlet duct, Martinsville, Virginia

-------
Because individual bags are pulse
cleaned at frequent intervals
(~1 minute), a more permeable
dust layer develops on the fabric
surface. Therefore, higher
velocities are possible without
prohibitive resistance to gas flow.

It should be noted that pulse-jet
cleaning can only be used
effectively when the dust is
collected on the outside of the bag
and the air pulse is admitted to
the inside, or clean side, of the
system. This method requires that
the bag or tube be supported  on
a wire cage to prevent bag
collapse during normal filtration.

Since the dust is cyclically
removed and redeposited during
pulse jet cleaning, the collection
efficiency may, in some
applications, be diminished. Hence,
system performance should be
carefully evaluated with respect
to certain hazardous materials and
local emission limitations. Because
of the  near uniform fabric loading,
however, the pulse-jet system will
operate within a narrower
pressure drop range, and thus
exhibit only minor fluctuations in
air flow and paniculate emission
rates. The above factors must be
reviewed carefully with respect to
local regulations,  existing fan
capabilities, and normal boiler
operating loads.

Gas stream temperature is very
important because it determines
the choice of fabric with respect to
gas-handling capacity, resistance
to gas  flow, mechanical strength,
wear resistance, and potential
corrosion problems. The
temperature of the air entering the
baghouse should  be above the
acid dewpoint, but should be low
enough to minimize the actual
volume of air filtered and to permit
the broadest possible selection of
fabric materials. Generally,  flue
gas temperatures should be held
between  300° and 400° F (149°
and 204° C),  a condition that is
usually satisfied when the boiler is
equipped with an  economizer
(water  preheater)  or an air
preheater. If neither an economizer
nor air heater is used, the air
stream may require cooling with
dilution air. Because only the
larger boilers are usually fitted
with economizers or preheaters,
dilution cooling would probably
be the most practical approach for
smaller units.

For most installations, insulation
is needed on all duct work,
manifold, baghouse, and hopper
surfaces to prevent the
temperature from falling below
the acid dewpoint. If the
temperature falls too low, dust and
condensed moisture may form a
nearly impermeable coating on the
bags, preventing a free gas flow.
The bags then may rupture or be
damaged irreparably by  chemical
(acid) attack, as may many
materials of construction in the
filter system.

-------
4.    Performance
Over the last several years, control
of fly ash emissions from coal-fired
industrial boilers with fabric filters
has been shown to be a viable
alternative to control by ESP's,
scrubbers, and other devices.
Operating experience gained to
date indicates tlpat low emission
levels are achievable, although not
without some operating problems.
The discussion that follows will
center on several functioning
installations and their operating
parameters and| mechanical
problems.
Table 2 lists 13 facilities for which
detailed operating data have been
obtained. All units use woven or
felted fabrics made of fiberglass
with finishes consisting of Teflon®,
silicone, or graphite, singly or in
some combination. (Teflon®, where
used, constitutes about 10  percent
by weight of the fabric.)

The reported air-to-cloth ratios
range from 2:1 to 4.5:1; the higher
values basically correspond to
those units using pulse cleaning.
                                   Baghouse ash collection hoppers, Waynesboro, Virginia

-------
Table 2.
Several Industrial Boiler Baghouse Installations and Associated Operating Parameters
Plant name
and
location
Caterpillar Tractor Co 	
Decatur IL
= Simpson Timber Co 	
: Shelton WA
University of Minnesota 	
St. Paul MN
Adolph Coors Co 	
; Golden CO
*. Amalgamated Sugar Co 	
1 Nyssa OR
Amalgamated Sugar Co ... .
: Twin Falls ID
s; Carborundum Co 	 	
; Niagara Falls NY
i Delco-Remy 	 	 	
Anderson IN
•; E. 1, DuPont 	
New Johnsonville TN
* E. 1, DuPont 	 	
" Parkersburg WV
- E, 1. DuPont 	
; Martinsville VA
~ E. 1. DuPont 	
Waynesboro VA
	 U.S. Steel Co. 	
Provo UT

Size(103 Ib
steam per hour)
260 (3 boilers)
	 170 (5 boilers)
140 (2 boilers)
	 250
	 200 (3 boilers)
400 (2 boilers)
75
50 (3 boilers)
	 270 (2 boilers)
450 (4 boilers)
	 380
645
	 1 440 (3 boilers)

Boiler
Type
Stoker
Dutch oven firing of
hogged fuel
Stoker

Stoker

Stoker
Stoker
Stoker
Stoker
Pulverized coal




date
Nov 1976
Feb 1976
Feb 1976
Jan 1977
Oct 1973
1974
1969
Feb 1977
Dec 1975
Dec 1 974
1977
1977
Mar 1978

Gas
flow
(103 actual
ftVmin) ;
150
230
63
175
92
200
35
24
130
??1
203
340
730

Available test data show that
paniculate emission rates vary
from 0.01  to 0.085 lb/106 Btu
(0.02 to 0.15 g/106 cal) with five
systems discharging less than
0.03 Ib/106 Btu (0.05 g/106 cal).
Based on limited data, bag service
life appears  to range from 10
months to 3 years. The facility
experiencing a 10-month bag life
attributed this result to operation
at 15 percent over design air flow
and rapid pulsing (18 cycles per
hour). When the pulsing rate was
reduced to 9 cycles per hour,
satisfactory results were claimed.

Cost information was generally not
available, but information that was
given showed  an average fabric
cost of about $30 per bag—$0.72
per ft2 of cloth—and a  total
installed cost of about $11 per
actual ftVmin.

One facility experienced fly ash
caking in hopper exit lines caused
by penetration of wind-blown rain.
To correct the  problem, plans are
being made to enclose  the
baghouse modules completely, to
pelletize the fly ash, and to store
the pellets  in a silo.
Excess flue gas moisture arising
from a pinhole leak in
the economizer at one boiler
installation led to steam
condensation in the baghouse
with irreparable bag damage.

One plant has experienced bags
ripping along the entire length of
the seam. The bags were partly
cleaned by shaking, but the failure
was attributed to defective bags
rather than to the shaking
mechanism (which is known to
have the potential to damage
woven glass fabrics).

-------
Table 2.

Several Industrial Boiler Baghouse Installations and Associated Operating Parameters—Continued
         Plant name
            and
           location
Temperature
Airj-to-cloth
  j ratio
  (gross)
 Cleaning
mechanism
 Fabric
material
- Caterpillar Tractor Co
1* Decatur IL
FSimpson Timber Co
f:.: Shelton WA
<* University of Minnesota ....
t=». St. Paul MN
^Adolph Coors Co
p^; ; Golden CO ~ "" 	 ",."-_ 	
jg Amalgamated Sugar Co
Is Nyssa OR
^Amalgamated Sugar Co 	
SiF -Twin Falls ID
I Carborundum Co 	
'I- Niagara Falls NY
t Delco-Remy 	
1' Anderson IN
* E. 1. buPont 	
J- , New Johnsonville TN
TE, I. DuPont 	
£ Parkersburg WV
rE.I. DuPont '. ''..".'. ."..'. ........
I" Martinsville VA
££. 1. DuPont 	
fj Waynesboro VA
fcU.S. Steel Co 	 	
£~. Prpvo UT
400
500
360
338

300
Stoker 285 to 300'
pulverized coal, 350
..150
.. 350

330
. . 330
.. NA
NA
360

44-1
4 5'1
2 1
2 3-1

3 6-1
2 5-1
• 	
Z8:i"
3:1

3 7-1
37-1
2 1
2 1
34 1


reverse air
Bassist



__.,-• . ' • ». = J- •• . --:'!» ' •» 1 •-" .. " ... • .„-.-. • .
Reverse air
On-line pulse

On-line pulse
On-line pulse
Reverse air/vibrator assist



Glass with 10% Teflon®
coating (24 oz/yd2)
Glass with 10% Teflon®
coating
No 0004 fiberglass with
sHtCQ.ne-graphite-Teflon®
finish
^ silicone-graphite finish

silicone-graphite finish
Fiberglass with Teflon®
coating
Glass with Teflon® finish

Teflon® felt (23 oz)
Teflon® felt (23 oz)
Glass with 1 0% Teflon®
coating
Glass with 10% Teflon®
coating
silicone-graphite finish
1

„
41
-

"
«

T-

•-•£
a
-.=
-
|
Another plant had significant
bearing wear in the screw
conveyor following the hoppers.
The problem was remedied by
installing hard iron bearings.

Because of "creeping" or
consistent rise in bag pressure
drop, one plant found it necessary
to wash the bags annually.
Installing double diaphragm valves
helped to promote more efficient
pulse cleaning and eliminated the
washing requirement.

The baghouse users surveyed
made other  suggestions that
would appear to be useful to
potential customers:
     It is highly recommended that
     3 inches or more (8 cm or more)
     of mineral wool insulation be
     used on all surfaces where
     excess heat loss might result in
     condensation problems.
     Injection of limestone to precoat
     the bags before starting up the
     system can aid the filtering
     mechanism before initial dust
     cake formatiori.
     Fine mesh cages—0.5 inch
     (1.3 cm) square—should be used
     in conjunction
     fabrics to prevent excess wear.
   with woven glass
      • Where practicable, collected fly
        ash may be mixed with other
        waste-line products from the
        plant to eliminate the need for
        a separate disposal system.

      It appears that many of the
      problems encountered at the
      surveyed facilities pertain to the
      transport and disposal of the
      collected fly ash. Hopper fires
      (caused  by overloading), sticking
      hopper doors, faulty screw
      conveyors, and caking fly ash are
      typical problem areas that should
      be controllable with proper design
      and operating procedures.

-------
Table 2.
Several Industrial Boiler Baghouse Installations and Associated Operating Parameters — Concluded
Plant name
". , ar>d
: location
Catepillar Tractor Co 	 	
: Decatur IL
; 	 Shelton WA
' University of Minnesota 	
St. Paul MN
~ Adolph Coors Co 	
: Golden CO
• Amalgamated Sugar Co 	
* Nyssa OR
; Amalgamated Sugar Co
: Twin Falls ID
Carborundum Co 	
: Niagara Falls NY
1 Delco-Remy 	
Anderson IN
* E. 1. DuPont 	
New Johnsonville TN
* E. 1. DuPont 	
.: Parkersburg WV
- E, 1, DuPont 	
i Martinsville VA
* E. 1. DuPont 	
a Waynesboro VA
U S Steel Co 	
ii Prove UT
i "SH = Standard Havens, Inc. C = The Carborundum Co.
* Note, — NA = not available.
Coal
• ... Sulfur Ash
BtU/'b (o/o) (%) 	
	 12,500 2.0 8.5
	 Hogged fuel (salt-laden
sawdust and bark)
	 8,600 0.85 10
12,000 1.0 9
	 8,750 0.4 18 to 25
	 NA 0.5 5.0
	 10 000 0 85 NA
	 '.'; 	 NA 	 2.5 " ' 15
	 NA NA NA
	 NA 3.2 7
	 NA 2.5 7
	 12,500 1.25 11
	 1 2,500 1 25 1 1
	 NA 0 65 6

Paniculate
emissions
date
(lb/106 Btu)
0.01
0.027
NA
0.027 to 0.085
NA
0.05
..NA " ,„;, 	 	 ;
NA
0.02
0.02
NA
NA
NA
WF = Wheelabrator-Frye Inc. WP = Western Precipitation.
Baghouse
manufacturer8
SH :
SH ;
C ' ' _^
WF
WF ;
WP 1
::.."„£"•'..' . ' „•
SH I
: SH j
SH
WP •
WP
WF ]


-------
5.   Economics
It was pointed out earlier that
fabric filter capital costs become
                                  competitive with
                                  when low sulfur
               those of ESP's
               coals are burned
                                  and where high collection
                                  efficiencies are desired. One
                                  control equipment manufacturer
                                  estimates that an ESP operating
                                  at 99.8 percent efficiency cannot
                                  be retrofitted economically to
                                  compete with a baghouse at 99.9
                                  percent efficiency until the sulfur
                                  content of the co'al exceeds  2.0
                                  percent. At the 99.5-percent
                                  efficiency level, an ESP would
                                  require a coal with more than
                                  1.5 percent sulfur.
Operating costs have traditionally
been less for ESP systems
because of their low pressure
drop—approximately 0.5 inch (1.3
cm) water in contrast to 4 to 6
inches (10 to 15 cm) water for a
baghouse. If the two control
systems are to be operated for
equal efficiency, however, the
overall power requirements for the
ESP may be  closer to those
observed for fabric filter systems
where fabric and dust cake
resistance necessitates a high
energy input. Fabric life now
becomes the deciding factor in
determining whether the baghouse
or the ESP has the lower
operating cost.
                                  Baghouse top support system, Waynesboro, Virginia

-------
The costs associated with any
fabric filtration installation depend
on several factors. A key variable
is the gas-handling capacity, for
which the cost per actual cubic
foot per minute generally
decreases as the size of the
system increases. Figure 2 shows
this relationship for an ESP and a
fabric filter system. The design
specifications used to develop
these cost statistics are given in
Table 3. The baghouse not only
performs  at a higher efficiency and
a correspondingly lower outlet
emission  rate, it also costs less to
install than  the ESP.

Additional cost data are presented
in Table 4 for several industrial
and utility boilers. It should be
understood  that comparison of
cost data  derived for several
different sources sometimes can
lead to erroneous conclusions
owing to  differences in the costing
base. For example, the Nucla
facility costs include such items
as new stacks, painting, and an
ash-handling system for both fly
and bottom ash; these items
normally  are not included in
capital cost estimates. Despite
these uncertainties, the data
provide a realistic range to aid the
filter user.

As shown in Table 4, average
capital costs (projected to April
1978 with Chemical Engineering
Cost Indexes) range between $10
and $11 per actual ftVmin, and
average operating costs are
between  $0.80 and $1  per actual
ftVmin. The projected April 1978
capital costs as a function of
airflow are  plotted in Figure 3.
Although a  strong correlation
between  cost and size is indicated,
the expected advantage of reduced
cost per unit size as gas volumes
increase  is  not clearly
demonstrated. If the data points
representing each information
source or specific working size
range are treated as separate
groups, however,  it appears that
                          Note.—See Table 3 for design criteria.

                          SOURCE: Industrial Gas Cleaning Institute.
   E
  S
  to
  CO
  r~
  o
  C/3
  O
  O
                                        ESP at 97.3% efficiency
        2 —
                 20
 40      60      80      100

SYSTEM SIZE (103 actual ftVmin)
                                                       120
                                                              140
Figure 2.

Installed Cost of Fabric Filter (FF) and Electrostatic Precipitator (ESP)
on Industrial Coal-Fired Boilers as a Function of Volumetric Flow Rate
capital investment increases
roughly as system size raised to
the 0.75 power. (See black lines).
It is suspected, therefore, that
differences in costing procedures
for the various reporting groups
tend to conceal the true
exponential decrease in unit cost.
            Available data show that stoker
            boilers of industrial size—
            < 300,000 Ib/h (40 kg/s) steam-
            typically operate at higher excess
            air levels than noted for  pulverized
            coal firing. Thus, emission control
            costs should be greater for stoker
            boilers because of larger flue gas
            flows per unit of coal fired.

-------
Table 3.
Design Criteria for Fabric Filter and Electrostatic Precipitator on Industrial Coal-Fired Boilers"
              Parameter
                                              Electrostatic precipitator                           Fabric filter
'Q3l                                                     i
s^ Sulfur	,,	._,   08%__
^ Ash	'...    75%               |
E" JWoisture	 - -    5 0%   _ _        L     „ _   „
te~~Btu/lbr.	7.'	"..."T	T2,800  "           ,             "
                   	.-...    "Spreader stoker, 10 to 250 x 1 p6 Btu/h
                                     10% by weight OO/um
                                     3/16 inch mild steel ^     ~* ~
                        -   *""""     2"inches      _^_

                       -  - -     ~ 350° F to eOO^F    .
        size
 Construction
 Insulation
 pearling method
   ejpperature
   •abnc
   utleLemissiorTrate
                                   """oTTETTo6 Btu"
                       0.8%
                       75%
                       50%
                      '12,800
                       Spreader stoker, 10 to 250 x 106 Btu/h
                       10_% by weight < 10/xm
                       CaTbon steel/compartment design
                       2 inches
                      " Pulse-jet/automatic cycling
                       350° F to 600° F
                      ^Suitable for temperature range (type not specified)
                      '001 lb/106Btu
                                                                                                                         4
 See Figure 2 for installed cost "as a function of volumetncTflow rate*
                      ^       ^    *~^-  _  _  i_ ^-m  ^  I
    RCE Industrial Gas Cleaning Institute
 Table 4.
 Summary Capital and Operating Costs for Utility anc
                                                         Industrial Boilers
     g:|^a||||W^~^,—.,-,,


      >,ip^£fta?wiii?li^ia;SHiili|ij
                                    rear™  *""ft3/min)   !  April 1978b/
                                    S^Spf-'SsWfS^iSp^rtffrnillions)
                                               '
                                                                         DitHf costs      ' -'-*:^ ™*  ^      '^Annual
                                                                         !^S^SS£^^S^^S^22'^^S;5'p'CfP'tih'g'co'':ta
                                                                         :..    Dollars per        "~       (dollars per
                                                                            actual ft3/min            actual ftVmin)
                          year   April 1978b   Base year   April 1978°
   Siifgfl!^w^i|ifi»iift»iii«i^s^l^^
   —-'-goilers: • "  .   -_'.-•-—   -  -	^   .   ^     ,
   „.Jpffitf"!'. ."7!7^;."".11.~.~"."7.'7r~rT9Z?; " ~~~1'^5''^ZXT 2^
   l"^MSuffictorii51orT-7'^"':Apr:.T977" TspO""'. .!. J~15.ff
   lSI^:'Sunbury"77'.;r7~7::'~lWarrj'976 .    '888'  ' ~^~  -•-§£
                                                                             9.25
                                                                           -g-gQ-

                                                                             6.20
                                                                            12:76
                                                             NA
                                                             0.33
                                                             0.77
                                                             1.34
   fe'lnpiustfial Gas Cleaning
   cK>: ;^^^'3%M^S'-^^fl?»».fg^=^M!a3i»^^«!a^=fel^«IB«p^^^J^=	^_™M-=..J	g,^	.,„ „. 1^
   »i« Institute  	   Jan. 1977         5.4      ;
   |t,^,^:/:,fc.m3fSi!,s!:b;-™S;S»(iM^i^p^«s^^fc«£s^««jS5g-2jJ-	" <--	
s«'S«fa^|ig^^^«'^)i^Wi«:^^
                 sgiss^teie*%i^sM4Ss^s^

                       12.60        14:62        1.56        1.67    ]
                    "    6'44   	   7.47        0.81         0.87   ' "
                                                                                          636
                                                                                                      1.56
                                                                                                      0.81
                                                                                                      Of2
                                                                                                      (370
      :"EjC"!7Trri7?!7!7r^^"••"'"•'.'&ug*'^&''~"M^_ "70'
    Includes .'elecTrj^'piov^r^aintenancejand repair4, "and bag replace^jelTt" Does not^inclucie arriortized capital costs, spacejaci^jpancy,

                                                                          B>y ^?y
                                                                          tnde

  depreciafion, and so fortrT".
 ^^A^^j^^^^'r-'^^^M^'f^f^^v^ -^MjE^^Ks. ^^ ^ "^ ™.^^sFs^m»^m ^ Fr^t^ pBsjfe^gT ^B>y ^ry J^"^^1  ^ ?"T^, "^^^g6. „ r
 |gcai|d from base_yea£u_sjng ChemicaTEngineeringFabmaled Equipment Costjndex                                    ^     3
 ^^i^|™g*^||^^"'^^^^e^^:£^/ne^J^"F^^l^fg3^^^m'ent clisTlnd^forljagTeplacernent cosr<20'percenFof  "*   ,
   flgatifJcosfrc^nslflSionTaiof Index for labor (55*pe7cent),'and Electric Rate Indexesjor power cost (25 percent)

      i jata supplied by.EPRi
        er read at '67th AnhuafMeetirig "of the Air Pollution Control Association
          '    ' "       '                 ''                -
                                            ^^^^        -^^ —
                                                                                                                    u

-------
                  10s
                  107

CO
              0.
              <
                 106
              UJ
£
                 105
                 104
                                                                 Linear regression line-
                                                                 r = 0.976 (log data)
                                                                 r = 0.954 (actual data)
                                   EPRI, Joy Manufacturing Co., and
                                   EPA, utility boiler data
                                    GCA case study,
                                    industrial boiler data
                                                                  EPA pilot plant study,
                                                                  industrial boiler data
                                                                   Industrial Gas Cleaning Institute,
                                                                   industrial boiler data
                              Note.—Black lines show approximate regression lines for individual data groups.
                                         _L
                                                 J_
                    103
                           104                   105

                                     SYSTEM SIZE (actual ftVmin)
                                                                                    106
107
Figure 3.

Capital Investment Versus System Size for Several Coal-Fired Boilers

-------
A second factor affecting system
economics is the operating
air-to-cloth ratio or filter face
velocity. As the ratio increases,
capital investment decreases
because  more gas flow can be
accommodated per unit fabric
area. On the other hand, the
pressure drop increases with
increasing air-to-cloth ratio, the
actual  rate depending on the type
of fabric  and the method and
frequency of cleaning.  Ultimately,
any advantage gained by reduced
capital costs will be offset by the
additional energy required to
overcome the higher fabric
resistance plus whatever energy
increase is incurred by the
fabric-cleaning process itself.

A third factor affecting system
operation is the initial cost for the
filter fabric material and its
expected service life. Currently,
bag prices range from $20 to $55
per bag—$0.40 to $3 per ft2—
depending on the cost of the fabric
per se and the related  sewing
costs as a function of size. For
a given fabric, tie overall cost per
square foot of f Itering area tends
to decrease witi increasing bag
capacity. An average price for
commonly used woven glass  fabric
is about $0.75 per ft2. Additional
costs of $10 to $15 per bag may
be incurred if  new seals, clamps,
or wire cages  are required at bag
replacement time. Fabric life will
impact significantly on system
operating  costs up to a bag life of
3 years. Beyond this point, the
cost advantage [lessens
considerably with increasing
bag  life.

In summary, st jdies have shown
that several factors affect overall
costs for the optimum filter
system; i.e., total gas-handling
capacity, air-to-poth ratio, and
fabric costs, including initial
investment and replacement  costs.
Although  increased air-to-cloth
ratios represent a viable means to
improve system economy, the
potentially adverse effect of high
air-to-cloth rati'o on paniculate
emissions must be considered.
Air-to-cloth ratios in the range of
2:1  to 4:1 are now predicted to be
optimum values when all variables
are considered. Installation costs
for fabric filter systems vary from
$6 to $15 per actual ftVmin, with
an average value of about $10 per
actual ftVmin.  The cost for a
similarly sized ESP system ranges
from $7 to $16 per actual ftVmin.
Annual operating costs for fabric
filter systems vary from $0.30 to
$1.50 per actual ftVmin with a
nominal average of about
$0.90 per actual ftVmin. The
corresponding annual operating
costs for an ESP vary from $0.10
to $0.35 per actual ftVmin. A
more severe emission requirement
will increase the operating costs
for  ESP's, whereas the values
cited for fabric filters would remain
unchanged.

-------
          This report has been reviewed
          by the Industrial Environmental
          Research Laboratory, U.S.
          Environmental Protection Agency,
          Research Triangle Park NC, and
          approved for publication. Approval
          does not signify that the contents
          necessarily reflect the views and
          policies  of the U.S. Environmental
          Protection Agency, nor does
          mention of trade names or
          commercial products constitute
          endorsement or recommendation
          for use.
This capsule report was prepared for the U.S. Environmental Protection
Agency by the GCA Corporation in Bedford MA. Principal contributors
were Douglas Roeck and Richard Dennis. The report was developed
by the Environmental Research Information Center and the Industrial
Environmental Research Laboratory's Particulate Technology Branch
in Research Triangle Park NC. E. I. du Pont de Nemours & Company,
Inc., has given EPA permission to reproduce the photographs of their
facilities that use fabric filters on coal-fired boilers.
•A- U.S. GOVERNMENT PRINTING OFFICE: 1979 — 659-245

-------