benitez cap10

8/11/2019 Benitez Cap10

http://slidepdf.com/reader/full/benitez-cap10 1/36

10

FabricFilters

IO.I NTRODUCTION

Fabric

iltration is a widely acceptedmethod

or

particulate

emissions

ontrol. In fabric

filtration, the

particle-laden as

lows through a

number of filter bags

placed n

parallel, eaving

the dust

retainedby the tabric. Extendedoperationof a

fabric filter, or baghouse,

equires hat

the

dust be

periodically

cleaned

off the cloth surfaceand removed rom

the baghouse.

After a new

fabric

goes

hrough a f'ew cycles of use and cleaning,

t retains a residualcake

of dust that

becomes he filter medium.

This

phenomenon

s responsible or the

highly efficient

filtering of

small

particles

hat characterizes

aghouses.

The type of cloth fabric limits the temperature f

operationof baghouses. otton

fabric

has he least esistanceo

high

temperature

about

355

K), while fiberglass

has he most

(530

K).

In many cases his

requires ha t the waste

gas

streambe

precooled

before entering he

baghouse.

The cooling system hen

becomesan integral

part

of the design.Conversely,

he temperature

f

the exhaust

gas

streammust be well above he dew

point of

any

of its condensable

onstituents s

liquid

particlesplug

the fabric

poresquickly.

Most of the energy requirements f the systemare

to overcome he

gas pressure

drop

across he bags,dustcake, nd associated'ductwork.

ypical valuesof

pressure rop range

rom I

to 5 kPa. The most mportantdesign

parameters the ratio of the

gas

volumetric

'low

rate o fab-

ric area,known as he ga.r-lo-clothatio. Other mportantprocess ariables ncludeparticlechar-

acteristics

such

as sizedistributionand stickiness),

as

characteristics

temperature

nd corrosivi-

ty), and abric

properties.

Electric

utilities have made significant

progress n recent

years

n designing

and operat-

ing baghouses

br the collectionof coal fly ash.

nterest n

baghouses

an be expected

o continue

increasing s air emissionstandards ecome

more stringentand concernsover

inhalable

particles

and air toxic emissions ncrease.

n many

cases.

he relative nsensitivityof

baghouses

o changes

in

fly ash

properties

make hem more attractive han ESPs

The oldest fly-ash baghousewas commissioned

n 1973. By 1990, here were close to

100

baghouses

n

operationon

utility boilers, representingmore than 21,000

MW of

generating

capacity

Cushing

et al. 1990).

414

Se

r

0

thc'

n te t

r

0 ,

l i - -

t - l L ' .

( ) \ s l .

l e i t .

l I a : . :

l C ] .

I I L

l 1 J - -

rc.1 .

e r .

10

ed ,

inr f

c l a .

i l l l d :

s - : - :

] ] ' :

'

. ; :

S)

',



Sec.

10.2 Types f

Fabric

ilters

415

IO.2

ryPESOF

FABRIC

ILTERS

Your objectiven studying hissections to

Describe he operational

haracteristics

f the three

nost common

baghouse

esigns:

I

)

shaker;

2)

reverse ir:

and

3)

pulse

et.

The most

mportant istinction

etween

abric ilters

designs

s the

methodused

o clean

the dust

f iom the

bags between

iltration

cycles.

The distinguishi ng

'eatures

f each

cleaning

method redescribed

ubsequently,

ollowing

Turneret al

(1987a).

10.2.1 hoker

leoning

For

any type of cleaning.

enougl.r

nergy

niust be

imparted

o the

fabric to

overcome

he

adhesion

brcesholcling

he dust

o the bags.

n shaker

leaning.

sed

with inside-to-outside

as

flow. this

is accornplished

y

suspending

he bag

from a

motor-driven

ook

or fiamework

ha t

oscillates.

he

motion creates

sine

wave along

he fabric,

which

dislodges

he

previously

ol-

lectedclust.Chunks

of

agglomerated

ust all

into a hopper

below

the compartment.

he cclmpart-

mentsoperate

n sequence

o hat

onecompartment

t

a t ime

s cleaned.

igure

0.l is a

schenrat-

ic diagramof

a shakerbaghouse.

Parameters

hat affect

cleaning

nclude the

amplitude

and

fiequency

of the

shaking

motion and he

tension

of the mounted

bag.

Typical values

of the

first two

parameters re

4 Hz for

frequencyand 50

to 7-5

urm fbr

antplitude.

The vigorous

oscillations

end

to

stress he

bags and

require

heavier

and more

durable

abrics.

n the United

States,

woven

fabricsare used

almost

exclusively

or shaker leaning,

whereas

n Europe

elted abrics

are used.

| 0.2.2Reverse-Air leoning

When fiberglass

'abrics

were ntroduced,

gentler

means

of cleaning

he bags

was

need-

ed

o

prevent

remature ag

ailure.Reverse-air

leaning

wasdeveloped

s

a less ntensive

a)' o

impart energy

o the

bags.Gas

-low o

the bags

s stopped

n the compaftment

eing

cleaned.

and

a reverse

low of

air is directed

hrough

he bags.

This

reversal

f flow

gentlycollapses

he

ba-ss

and the

shear

orces developed

emove he dust

from

the surface

of the

bags.

The

reverse

air tilr'

cleaning

omes rom

a separate

an capable

f supplying

lean,dry

air

for one

or t\\

o compart-

ments t a

gas-to-clothirtiosimilar

o that

of the orrva rd

as

low.

Many

reverse-air

achouses

re nstalling

onic

horns

t-rntprove

hc

performance

f the

cleaning

tep.

Sonic

nergv

sins

horns ur ing

he

everse

lou concent l r te\

l lore

leuning

ner-

gy

at

the bag-dust

ake

nterface.

he

esulting eciuction

n residual

ustcake

density

educes

he

pressureropb1 50. i to 60.r rCur landSmith 9Slbr Sonic orns perate)n ompressedirat

,t



416 Chap.

0 Fabric

ilters

Figure

10.1Schematic iagram of a shaker

baghouse

pressures

of about 400 to 900 kPa. It appears hat

reverse-air

cleaning

with sonic assistance

s the

cleaningmethodof choice

or

baghouses n

pulverized

coal-fired

boilers

Cushing

et

al. 1990).

|

0.2.3

Pulse-Jel

leoning

This form of cleaning orces

a

burst of compressed ir down through he bag

expanding

it violently. As with shaker

baghouses,

he fabric reachests extension

imit

and

he dust separates

from the bag. n

pulse

ets,

however, iltering flows are opposite n direction

when compared

with

shakeror reverse-air esigns,

with

the outside-to-inside

as

low.

Bagsare mountedon

wire

cages

to

prevent

collapse

while the dusty

gas

lows through hem.The top of the bag and

cageassembly

is attached o the baghouse tructure,whereas he bottom end

is loose and tends o

move in the

turbulent

gas

low.

Most

pulse-jet

baghousesare not compartmented.Bags are cleaned

by rows

when a

timer initiates he burst of

cleaning

air through a

quick-opening alve.

Usually

lOToof the

collec-

tor is pulsed at a time by zones.A pipe aboveeach row of bags carries the compressedair. The

Sec

l : I a

. .

:

_ f i

r e

n : : : .

L.rs'

P. '

' r0,3

F

n n t i n r ' - -

Althous

desi

ne;

10 ,3 ,

\ o

u n l I i . ' ' ;

i t ; i < r . - : : ' .

. : - .



Sec.

10.3 Fabric iltrationheory

417

pipe

s

pierced

aboveeachbag so that cleaningair exits directly downward

nto the bag.The

pulse

opposesand

interrupts forward

gas

flow for

only

fractions of a second.

However. the

quick

resumptionof

forward

flow redeposits

most

of the dust back

on the

"clean"

bag or on

adjacent

bags.Pulsejetsnormally operate t about wice the

gas-to-cloth atio

ofreverse-air

baghouses.

IO.3FABRICITTRATIONHEORY

Your

objectives

n

studying his section re

o

1. Describe, n

qualitative

erms, he mechanisms f

particle

penetration

througha

fabric filter.

2. Estimate he tubesheet

ressure

rop n baghouses.

The key

to designinga

baghouse s to

determine

he

gas-to-cloth atio

that

produces

he

optimum balancebetween

pressure

drop

(operating

cost) and baghouse

ize

(capital

cost).

Although collection efficiency is another mportant measureof

baghouse

erformance,

roperly

designed nd well-run baghouses re

highly

efficient

I0.3..| enetrotion

Design overall efficiencies

or fabric filters range from

98Va

o more

than

99.97o.

Most

units currently

n

operationmeet or exceed

heir design efficiencies

Cushing

et al. 1990).Two

basic mechanisms ccount or the

particles

hat

pass

hrough he baghouse.

Particlescan escape

collection through

leaks in the ducting, tubesheet,or bag clamps,

or through holes, tears,

or

improperly sewn seams n the bags hemselves. he second

mechanism or emissions

s

particle

movement hrough the dustcakeand the fabric, also

known as bleed-through.Bleed-through s

primarily

a

function

of baghouse esign

and

particle

morphology.Smooth,

spherical

particles

are

less

cohesive, nd can seep

hrough he dustcake nd abric easily,

ncreasing missions.

Well-developedequationsexist in aerosol

physics

o

calculate he fractional

penetration

of

particles

on various structures y filtration

forces

(Crawford

1976;Flagan

and Seinfeld 1988).

However, each requiresdetailedknowledge of the

geometry

of the collection

medium.

Because

the structureof dust cakes

s not

defined-and,

in fact, appears o be

systemdependent

n ways

that are now not

understood-none

can be used o

predict ractional

penetration.

Figure 10.2 shows actualbaghouse

enetration

ata reported

by Ensor et al.

(1981).

Considering

he dominantcollectionmechanisms, iffusion

(which

decreases

ith increasing

ar-

ticle size).and mpaction and nterception

which

increase

with particlesize), he classic heoreti-

cal

penetration

urve exhibits

a singlemaximum n the 0.

-

to 0.3-pm

range.

Thus, additional

mechanismsmust act to create he bimodal curve of Figure 10.2.A reasonable xplanation s to

attribute

he arge-particle

enetration

mode o bleed-through

Can

and Smith

1984a).

t

li

.-t



418

Chap. 0

Fabric

ilters

Figure

10.2

Measured

baghouse

ractional

penetration

From

Carr,

R.

C.,

and

Smith,

W. B.

IAPCA,34:79,7984;

reprinted

with

permis-

sion from

IAPCA)

Sec.1

Exampl

wh

Th

kg/

an

Soluti

S

Th

re

ti m

0.

0.0

\c

. 2

a

=

'5

?

LL

0.1

I

Particle

iameter;rm)

Dennis

and Klemm

(1979)

proposed

he

following

semiempirical

verall penetration

equation

based

on

the observations

hat penetration

ncreased

with

increasinggas-to-cloth

atio

and

decreased

s

he

dust cake

on

the fabric

grew:

Pt= Pt.\

+

(116

pt,\e-,,w

+

pt1,,

(

1 . 1 )

where

P/

=

overall penetration

Pl.

=

low-level

penetration

ue o pinholes

n

the

cake-fabric

tructure;

ncreases

with

the

gas-to-cloth

atio

P/0= penetrationhroughajust-cleanedhbricarea,

Prl,r

=

penetration

ue

o bleed-through;

function

of parlicle

morphology

w

=

dust

massper

unit

bag surface

area,

also

known

as areal

densitv,

ks/m2

r,r

=

cakepenetration

ecay

ate

Viner

et al.

(198'1)

ound

that Eq.

(10.1)

overpredicts

enetration

ust

after

a compart-

ment

hasbeen

cleaned:

ut.

as

he dust

cake

on

the ilter grows,

he

discrepancy

etween redict-

ed

and measured

alues

diminishes.

ll the

terms

n Eq.

(10.

)

must

be determined

rnpir ically

for

a

given

fabric-dust

ake

combination.

he

following

example

llustrates

ypical

con<Jitions

n

a reverse-air

ilter.

\ l

"6



Sec.

10.3 Fabric

iltration

heory

419

Example

10.1 Time

Dependency

of Penetration

in Baghouse

The

waste

gases rom a coal-fired

boiler

flow through

a reverse-air

aghouse

or

111'-ash

removal.

he bllowing

empirical

quation

or overall

penetration s a function

of operation

time between leaning ycleswasdeveloped:

p l

=

t60y r ' l

+ 0 . 0

t60y r ' ' r

exp { - t80c ;v r ) *

: . l u

where

V

-

gas-to-clothatio,

m/s

Ci

=

inlet dust

oading,

g/m3

I

=

operation

ime since

astcleaning

The baghouse

perates

t an air-to-cloth

atio

of 0.0

m/s and he

nlet

dust oading

s 0.004

kg/m3.

Eachcompartment

s

cleaned

everl'

20 min.

Plot

penetration

s a function of time.

andcalculatehe average verallpenetrationor thecycle.

Solution

Substituting

he

valuesof V

and C;

n the

penetration quation:

P/

=

0.0038

0.0963

xp

-0.432r ' )

where

' is the

ime n

minutes.

igure10.3

hows he

variation

n

penetration

ith t ime.

Most of the

particle

emissions

ake

place

shortly

after

he bags

arecleaned.

To calculate

he

average

enetration.ntegrate

ver a cycle

and divide

by the

ength

of the cycle.

20

i

o.oo:s

0.0963

"^p

o.+:zt

ai

)

Vrtnn

Bush

t al t1989)der ived

n

expression

o

quant i ty he

exture

r sur face

r lorphol-

..

:

1r.l\ .r.h

lntflr ':.

Their

morphology

actor

,41)

s a dimer.rsionless

uantitl based

n

rolu-

' - : : : - . . . ' r - , l l , t n l i r : t , , r ' r J . r t . L nt e a . L r r e d u i t h a C o u l t e r C o u n t er ( a d e v i c e t h a t m e a \ u r e \ p a n i c l e s i z e

- . : i

:

: i . i , . : , , . ' : r -

: . . i \ l r \ l t r

r .

pec i f i c

u r l ' ace

rea

A) .

and

hc

fue

den \ i t \o t ' t he

ash

' '

.

-

:

0.0

49

t

i

I

i

:

I

:

t

q{ffi&s,,



420

Chap."

0

Fabric

ilters

Figure

10.3

Penetration

ersus

time

in

a

reverse-air

baghouse.

(r0.2)

Sec.1

reglons,

a

tubesheet

and

the m

constant

F

tion

throu

continuo

mal ly

use

is cake

il t

the actual

to-cloth

r

tion

betu

pressure

T

maximur

designer

sure

drop

where

surface

u

i're:;

L

,,i

::

x

()

0.01

8 1 2

Time

min)

n;

D];

ln Dpi

u

=1"*v

n;D];

where

ni

=

the

otal

number

f

particles

ounted

n Coulter

hannel

Doi= thegeometricmean iameter

f

Coulter

channel

,

pm

A

low

value

of

M

indicates

hat

he

particles esemble

mooth

pheres,

hile

a

high

value

ndicates

hat

he

particles

ave

very

rregular

hapes

nd

surfaces.

or

perfect

spheres,

=

1.0.

Field experience

hows

hat

ashes

ith

ow values

f

M

(2.0

o

2.7)seep

hrough

he

abric,

increasing

missions

Felix

and

Menitt

1986;

Milleret

al.

1985)'

The

mechanism

fbleed-through

s not

well understood

et

and here

s

no

general heo-

retical

model o

predict

ts

effect

on

particle

penetrationhrough

baghouses'

t

is

fortunate hat

fabric

filters

properly

designed

ased

n

pressure

rop

considerations

enerally

xhibit

penetra-

tions

well below

he

applicable

SPS.

10.3.2

ressure

roP

There are

several

contributions

to

the total

pressure

drop

across

a baghouse

ncluding

the

pressure

drop

from the

flow through

the

inlet and

outlet

ducts,

from

flow through

the hopper

; - t

,l|



Sec.

10.3

Fabric

iltration

heory

421

regions,

and

from flow

through

he

bags.The

pressure rop

across

he bags

(also

called

he

tubesheet

ressuredrop)

is a complex

function of

the

physical

propertiesof the

dust

and

fabric

and

the manner

n

which the

baghouse

s designed

and

operated.

he duct

and

hopper

ossesare

constant

nd

can be

minimized

effectively

hrough

properdesign'

Fabric

filtration

is inherently

a batch

process hat has

been

adapted

o

continuous

opera-

tion

through

clever

engineering.

he dust collected

on the

bags

must be

removed

periodically

br

continuous

operation.

Shaker

and

reverse-air

esigns

are

similar

in the

sense

hat they

both

nor-

mally

use

woven

fabric bags

un at

relatively

ow

gas-to-cloth

atios,

and

he filtration

mechanism

is cake

iltration.

The

fabric merely

serves

as a support

or the

formation

of

a dust

cake.

which

is

the

actual

iltering

medium.

Pulse-jet

aghouses

enerallyuse

elt fabrics

and

run

with a high

gas-

to-cloth

ratio.

The f'elt

abric

plays a much

more

active

role in

the flltration

process.

The distinc-

tion

between

cake

filtration

and

fabric

filtration

has important

implications

for calculating

he

pressure rop across

he

filter bags.

The design

of a

fabric filter

begins

with a

set of specifications

ncluding

average

and

maximum

pressure rop, otal

gas 1ow,and

other

requirements.

asedon

these

equirements,

he

designer

pecifies

he

maximum

face

velocity

allowed.

The standard

way

to relate

ubesheet

res-

suredrop

o

cas-to-cloth

atio

s

^P( r )=(dY

(

1 0 . 3 )

where

AP(r

)

=

the

pressure rop through

he

filter, a

function

of time,

t

S(r

)

=

the

drag hrough

he fabric

and cake,

Pa-s/m

V

=

the average

r design

gas-to-cloth atio,

m/s

The

drag across

he

fabric and

cake

s a function

of

the amount

of

dust

collected

on

the

surface

f the

bags.

Assuming hat

he

filter drag

s a

linear

unction

of the dust

oad,

s(r)=

.

+ <zw(t)

(

10 .4)

ri

here

S"

=

the drag

of a

dust-free

freshly

cleaned)

ilter

bag

Kz.= the

dust

cake low

resistance,

-l

The dust

mass

as a function

of

time

is W

=

C;Vf,

where C;

is the

inlet dust

concentratlon

,kg/m3).

Substituting

his expression

n Eqs.

10.3)

and

10.4):

A P ( r )

S " V

+ K ; C ; V 2

(

1 0 . 5

The constants

,

and

K; depend

n the

fabric and

he

nature

and size

of

the dust'

Th e



422

Fabric

ilters

relationship

between heseconstantsand the dust and fabric

properties

are not understood

well

enough o

permit

accurate

redictions

and so must be determined mpirically.

Example 10.2Estimation

of

Parameters n Filter

Drag Model

(a)

Estimate he

parameters

n Eq.

(

10.4)

basedon the following data or a freshly cleaned

fabric. The

gas-to-cloth

atio is

0.0167

m/s and he nlet

dust concentration

s

0.005

kg/m:.

(b)

Estimate he

pressure

rop for

the sameconditionsafter 70 min of continuous peration.

Test Data

Time

(min)

AP

(Pa)

9.00

22.80

30.30

36.60

41.40

-59.40

0

0.025

0.050

0. r00

0.150

0.030

Sec

l

Exam

Solut i

Chap.

0

0

5

l 0

20

30

60

r50

380

50 5

6 1 0

69 0

990

Solution

(a)

Basedon the testdata,

generate

plot

of filter drag versusarealdensity.The

following

table llustrates

he data o be

plotted.

S = LP/V kPa-s/m) W =CiVt (kg/rn2)

\

:

Figure 10.4showsan nitial characteristic urvature ollowed by a linear dependency f fil-

ter

drag on

areal

dust density.The nonlinear

portion

of the curve s due o a

non-uniform

initial flow

through he

fabric. The previous

cleaningcycle usuallydislodges he

cake n

,'



Sec.

10.3

Fabric

iltration

heory

S = 2 . 1 . 5 6 8 +

l 5 . 7 1 W

423

60

30

l 0

a

-o

l t

Figure

10.4

Filter

drag

versus

dust

areal

density

or

Example

10.2

|

. | |

|

|

| r I

r

I

l:

lT_rr_rt_rrrT,T-r1

0 .05

0 .

0 . l . s

0 .2

0 .25

0 .3

0 .3 -5

Areal usr cnsiry'(W.g/ml

irregular

hunks.

earing

someparts

ofthe

bagvery

clean

and

others

ti l l quite

dusty,

resulting

n

spatial

ariations

n

the

nit ial

gas

low.

A least-squares

it

of

the inear

portion

of

the

curveyields:

S"

=

21.51

Pa-s/m,

nd

K2

=

I 15.7

pa_s_m/kg

l. l6 x

l0s

s

r.

(b)

Equation

10.-5)

ives

he pressure

rop

after

70

min

of

continuous

peration:

p

=

24,510(0.0167)

( l . tU

t

1gs;(0.00-5)(0.0167)r(70X60)

1.090

a.

Example

10.3

Cycle

ime

for

Reverse-Air

Fabric

Filter.

A reverse-air

abric

llter

has

1,000

ml of

f i l ter in-s

rea

and reats

0 m3/s

f

air

carrying

dust

concentration

f

0.005

kg/m3.

Assume

S,, lg.g

kpa-s/m

nd

K2

=

I

.0

x

l0s

s

l.

I f the

filter

must

be

cleaned

hen

hepressure

rop

eaches

.0kPa,

after

what

period

must

he

c leun ing

ccur?

Solution

Thegas-to-cloth

at io

s

11

1971.000

0.01

m/s.

Solving

q.

10.-5)

br

the

irne

f

opera-

t ion

befbre

he

pressure

L'op

rceecls

.000

Pa.

=

[2.000

(20.000)(0.01) l / t (105)(0.005)

(0 .01 ) r l

36 .000

=

l [ )

h

sjffiknr



424

Chap.

0 Fabric ilters

Pulse-jetbaghouses ave been designed o operate n

a

variety

of modes.Some

remain

on line at all times and are cleaned iequently.

Othersare aken off line for cleaningat relatively

long intervals.A

complete

model

of

pulse-jet

iltration

therefbremust account or the composite

dust-fabric iltration occun'ingon a relatively clean bag. the cake filtration that result fiom

pro-

longedperiods n ine.and he ransition eriodbetweenhe wo regimes.

If a cornpartments takenofl ' l ine fclr

cleaning. he dust hat s removed rom the bags

falls into the hopperbefbre orward

gas

low resumes. f

a comparlment s cleanedwhile on

line,

only a small fi-actionof the dust removed iom the bag falls

to the

hopper.The remainderof the

dislodged ustwill be redeposited

i.e.,

recycled")

n the bagby the brward

gas

low. The

rede-

posited

dust layer has clifl 'crent

ressure

rop characteristic shan the freshly deposited ust.

Dennis

and Klemm

(

1979) roposed

he bllowingmodelof

dragacross

pulse-jet i l ter:

s

=

s,,

(xr),w,.

K2w11

(

10.6)

where

(K1),=

specif ic ust esistancefthe recycling

ust

W.

=

arealdensityof the recycling

dust

K2

=

specif ic ust esistancef

the

ieshly

deposited ust

Ws

=

arealdensityof the freshly deposited

ust

Th is

mode l can eas i l y accoun t o r

a l l t h ree eg imeso f f i l t r a t i on n a

pu lse - je t

baghouse.The

ressure

rop can thus be expressed s the sum

of a

relatively

constant

erm and a

term hat ncreas es ith dustbuilduo.

Sec.

10

Eranrp l

1 -

Solut ion

E r a m p l c

and

RifT

\

- . :

.:

-

' . :

:

where

(

0.7)

(

10 .8)

a p = ( p g ) r "

+ K 2 w o v

(pst,,

is"

+

xz),.w,lv

The

disadvantagefthe model epresen ted

y Eqs.

10.7)

and

10.8)

s that he constants

S",

K2,

and W, cannotbe

predicted

at this

time. Consequently, orrelations f laboratorydata

must

be used o determine he value

of

(PE)1,,

For one fabric-dus tcombinationof Dacron felt and

coal ly ash.DennisandKlemm

(

1980)

evelopedhe ollowingcorrelation:

(f'f)o,,

=

1,045

r,

n ot

(PE)r",

is in kPa

P;

-

gaugepressure

f the cleaning

pulse,

n kPa

Y = gas-to-clothatio, n m./s

(

0.9)



Sec.

10.3

Fabric

iltration

heory

425

Example

10.4

Tubesheet

Pressure

Drop in Pulse-Jet

Baghouse

A

pulse-jetbaghouse

sesDacron

elt bags

or the control

of fly-ash

emissions.

he

gas-to-

cloth

atio s 0.02,1

m/s and he

nletdust

oading s 0.01

kg/m3.

The bagsare

cleaned

t l0-

min ntervals

with

pulses f air at 650

kPa

gauge' f K:

= l'5 x 105 l' estimatehe maxi-

mum ubesheet

ressure

rop.

Solution

The conditions

resimilar

o those

or whichEq.

(

10.9)wasdeveloped.

hen,

PE)a"

-

(

1,045)(0.024)(65Ofo

:

=

0.372

Pa.At the

endof the

l0-min cycle,

4ze

(0.01X0.024)

(600)

=

0.144

g/m2.

Equation

10.7)

gives

he ubesheet

ressure rop:

AP

=

0.312

+

(

L5

x

1

05)(0.

44)(.0.024)

t.000

=

0.89

kPa.

Example

10.5

Effect of

Filter

Inhomogeneities

n

Filter

Drag Model(Cooper

and

Riff

1983)

Knowing

the mean

of a

property,suchas

dust arealdensity,

may be misleading

br

predict-

ing

the effect of

that

property.

When

estimating

he

specific

dust resistance

or a filter cake

(Kz)

fiom

pressure

rop

data,Eq.

(

10.7)

predicts hatK2

n

W

|

. However,

n a typical

bag-

house he

arealdensity

s not

homogeneous,

ut changes

rom

bag o bag.

For example,

Ellenbecker

.1919)

measured

realdensities

n a

pulse-jet

baghouse

nd ound

that he

data

could

be fitted

by a log-normal

distribution

with a median

of 0.64

kg/m2and

a

geometric

standard

eviat ion, * ,

of

1.15.

In the

presence

f

inhomogeneities,

he

average

alue of l{

shouldbe

used

br estimating

K2.

The

proper

average

o use

s the harmonic

nrcon.not he arithmeticmean.For a log-nor-

mal distribution.

he

ratio of the

harmonic

o the

reciprocalof

the arithmetic

mean

s

=

exp

(h206')

(

10 .0 )

where he

harmonic

verage

br a og-normal

istribution

s:

g

l

=

I i /<11

xp

{-0 . :

tn

o .

)

l

r 1 0 , I I t

W :

- l

w

^b



Sec.

10.4

Practical

esign

onsiderations

Toble

0.l

Gos-to-Cloth

Rotioso

cm/s)

427

Shaker/Woven

Reverse-AirAVoven

Pulse-Jet/Felt

Alumina

Asbestos

Cocoa,chocolate

Cement

Coal

Enamel

frit

Feeds, rain

Fertilizer

Flour

Fly

ash

Graphite

Gypsum

Iron

ore

Iron

oxide

Iron

sulfate

Leather

dust

Lime

Limestone

Paint pigments

Paper

Rock

dust

Sand

Sawdust

Sil ica

Soap,detergents

Starch

Sugar

Talc

Tobacco

Zinc

oxide

1 . 2 7

t . 5 2

1.42

1.02

l .2 l

1 . 2 7

1 . 7 8

t . 5 2

t . 5 2

1.02

1.02

1 . 0 2

t . 5 2

1.21

1.02

1.78

1.21

1 . 3 1

t .27

r . 7 8

1 . 5 2

1 . 2 1

1 . 7 8

1.21

1 . 0 2

1 . 5 2

1.02

t .2 l

t . 7 8

r . 0 2

4.01

5.08

6 . 1 0

4.07

4.01

4.51

' 7 . 1 1

4.01

6 . 1 0

',

<A

2.54

5.08

5.59

3.56

3.05

6 . 1 0

5.08

4.01

3.56

5.08

4.51

5.08

6 . 1 0

3.56

2.54

4.01

3.56

5.08

6.61

r 5 J

a Generally

afe

design

alues.

Source: rom

Turner

J

Reprinted 'ith permissionrom JAPCA.

H. e t a l .JAPCA.

i7 :7-19987)

t



Chap:

0

Fabric

ilters

Sec.

Solutio

Tab

clor

l \

1 -

Co

\ , .

j e : :

10.4.2

of che :-.

chent i ; . .

k n i | .

r .

. .

Oh

pL l : t

; r

membr ' :

act ion :

th is r

r r '

_

Cott tn : - ,

t he

n r . . . : :

t a b r i c : : .

428

Solution

Table

10.

1

gives

v

=

Q/A,

=

2.5,1

m/s

br

filtration

of

fly-ash

n

pulse-jet

ilters'

Then,

he

netcloth

area

s

A,

=

(13

6)/(0'0254)

930

m2 '

For

conttnuouslv

operated

shaker

and

reverse-gas

leaned

ilters,

the

area

must

be

increased

o allow

tor

snuitin_l

lown

of one

or

more

compartments

or

cleaning

and

maintenance'

A typical

compartmenrr-,r.,

*.gr

0.3

m

in

diameter

nd

10.7

m

long,

with

the

number

of

bags

per

compartment

anging

rom'10

L O+S.

able

10.2

provides

a

guide

or adjusting

he

net

area

o

the

grossarea

br

these

\\

o t)'pes

of

baghouses.

ecause

ulse-jet

ilters

are

cleaned

on

line'

no addi-

tional

flltratron

area

s required,

and

he

net

and

gross

cloth

areas

re

equal'

t . 5

t .25

t . 1 1

1 . 1 2 5

1 . i l

1 . 1 0

1.09

1.08

1.01

1.06

1.05

1.04

Toble

10.2

Guide

o

Eslimofe

hqker

And

Reverse-Air

oghouse

Gross

Cloth

Areo

Net Cloth

Area

(m2)

Multiply

Net

Area

by

I

-170

z

3 7 1 - 1 , 1 1 5

t,16-2,230

) ) ? r - 1 1 5 0

3,351-4,460

4,461-5,580

5,58

-6690

6691-7 ,810

7,811*8,920

8921-10,040

t0,041-12.210

t2,211*16,130

> 16,730

ted

with

Permission

rom

JAPCA

Example

10.7

Gross

Cloth

Area

of

Reverse'Air

Baghouse

Estimate he grosscloth area equired or the conditionsof Example10'6 f a reverse-atr

baghouse

s

sPecified.

Exanrpl

l - .

: -_

So lu t io

T l -

-

l : - :

c u . ;

\ I ; : -

-



Sec.

10.4 Practical esign onsiderations

429

Solution

Table 10.

gives

V

=

I .02cm/s

or fly-ash iltra tion n reverse-air aghouses.

hen. he

ne t

clotharea

equireds A,

=

(23.6X0.0102)

2,314

m2.From Table 10.2, he

sross

lotharea

i s

2 , 3 1 4 ) ( 1 . 1 7 ) =

. 7 1 0 r .

Comment

Notice that he reverse-air aghouse

equiresalmost hree imes as

much areaas he

pulse-

jet

to

process

he same

lue

gas

low rate.

10.4.2 iltermediq

The ypeof filter material sed n baghousesepends n the specific pplicationn terms

of chemical

composition f the

gas,

operating emperature.

ust oading,

and the

physical

an d

chemical

haracteristicsf the

particulate. varietyof fabrics. ither

eltedor woven

sometimes

kni|,

is

available.

he selection f a specif icmaterial.

weave. inish,or weight

s based

rirnarily

on

past

experience.

or somedilficult applications, ore-Tex.

polytetraf'luoroethylene

PTFE)

membrane aminated o a substrate abric

(f'elt

or woven) may be used.

Becauseof the

violent

action of mechanical hakers,

pun or heavy weight staple

yarn

fabrics are commonly

r.rsed.r th

this ypeof cleaning. ighter-weight

lament

arn

abricsareused

with reverse-airleaning.

The ypeof

materialwill limit the maximumoperating

as

emperature

br the baghouse.

Cotton abric s among he

east esistanto high em peratures

about

355

K) while

f iberglass s

the most esistant

about

530

K). Table 10.3

gives

he

maximum emperature

imits of the

eading

fabric materials.

Example 10.8 Fabric Filter for

Open-Hearth

Steel

Plant

(Licht

1980).

The flue

gases

rom an open-hearth

teel

plant

low at the

rate of I l0 m3/s

at 1,000K and

l0l .3 kPa with an ron oxides

articulateoadingof 0.0026

g/m3.The

watercontent f the

gases

s

87c.

Designa tabric i l ter

system o reduce he

particulateoading

f the -l lre

a'es

to thecorresponding

SPS.

Solution

The SPS 1'or

l r t iculate

r l is: ions

ronrsteel

lants

s 50

mg/dscm

:ec' . ih, le .1r .

t r

c l l -

;u l . i te heoreral l err t , r l . ' f ic iencr er lu i led.( ) l lectheactual ar t icLr l r teo.rJinStrdr 'y.

. i . r nJ .L rJ . , r n . i i t r i . r : .

1 .6 (

r

l . i l t

t r 0

t ) l r r l - - l r ]=

10 .3 - i l

t s /d \ c l l t

I t c , r e r r i l l i i i e

enc r



430

Chap.

0

Fabric

ilters

Toble

10.3

Properties

f

Leoding

Bqghouse

Fobric

Moteriols

Fabric

Temp.u

K)

Acid

resistance

Base

resistance

Sec .1

that

dias

.\

ca1

( l l t t

a t I '

l 0 t

/ { . r .

n i h : :

l l f i . :

T; -

i .

i .

: '

n'.-

-

{ r j ' -

r -.:

Cotton

355

Creslanb

395

Dacronc

41 0

Dynel'

Fiberglasd

Filtron'

Gore-Text

Nomexc

,165

Nylonc

36 5

Orlon'

400

Polypropylene

36 5

Teflon' 505

Depends

n backing

Depends

n backing

3,15

-530

-105

Dependsn

backing

Poor

Good

Good

Excellent

Fair-good

Excellent

Fair

Fair

Excellent

Excellent

Excellent

Very

good

Very good

Very

good

Fair

Fair-good

Very

good

Excellent

Excellent

Fair-good

Excellent

Excellent

Wool

365

Very

good

Poor

"

Maximum

continuous

operating

emperature.

b

American

Cyanamid

egistered

rademark.

c

Du Pont

registered

rademark.

d

Owens-Corning

iberglass

egistered rademark'

e

W.

W. Criswell

Division of

Wheelabrator-Fry

nc.

r

W

L. Gore

and Co.

registered

rademark.

Source:

From

Turner J.

H. et

al.JAPCA,

37

749

(1987).

Reprinted

with

permission

rom

JAPCA.

is 11,

=

(10,352

50y10,352

0.9952

99.52E;).

his eff iciency

s within

the imits

achiev-

able

by a

properly designed

abric

filter. However,

he

temperature

f the

gases s well

above

he operating

imits of available

abric

materials.

A

precoolingsystem

s required,

ut

a fabric

to

operate s

hot as

possibleshould

be

sought.

Table

10.3shows

hat

iberglass

agscan

operate

ontinuously

at temperatures

p

to 530

K. Previous

experience

hows

hat

iberglass

ilters

can remove

ron oxides

dust with

effi-

ciencies

higher

han 99.67o

n sonic-assisted,

everse-air

aghouses

at

an

air-to-cloth

atio

of

| .21 cm/s

(Licht

1980).Choose

iberglass

agsat

an operating

emperature

f

530

K.

There

are,

basically,

hreealternatives

or cooling

he

gases o 530

K:

(l)

dilution

with

ambient

alr;

2)

evaporation

f a water

spray

nto

the

gas and

(3)

cooling

n

a waste

heat

boiler.

Dilution

with ambient

air will

more

han riple

the

filter area

equired,

while

evapo-

rative cooling

will increase

t by

about30Vo

see

Problem

10.

0). They

both

waste

he

pre-

cious

hermal

energy

contained

n the

gases.A waste

heatboiler

recovers

lose

o 507o

of

t : :

l i



Sec.

10.4

Practical

esign

onsiderations

431

that

energy

with

no penalty

n

terms

of

filtering

area

equired.

igure

10.5

s

a schematic

diagram

of

the proposed

ystem.

Assume

hat

he

waste ases

ehave

ike

air

(p

=

it.:S2+

kg/m3,

Cp

=

l.0g

kj,/kg_K)

nd

calculate

he

enthalpy

hange

hey

experience

n going

iorn 1,000K to 530 K: AH* =

( l l 0 ) (0 .3524) (1 .08 ) (530-

1 .000)

-19 ,671kJ .Theenrha lpychangeo f

l kgo f

l i qu idwa te r

at

298

K

that

s

convertecl

o saturated

ream

t 445

K

(g30

kpa)

s Z,1AS.+

J.

Assurning

10Vo

nerf,y

osses,

he

amount

f sleam

roducecl

n

rhe

vaste

eat

boiler

s

(0.9t(19.677i

l(2'665'4)

=

6.64

kg/s

or 23,900

g/h.

To

estimate

he

heat-transfer

rea,

alculare

he

oga-

rithmic

mean

emperature

lifference:

LMTD

=

[(

1.000

4.15)

(-530

2gL]/1n(55-5/23t)

370

K.

Assume

hat he

overall

heat-transfer

oefTicient

s

300

Wm2-K.

The

heat-transfer

area

s

(19,677) l [ (0.3)( .310))

t ] t

mz.

The

gas

low

rate

entering

he

baghouse

s

(110X530/1.000)

-5g.3

m3/s;

he dust

oading

is

(0.0026)(

l0/58.3)

=

0.005

kg/m:.

For

a

gas-to-cloth

atio

of l .2l

cm/s,

hener

clorh

area

is

58'3/0 '0121

4.590

m2.

From

Table

10.2.

hegross

loth

area s

(

. l l

) (4,590)

5.

00

m2.

A

reasonable

esign

might

use 0 compartments.\'ith50 bags ercompartment ac h

0.3

m

diameter

nd

10.7

m long.

The

length

of the

filtration

time

depends

n the

nraximum

allowable

pressure

rop

through

he abric-dustcake.

ssume

hat.

or

thrs

case.

he

maximum

s

2.,5

pa

For

ro n

oxides

dust

on

fiberglass.

he paramcters

n

the firter

drag

moder-

Eq.

(10.5)-are

s"

-

142

Pa-s/m

nd

K2

=

l.2l

x

t06

s

I

(Lichr

9g0).

Then,

.5

=

(142)(0.012j)

(1.21

106X0.005X0.0121

21/1,000.

Solving,

='713

s

=

l2

minutes.

Saturated

team

.1'15

K

Water

298

K

Iron

oxides

dust

Figure

10.5

Vaste-heat

oi ler

of

Example

0.E

Waste

gases

1.000

K



432

Chap.

0

Fabric

ilters

IO.5

COSTS

F

FABRIC

ILTRATION

Yourobjectivesn studying hissection re to

1.

Estimate

he

equipment

ost

of

the

most

common

baghouse

ypes'

2. Estimate

he

corresponding

CI .

3.

Estimate

he

TAC

associated

ith

fabric

filtration'

The

annualizecl

ost

of owning

and

operating

a

baghouse

an

be

very

high.

The

capital

cost

fbr

a fabric

f i l ter

system

can

be estimated

based

on the

gross

cloth

area'

The

cost

also

depends

n

the

type

of

baghouse,

whether

t is

made

of

stainless

r

mild

steel,

and

whether

or

not

it is

insulated.

Furthermore,

tandardized

modular

baghouses

re

ess

expensive

han

custom-built

units.

10.5.1

EquiPment

osf

Baghouse

equipment

costs

consist

of

two

components:

he

baghouse

unit

and

the

bags'

Both

costs

are

unctions

of the

gross

cloth

area,

Ac.

The

baghouse

nit

cost

s'

in turn'

comprised

of

the

basic

baghouse

ost

and

he

costs

of

"add-ons"

or stainless

teel

and

nsulation.

The

prices

for the

basic

unit

or

add-ons

are

inear

unctions

of

the

grosscloth

area

Vatavuk 1990),

E C

=

a +

b A ,

i l 0 . 1 2 )

where

a

and

b are

regression

onstants

isted

in

Table

10'4.

Notice

that

the

parameters

n Table

10.4

apply

only to certaingrosscloth area anges'

Table

10.5

gives selected

ag

prices

(in

$/m2)

br

pulse-jet,

shaker,

and

reverse-air

ag-

houses.

The

pulse-jet

bag

prices

n

Tabie

10.5

do

not

include

he

prices

of

protective

cages'

Mild

steel

cages

will

add

approximately

$13/mz

of

filter

area

June

1990

dollars).

Stainless

teel

cages

will

cost

about

$32lm2.

Example

10.9

Equipment

Cost

of

Reverse-Air

Baghouse

Estimate

he

cost

of the

reverse-air

aghouse

f

Example

10.8.

The baghouse

hell

material

can

be

mild

steel.

nsulation

s

required.

The

bags

should

be

0.3-m

diameter

with

sewn-in

snap ings.

Sec.

10 .

Toble

0

Baghouse

Shaker-inl

Shaker-:.,

Pulse-1et

Pulse-jet-

Reverse-

Custom-b

u

Prices

Reprinte

Solutio

For

me

Ba

Inr

Fo r

B a

,f



Sec.

10.5

Costs

f Fabric iltration

Tqble 0.4 Porometers

or FobricFilterEquipment

Cosfso

433

BaghouseType Area Range m2

Component

a

Shaker-intermittent

370-1.-500

Shaker-continuous 370-5,600

Pulse-jet-common-housing

370-

500

Pulse-jet-modular 370-

I

,500

Reverse-air

930-7,500

Custom-built-alltypes

9,300-37,200

Basic ni t

stainless teel

Insulat ion

Basicunit

Stainless

teel

Insulation

Basic

n i t

Stainless teel

Insulat ion

Basic

n i t

Stainless teel

Insulation

Basicunit

Stainless

teel

Insulation

Basic ni t

Stainless

teel

Insulat ion

4,t20

14,000

2,200

43,800

29,100

0

l l ,280

t2 ,100

1.610

55.

40

29,300

3,500

34.200

r6.500

1,320

263,000

108,400

70.200

8+.6

l l . 9

.5.7

9 3 . 8

6 l . l

1 r 1

69.8

59.

n .1

92.0

81.1

2 6 . 1

88.i)

68 .5

r 0 . 0

69.3

28.1

8.0

o

Prices pdatedo June1990. ource:

rom

Turner .

H.

et al .JAPCA,37:1

05 1987).

Reprintedwith

permission

iom JAPCA.

Solution

For a reverse-airaghouse i th 930

<A,.<

,-500

2,Table 10.4

gives

he

bllor.ing equip-

ment costs:

Basic

unit: 34,200

(8UX5.100)

$483,000

Insulat ion dd-on:

.320

(

10X5.100)

S-52.320

For 0.3-mdiameter

iberglass ags

.r

th rings.Table

10.-5

sir

s

B a s s o s t :

1 2 . 2 ) ( 5 . 1 0 0 i =6 l . l l 0



434

Chap.

0

Fabric ilters

Sec.

Tobfe10.5

SelectedBoq Prices

$

in

June l99o/mz)

Pulse et

Shaker

Reverse

Air

Material"

TRb

BBR.

Strapd Loop.

w/Ringsr

w/o Rings

PE

PP

NO

FG

TF

CO

5.1-13

5.5-7.-5

19.2-23.1

13.3--5 .9

83.8-

l

1

NA

-+.0-4.6

.1 .1J .9

r4.6-

.9

r

1 . 7 - 1 5 . 3

82.7-108

NA

5.7-5.8

NA C

20.1-21.2

9.3-12.2

NA

NA

4.0

NA

14.3-14.8

6.5-8.s

NA

NA

Solut i

T

l i

B

l n

r )

Il.l

10 .5

\ \ \ : .

-

Eram

So lu t i

5.6

5.9

15.7

NA

NA

5.5

5.1

s.6

14.5

NA

NA

4.8

a

Materials:

PE

=

polyester;

PP

=

polypropylene;

NO

=

Nomex; FG

=

Fiberglass

with

10%

Teflon; TF

=

Teflon

felt;

CO

=

cotton.

u,.

Bag removal

method:TR

=

top bag

removal;

BBR

=

bottom

bag removal.

d'"

Bag top

design. Prices

are

given

for

bagswith

and without sewn-in snap rings. c

Not applicable.

Ranges

shown reflect

different bag

diameters. ulse-jet

ag

diameters:

.1-0.2 m; reverse

air:0.2-0.3

m; shaker:

.13 m.

Source:

rom Turner

J. H.

et al . JAPCA,37:1105

1987).

Reprinted

with

permission

from

JAPCA.

Example

10.9

Continuation

The

equipment

ost s EC

=

483,000

52,320

62,220

$598,000

n

June 990.

Example

10.10

Cost of

Pulse-Jet

Baghouse

or Municipal

Solid Waste

(MSW)

Incinerator

The air

pollution

control

systemof

a

MSW

incinerator

onsists

f a spraydryer ollowed

by

a

pulse-jet

abric filter

(Frame

1988).The

flue

gas

entering

he baghouse

lows at the rate

of

35 m3/s

at 500 K and

1 atm and

contains ignificant

amounts

of HCl. A design

gas-to-cloth

ratio

of

0.025 m/s s specified

when

usingTeflon f'elt

bags,which are

extremely esistant

o

the harsh

environment revailing

n

the baghouse.

stimate

he

fabric

filter equipment

ost.



Sec.10.5

Costs f Fabric iltration

435

Solution

The net

iltrationarea s

(35)/(0.025)

1,400

m2.Assuming hat hecleaning s done

on -

line, the

gross

iltration

area s also 1,400

m2 and a commonhousing

design s appropriate

Because f

the corrosiveconditions

and high temperature, tainless

teelconstruction nd

insulationare required.

The diameterof

the bags s 0. m, and hey

are of the top removal

type. Stainless teel

cagesare also equired.

The following table summarizes

he disburse-

ments hat comprise

he equipment

ost

or

this system.

Item

Cost

$

n June1990)

Basic

unit

Stainless

teel

add-on

Insulation

dd-on

Teflon f-elt

bags

Stainless teel ages

Equipment

ost

108,580

95,140

18,050

111,320

44,800

384.190

r0.5.2 CI

The total

capital nvestment s

estimated rom a series f factors

applied o the

purchased

equiprnent

ost o obtaindirect

and

ndirect

ostsof installatio n.

he TCI is the sumof these hree

costs. able 10.6 ives

he requiredactors.

Example

10.11 TCI for Reverse-Air

Baghouse

Estimate

he otalcapital nvestment

or thebaghouse

f Examples 0.8 and 10.9.Assume

that no

specialsite

preparation

r buildingsare needed. he

cost of auxiliary equipment,

excluding

hewasteheatboiler, s

$150,000.

Solution

From

Example10.9, he

equipment ost or thebaghouse-includ ing hebags is

Sr98.000. he

costof thewaste eatboiler

o

produce

3,900 g/h

of steam aturated t

r-+-5

(830

kPa)mustbe estimated. ased

n data

published

y Peters nd

Timmerhaus

199) . thecos to f

hebo i le r fb r s teampressuresup to.800kPacanbees t i r na ted f rom:

-

-

l

I

l \

l r

t \



436

Toble10.6

Copitol Cost

Foclors or Fobric

Filters.

Chap.

0 Fabric ilters

Sec.1

Costs

Factor

Direct

costs

Purchased quipment ost

Fabric ilter,

plus

bags.

lus

auxil iary

equipment

Instruments ndcontrols

Taxesand ieight

Total

purchased

quipment ost

(B)

Installation irectcosts

Total direct costs

Indirect

costs

Totalcapital nvestment

A

0 . 1 0 4

0.08A

1 . 1 8

0 . 1 2 8 + S P o + B l d g b

1 . 1 2 8 + S P + B l d g

0.45

B

2 . 1 1 + S P + B l d g

shiti. \1.r

andNer. '

i n q l r i e , r '

1987b.

drop

thrrr

work

an.l

appro\rn

drag

coei

number

.

gas-to- .

is

abour

a shaker

w

here

boi ler e;

k P a .

r : .

recr ie;

ther r l :

.

r e p l a c r l

i l D C C . . : l - .

t h e u i r : .

Eramp

" l

* #

u

SP

=

sitepreparation.. Bldg

=

buildings.Source:From TurnerJ. H. et al.JAPCA,

37:l105

(1987).

Reprinted

with

permission

rom

JAPCA.

nc

=

qo(,;r)o

a

l,4oo< ,it,

ksh

<

18o,ooo

( 1 0 . 1 3 )

where

EC

=

cost

of

the

boiler,

n

June

1990

m

=

rate of steam

generation,

g/h

The costcalculatedwith Eq.

(

10. 3) is for

a complete

package

oiler

plant,

which includes

feed-water eaerator, oiler f-eed umps,chemical njection system,and stack.For a rateof

steam

roduction

f 23,900 g/h ,Eq.

(10.13)

ives

a boilercost

of

$190,500.

rom Table

10.6,

=

598,000

190,500

150,000

$938,500.

he

purchased

quipment ost s B

=

I 18A

=

$1.107,400.

CI

=

2.11B

=

$2,403,000

in

June1990).

r0.5.3

AC

Direct annual costs nclude operatingand supervisory

abor, replacement ags,

mainte-

nance abor and materials,

utilities, and dust disposal. Indirect annualcosts nclude capital ecov-

ery,

property

ax, nsurance, dministrative osts,and overhead.

Typical operating abor requirements re 2to 4h

per

shift

for

a wide

rangeof filter sizes.

Supervisory

abor s

taken as

157o

of operating abor. Maintenance abor varies rom I to 2 h

per



Sec.10.5

Costs f Fabric il tration

437

shift.Maintenance

materials

ostsare

assumedo be

equal o maintenanceabor

costs

Vatavuk

and

Neveril

1980). he rr-rajor

eplacenent aft

tems

are ilter

bags,

which

have

a typicaloperat-

ing life

of

2

yr.

The bag replacement

abor s

approximately

2.00/m:

of bag area

Turner

et al.

r

987b).

Electricpower is required o operatesystem 'ansand cleaningequipment.The pressure

drop through

he system s

the sum of the

pressure

rop

through he baghouse

tructllreand

duct-

work

and he

tubesheet

ressure

rop. The

contribution

of the baghouse

tructureand ductwork rs

approximately

kPa. The

tubesheet ressure

rop is estimatedwith

Eq.

(10.3)

when

the f l l ter

drag

coefficients

are

known.

Cleaning

energy or

reverse-air ystems

an be calculated iorri

the

number

of compartments

o be

cleaned t one

ime

(usually

ne.sometimeswo),

and he re\ierse

gas-to-cloth

atio

(tiom

one o two times

he forward

gas-to-cloth

atio). Reverse-air ressure

rop

is

about1.7kPa.The

reverse-airfan enerally

uns

continuously.

ypicalenergy

onsumptionor

a shaker

leaning

ystem anbe

estimatediom

(Turner

t al. 198.1b):

l A /

- A

( z t A - 5 , r

r r

\

-

u . J

-

t v n l

( r0 . r4 )

wnere

lV.

=

power.

n kW

A.

=

grosS

loth area. n ml

Cooling

process

-sases

o acceptable

emperatures y water

evaporation r in

a

waste

heat

boiler require

plant

water. Pulse-jet

ilters

use compressed

ir at

pressures

f about -500 o 800

kPa.

Typical

consumptions

about2 standard

3/1,000

ctualm3. f the collected

ustcan not be

recl 'cled

r sold, he

costs br final

disposalmust

be ncluded.Disposal

ostsaresirespecific.

ut

they are ypically

about

$30 to

$50/metric

on including

ranspofiation.

Capital recovery

costsare basedon

the total depreciablenvestment

ess he

cost of

replacing

he

bags.The if'etime

f the abri c ilter

system s typically20

yr.

Property

axes. nsur-

.rncc.

ord administrative

hargesare about 4c/c

f the TCl. The

overhead s calculated

as 60% of

ihe sumof al l laborand he maintenance aterials.

Example

10.12TAC for Reverse-Air

Baghouse

Estimate

he TAC for the baghouse f Example

10. 1. The systemoperates ,640h/;-r.The

operatingaborrate

s

$12lh,

he maintenance

aborrate s

$15/h.

The costof electricitl s

$0.06/kW-h,

process

water

costs

$0.26lmetric

on. The

dust

recovered

an

be returned o

the

process

t

no

additional ost.The recovery

redit

br

the

waste-heat

team

s

Sl/nretlic on

The

pressure

rop hrough

he

uasteheat

boiler s 2 kPa.The minimumattracti\e ateof

return s lZc/clyr',

he sr:tenr

it'etine

s l0

vears.

he rnechanical

fficiencroi

the ans s

65%.

ilhrffrid',

A



438

Chap.

0 Fabric ilters

Solution

Table 10.7

gives

he direct

and

ndirect

annualcostsas calculated rom factors

given previ-

ously.To calculatehe annualized

ostof

replacing

heba-es,

he otal

cost

of

replacement,

including abor, axes,and reightmust be estimated. he cost of purchasinghe bags s

multiplied by a factor ol 1.08

o account

or

taxesand reight. The

cost

of replacementabor

is

estimated t a

rate

of

$2/mr

of filtering area.The

capital ecovery actor for a 2-yr

life-

t imeand

=

O.l2lyr s 0.592.

To

calculate

he

electricitycost o operate he main fan,

the total

pressure

rop through

the system

must

be estirnated.rornExample10.8,

he maximum ubesheet

ressure

rop s

2.5 kPa.The otal

pressure

rop

ncluding

he waste eatboiler, s

6.5

kPa.Assuming

ha t

the fan is locatedat the entrance

fthe baghouse,he

gas

low rate

hrough he

fan is 58.3

m3/s. herefore,he

power

o opelate t is

(58.3)(6.5X0.65)

583 kW. A reverse-air-a n

will

operate ontinuouslyo cleanone

out

of

10)

compartment t a time.The reverse

as-

to-cloth atio is equal

o the

forward

gas-to-cloth

atio: 0.0127m/s.The reverse

ir

f-low s

(5,100)(0.1X0.0127)

6.48mr/s.The

corresponding

ressure

rop s 1.7kPa, herefore,he

power o operatehe everse-airan s (6.18)(1.7X0.65) 17 ryy.

The capital ecovery actor

(0.

34)

for

the depreciablenvestment orrespondso a life-

t ime

of 20

yr

anda minimumattractiveate

of

return

of 12 Volyr. s usualwhencalculating

the capital ecoverycost,

he otal cost ofreplacing the bags s

subtracted

rom the total

depreciable

nvestment

o avoid

doubleaccounting.Because he recovered ust can be

returned o the

process

t no additionalcost. here s no

cost associated

ith final

disposi-

tion

of the solid

residue.

There s a substantial ecovery

credit br the wasteheatsteam,

which

helps o

partially

offset he annualcosts.The total

annual

cost br

the

system s

$504,800.

he

otalamount f

particulate

emoved

s

(0.0026)(

l0)(3,600)

8,640X

,000)

=

8,900

metric

ton/yr. Therefbre, he cost

efTectivenessf the system s

$56.7/metric

on of

dustcollected.

r

0.6coNcLUStoN

Fabric filtration

s rapidly becomingan

accepted nd frequently

preferredparticulate

mattercontrol echnology. nterest

n baghouses

ill

continue

ncreasing

s air crniss ions

tan-

dardsbecomemore stringent

and concernsover fine

particulates

ncrease. hey

are

gaining

acceptance

or

usedownstream f dry FGD s),stems

nd

'luidized

edboilers. he baghouse

ro-

vides additionalSO2 emoval.

and

particulate

ollection s not

affectedby the

high

alkali

content

of the

products

of dry SO2 emoval.The relative nsensitivity

of f'abric ilters to

variations

n dust

electric resistivity makes hem more

attractive han ESPs n many applications.ElectricuLly

enhancetl

fabric ilters,

an

emerging echnology, perate t higher

gas-to-cloth

atios han conven-

tional baghouses, nd at greatly educedpressure rops.

Sec.

Toble

DAC

Oper

Supe

- \ t a tn l

L"

Repl

U t i l i r

r '

f , - -

P r

Total

D

\

IAC

Or

e : :

Prope:

Capr i . .

> -

Total

\ [

Recor

: ,

T.{C



Sec.

10.5 Costs

f

Fabric

iltration

439

Toble

10.7

AnnuolCosts

or he

Fqbric

Filter f

Exomple

0' 2

DAC

Operating

abor:6

x

360

x

$12

=

Supervisoryabor:0.15 25,900=

Maintenance

L a b o r : 3 x 3 6 0 x $ 1 5 =

Materials

Replacement

arts,

ags:

$62,200

1.08

+

$2

x 5,100)

0.592

=

Utilities

Electr icity:

583

+ 17)

x

8,640 $0.06

=

Process

water

or boiler:

23.9

x

8,640

x

$0.26

=

Total

DAC

IA C

Overhead:

25,900

3,900

+ 2

x 16,200) 0.60

=

Property ax,

nsurance,

dministration:

.04

x

$2,403,000

Capital

ecoverycost

($2,403,000

$2

x

5,100

$62,200

1.08)

0'134

=

Total

IAC

Recovery

credit

or

wasteheatsteam:

23.9

x 8,640

x

$2

=

TAC

Sl-\ .900

-r.900

16.100

16.100

4 5 . 8 r 0

31

,000

53,70t)

$412,100

37,300

96,120

3

1,6,10

$44s,100

$413.000

$s04.800

r)

,'



442

Chap.

10

Fabric

ilters

PROBLEMS

The

problems

at

the end of each

chapter

have been

grouped into

four classes

designated

by a superscript

after the

problem

num-

ber)

Class

a: Illustrates

direct

numerical

application

of the

formulas

n the

text.

Class

b:

Requires lementary

nalysis

f

physical

situations,

ased

on

the subject

material

n the chapter.

Class :

Requires

omewhat

more

matureanalysis.

Classd:

Requires omputer

olution.

10.lu. Effect

of the

gas-to-clothatio

on

penetration

If the

gas-to-clothatioof

Example

0.1ncreaseso 0.015

m/s, alculate

he

aver-

age verall

enetrationor

a

20-min ycle.

Artsu;er:

.65Vo

10.2b. Effect of

operation cycle

ength

on

penetration

Determine

he operation

ycle

ength hat

reduces he

average

verall

penetration f

Example

0. to

| 07

Answer:

36 min

10.3c.

Penetration and

pressure

drop

through

a fabric

filter

The filter drag

parametersbr the

fly-ash and

abric combination

of

Example

10.1

are,S"

100kPa-s/m

nd

K2 106 -I. Choose

gas-to-clothatioandoperation

im e

such hat the

average verall

penetrationor

the

cycle does

not exceed

1.IVa

and he

pres-

suredroo does

not exceed

.5

kPa.

Prob l

10..1.

( e

(

0 . l

r

r

where

.r, I

\ { r

- \ l \ .

( b

Teras

U

compa

than-er

sonabl i

operati

was

me

1 9 8 9 ) .

Dpi

r

*m

F

0.05

0.08

0 .10

0.20

0.30

0.40

0.60

Est

whether

morphol



Problems

10.4c.Alternative

form

of the

morphology factor

(a)Show

hat an alternative xpression

or the morphology actor definedby'

Eq.

(10.2)

s

443

,=+ ,

o'l'= vtco

where

;r; s the mass raction collected

n

chanel

MGD

=

mass-weighted

eometric

meandiameter

(b)The

Monticello GeneratingStation

s

a coal-fired

power plant

operated y

the

TexasUtilities GeneratingCompany

Felix

et al. 1986).Part of the

flue

gases o

to a 36-

compartment

hakerbaghouseor

particle

emoval.The baghouse xperienced

igher-

than-expected

ressure

ropsand

arge

penetration

pikes,even hough

t

operated

t rea-

sonably low gas-to-clothatios. n an effort to find an explanationo theseunexpected

operational

roblems,

samples f the fly ashwere analyzed.

he specificsurface

area

was measured t 0.85

m2lg and he true

particle

density at2,420

kg/m3

Vann

Bush et al.

1989).The ollowing able

gives

a typicalsize

distribution nalysis.

Dpi

(pm)

MassFraction

DpiQtm)

Mass Fraction

0.02

0.05

0.08

0 .10

0.20

0.30

0.40

0.60

5x10 :

2x l } t

1 .6 10+

0.0010

0.0008

0 .0010

0.0070

0.0100

0.80

2.00

5.00

7.00

10 .0

20.0

50.0

70.0

0.0503

0.0600

0 .1500

0.3540

0.2640

0.1020

0.0010

8x10s

Estimate he value of the morphology actor

or

the

Monticello

fly-ash and discuss

whether he operational

roblems

experienced

an be explained

n termsof the

particle's

morphology.

Answer:

M

=

2.22



444

Chap.

0

Fabric

i l ters

10.5a.Morphology

of

particles

from

atmospheric fluidized-bed

combustor

(AFBC)

The

TVA atmosphericluidized-bed

ombustor s

a 20 MW

power plant,

which

burns

eastern ituminous

coal.

The

flue

gases

low

througha reverse-air

aghouse

or

particle emoval.The f1y-ash as he following properties: rue density,2,620 g/m3;spe-

cific surfacearea,11.44

ml/g; mass

geometric

meandiameter,

.0

pm.

Estimate he mor-

phology

actor or

this f1y-ash

see

Problem

10.4).Do

you

expect

penetration roblems

owing

o bleed-through?

Answer : L5

10.6b.Filter

drag model

parameters

estimation

Estimate he

parameters

n Eq.

(10.4)

based n

the ollowing estdata.

Gas-to-clothatio

=

0.013m/s

Inlet

dust

oading

=

0.005kg/ml

10.9b.

A

l q r - I

:

0

I

I

:

l

Probl

Sh

geometr

( b t l

mean.

( c t I

Time

min)

5 10

15

20 25

AP

(kPa)

0.33

0.49

0.55

0.60 0.64

Ansvyer:Kt

=

I .96 x 105 s-1

10.7a.Maximum

and

average ubesheet

pressure

drops

The fabric

filter

of

Problem

10.6

operateswith an nlet

dust oading of 0.01 kg/m3.

The filtration

time is I h. Estimate

he maximum

and average

ubesheet

ressure

rops

during the filtration

time.

10.g"

Tirbesheet

pressure

drop

in

a

pulse-jet

baghouse

Ansvver: P'"*

=

1'58

Pa

A

pulse-jet

baghouse

sesDacron elt

bags

or

the control

of

fly

ashemissions. he

gas-to-cloth

atio

s

0.030

m/s

and he nlet

dust oa ding s

0.02

kg/m3.

The bagsar e

cleaned

t lO-min ntervals

with

pulses

f air at 690 kPa

gauge.

f

K2

=

2.0

x

l0: s

r,

esti-

mate

he

maximum

ubesheet

ressure

rop.

Answer:2.61kPa

0.70

10.10b.

( a t l

298

K. e

the mean

effectof I

( b r

cooled

r

10.11b.

Th .

into

a

r

a: '



Problems

10.9b.

Areal density

distribution

in

a baghouse

(a)The

ollowine areal

densitv istribution ataweremeasuredn a bashouse:

Areal

density

{2,

g/m2)

Cumulat ive

ercent

ess han

l

445

0.56

0.11

l . l 0

1 .60

2.40

4 .10

I

-5

20

,5 0

80

98

Show hat thesedata ollow

a

log-normal

distributionand estimate he medianand

geometric

landard

eviat ion

(b)

Estimate he ratio

of the harmonicmean o the reciprocalof the arithmetic

mean.

(c)

Estimate hearithmeticmeanof the distribution

Answer:

L

l8

Answ'er: .71 kg/tn2

10.10b.Precooling of waste

gases

upstream of

fabric

filters

(a)

f the waste

gases

f Example10.8

are

precooled

y dilution

with

ambient

ir at

298 K,

estimatehe

volumetric

low rateof the

gases

t the baghouse

nlet.Assume hat

the

mean

heatcapacity f air n the ange298 o

530

K is 1.02 J/kg-K.Neglect he

effectof heat osses.

Attsv'er: 83 f/.s

(b)

Estimate he flow rate

of

the waste

gases

t the baghouse

nlet if they are

pre-

cooledby evaporation f waterat 298 K.

Art: t et ' ; 74.8

tt t l / . t

10.11b.Baghouse eat

exchanger

The

J.

E. BakerCompany f York,Pennsylvania

rocesses

olomitrc

imestone

into a varietyof agricultural nd ndustrial

roducts

Krout,

B.. and

Kilheffer.

. JAPCA.

r)



446

Chap.

0

Fabric

ilters

3l:293,1984).

They operate

wo

rorary

coal

and coke-fired

kilns.

In 1978,

hey fined

Numer

One

kiln

with

a fiberglass

everse-air

aghouse

or

particulate

matter

control.

The

system ncluded

a heat

exchanger

o lower

the

exhaust

as

emperature

rom

756 K to

533

K before

entering

he baghouse.

he cooling

luid

is ambient

air

which enters

he

heatexchanger t 300 K and eaves t 590K. Theunit hasa heat-transferreaof1,374

m2 or

an nlet

flow rate

of 56.7m3/s.

Estimate

he overall

heat-transfer

oefficient

or

the unit.

Assume

hat the

exhaust

gases

ehave ike

air.

Answer:23

W/m2-K

10.12".

Gross

cloth area for

a

reverse-air

fabric

filter

Estimate

he

gross

cloth

area or

the reverse-air

aghouse

f Problem

10.11.The

particles

coming

out

of the kiln

are

a mixture

of lime

dust

and ly-ash.

Answer: ,400

m2

10.13a.

Cloth area for

a

pulse-jet

baghouse

The

exhaust

ases

rom

a cement

plant

flow

at

the rate

of 130m3/s

at 500 K.

Calculate

he net

cloth

area or

a

pulse-jet

baghouse

o remove

he cement

dust.

Assume

on-line

cleaning

f

the bags.

Answer:

,190

m2

10.14u.

Reverse-air

baghouse

or

a coal-fired power

plant

Consider

he

coal-fired ower

plant

of Example

5.1.The plant

will

consist

f two,

150-MW

boilers.

Design

a

reverse-air

abric

filter

for

eachboiler

to remove

at least

99

percent

of the fly

ash n

the flue gases.

he

bagsare

0.3 m in

diameter

and 10.7m

long.

There

are 120

bags

per

compartment.

pecify

he

gross

cloth

area

and he number

of

compartments

or

eachbaghouse.

Answe 22

compartments/unit

Probl

10.15b.

Ac

tion

with

and mat

where

T/

T

(-

N,I

C

0.1 6

Flour

Tobacctr

Grain

Sari ust

E : t

open-h

1 9 8 0

.

l0 . l6h .

T l -

-

I t - -

dn' ing

L



448

Cha p. 0 Fabric il ters

fuel savings

are approximately

100,000/yr.

he annual

operationand maintenance

expenses

f the

heat

exchanger

re

$20,000/yr.

Estimate

he

profitability

ndex for

the

capital nvested

n the heatexchanger.

ssume hat

he useful ife

of the unit

is 15

yr

with

no salvage alue.The

following equation

an be

used o estimate he cost of heat

exchangersVatavuk,W. M. andNeveril,R. B. Chemical ngineering,p)2g,July 12,

1982):

rc= r4.84A-o' ' '

"*p[o.oozz(ni lz)

where

EC

=

cost n 1978

dollars

A

=

heat-transfer

rea,m2

Assume

hatTCI

=2.34

EC.

10.17^.

TCI for

a

fabric

filter system

Estimate

he total capital

nvestment or

the

fabric

filter

systemof Problem 10.14.

Because he fabric

filters will

be

ocated

upstream

f the FGD

system, he flue

gases

re

acidic

at this

point,

which mandates

tainless

teel abrication.

At the temperaure f

oper-

ation

(530

K)

insulation s required

and he

bagsmust be made

of

Fiberglass.

he cost

of

auxiliary

equipment or

eachbaghouses

$300,000.

Answer:TCI

=

$19.9

millions n

June

1990

10.18b.

TAC

of a

reverse-gas

aghouse

Estimate he total

annual

costsand he cost effectiveness

$/metric

on)

for

the bag-

housesystemof Problem10.17.The systemoperates ,640Wyr.The operatingabor rate

is

$25lh,

he maintenance

abor rate

s

$28lh.

The

cost of

electricity

s

$0.06/kW-h;

he

total

pressure

rop through he

system s

5

kPa. The

cost of final fly-ash

disposal

s

$20lmetric

on. The minimum

attractive ate

of return s

lT%alyr; he system ifetime is

20

yr.

The mechanical

fficiencyof the fans s

65

percent.

Answer:

$77/metric

on

10.194.

CI for

a MSW incinerator

pulse-jet

aghouse

Estimate

he otal

capitalnvestmentor

thebaghousef

Example

0.10.

ssume

Probl

that he

t ion s

n

10.20"

I n

control

o

conside

is capa

Proble

tubeshe

s-l and

a

20 r'r. e:

$

15,4r

n

process

$7.00/

t

( a t

( b t

( c

Ch

merit.

C



Problems

449

that he

cost of the

auxiliary

equipment s

$90,000.

No

buildingsor specialsite

prepara-

tion s

needed.

Answer:

$1,214,000

n

June1990

10.20c

Design

of a

pulse-jet

filter

for

a

Portland

cement kiln

In Problem

8.17,a multiple

cyclone system

was

designed s a

precleaner

or the

;ontrol

of

particulate

missions rom

a Portland

cementkiln.

For each

of

the alternatives

:onsidered

n

that

problem,

design

a

pulse-jet

abric

filter

such hat he combined

system

rr

capable

f satisfying

he

particulate

NSPS or

the source.

Use the correlation

of

Problem

10.15

o estimate

he

gas-to-cloth-ratio

or

eachalternative.Assume

hat he

:.-ibesheetressure

ropcan

be calculated

rom Eqs.

10.7)

o

(10.9),

with

K2

=

1.5

x

105

.-

and

a

cleaning

nterval

of 10

min. The lifetimes

of the bagsand system

are2

yr

and

it) r r. respectively, ith no salvage alue.Operatingand maintenanceabor ratesare

>1,s,tr

nd

$20lh,

respectively.

he dust

collected n

both devices

can be

recycled

o

the

::ocess

at no additional

cost.

Compressed

air

(dried

and iltered)

is availableat

)-.00/1,000

standardm3.

For each

aTtemative,

pecify he ollowing:

(a)

Total iltration

area

lb)

Filter

media,

numberand

dimensions

f the

bags

tc) TCI

and TAC

of the combined

multiple

cyclone-fabric ilter

system.

Choose

between he

threealternatives

uggested

asedon the EUAC measure

f

-

::rI.

Compare

our

results o

those f Problem

.22.

benitez cap10

Documents