cake filtration
TRANSCRIPT
![Page 1: Cake Filtration](https://reader034.vdocuments.mx/reader034/viewer/2022042508/577ccd161a28ab9e788b75dc/html5/thumbnails/1.jpg)
PIERO M. ARMENANTENJIT
Cake Filtration
![Page 2: Cake Filtration](https://reader034.vdocuments.mx/reader034/viewer/2022042508/577ccd161a28ab9e788b75dc/html5/thumbnails/2.jpg)
PIERO M. ARMENANTENJIT
Cake Filtration• Cake filtration consists of passing a solid
suspension (slurry) through a porous mediumor septum (e.g., a woven wire). The solids inthe slurry are retained on the surface of themedium where they build up, forming anincreasing thicker cake.
• As more slurry is filtered the solids retained onthe medium provide most of filtering action. Incake filtration the cake is the real filteringelement.
![Page 3: Cake Filtration](https://reader034.vdocuments.mx/reader034/viewer/2022042508/577ccd161a28ab9e788b75dc/html5/thumbnails/3.jpg)
PIERO M. ARMENANTENJIT
Cake Filtration (continued)• As time goes by the thickness of the cake
increases, as more solids are filtered. Thisresults in a corresponding increase of thepressure resistance across the cake.
• If the cake is incompressible (i.e., it does notchange its volume as pressure builds up) thepressure resistance increases proportionallyto the cake thickness.
• However, since most cakes are compressiblethe pressure across the cake typicallyincreases even faster than the cake build-up.
![Page 4: Cake Filtration](https://reader034.vdocuments.mx/reader034/viewer/2022042508/577ccd161a28ab9e788b75dc/html5/thumbnails/4.jpg)
PIERO M. ARMENANTENJIT
Cake Filtration (continued)• The cake is removed intermittently during
batch filtration processes. This is done bytaking the filter off line and manually orautomatically collecting the cake.
• The cake is removed continuously incontinuous processes, for example byscraping the cake with blades, as in rotatingfilters.
• Cake washing and drying operations can alsobe incorporated in the operation of mostfilters.
![Page 5: Cake Filtration](https://reader034.vdocuments.mx/reader034/viewer/2022042508/577ccd161a28ab9e788b75dc/html5/thumbnails/5.jpg)
PIERO M. ARMENANTENJIT
Examples of Cake-Forming Filters• Filter presses
• Belt filters
• Vacuum filters:
- Rotary vacuum belt filters
- Rotary vacuum precoat filters
- Vacuum disk filters
![Page 6: Cake Filtration](https://reader034.vdocuments.mx/reader034/viewer/2022042508/577ccd161a28ab9e788b75dc/html5/thumbnails/6.jpg)
PIERO M. ARMENANTENJIT
Example of a Filter Press
After Metcalf and Eddy, Wastewater Engineering, 1991, p. 869
![Page 7: Cake Filtration](https://reader034.vdocuments.mx/reader034/viewer/2022042508/577ccd161a28ab9e788b75dc/html5/thumbnails/7.jpg)
PIERO M. ARMENANTENJIT
Cross Section of a Filter Press
After Freeman, Standard Handbook of Hazardous Waste Treatment andDisposal, 1989, p. 7.9
![Page 8: Cake Filtration](https://reader034.vdocuments.mx/reader034/viewer/2022042508/577ccd161a28ab9e788b75dc/html5/thumbnails/8.jpg)
PIERO M. ARMENANTENJIT
Example of a Belt Filter
After Freeman, Standard Handbook of Hazardous Waste Treatment andDisposal, 1989, p. 7.10
![Page 9: Cake Filtration](https://reader034.vdocuments.mx/reader034/viewer/2022042508/577ccd161a28ab9e788b75dc/html5/thumbnails/9.jpg)
PIERO M. ARMENANTENJIT
Schematic of a Belt Press Filter
After Vesilind, Treatment and Disposal of Wastewater Sludges, 1979, p.156.
![Page 10: Cake Filtration](https://reader034.vdocuments.mx/reader034/viewer/2022042508/577ccd161a28ab9e788b75dc/html5/thumbnails/10.jpg)
PIERO M. ARMENANTENJIT
Example of a Rotary Vacuum Belt Filter
After Freeman, Standard Handbook of Hazardous Waste Treatment andDisposal, 1989, p. 7.8
![Page 11: Cake Filtration](https://reader034.vdocuments.mx/reader034/viewer/2022042508/577ccd161a28ab9e788b75dc/html5/thumbnails/11.jpg)
PIERO M. ARMENANTENJIT
Example of a Rotary Vacuum Precoat Filter
To Vacuum
Precoat
CakeDischarge
Slurry
Knife
Filtrate
![Page 12: Cake Filtration](https://reader034.vdocuments.mx/reader034/viewer/2022042508/577ccd161a28ab9e788b75dc/html5/thumbnails/12.jpg)
PIERO M. ARMENANTENJIT
Precoats and Filter Aids• A precoat is a layer of fine particulate material
(e.g., perlite) added on to the filter septumbefore filtration to form a coating cake
• During filtration the filtered solids in the slurrymay clog the filter and reduce the rate offiltration. This happens especially if theresulting cake is very compressible
• In such cases a filter aid made of fine particlesof a hard but porous material (such as perlite)having good filtering properties is added to theslurry to prevent and filtered with the slurrysolids
![Page 13: Cake Filtration](https://reader034.vdocuments.mx/reader034/viewer/2022042508/577ccd161a28ab9e788b75dc/html5/thumbnails/13.jpg)
PIERO M. ARMENANTENJIT
Common Types of Precoats and FilterAids
• Diatomaceous earth (diatomite)
A light siliceous material derived primarilyfrom sedimented diatoms (minute planktonicunicellular or colonial algae with silicifiedskeletons). Typical bulk density: 0.32 g/cm3
• Perlite
A volcanic glass made of siliceous rock havinga concentric shelly structure. Typical bulkdensity: 0.16 g/cm3
![Page 14: Cake Filtration](https://reader034.vdocuments.mx/reader034/viewer/2022042508/577ccd161a28ab9e788b75dc/html5/thumbnails/14.jpg)
PIERO M. ARMENANTENJIT
Analysis ofCake Filtration
![Page 15: Cake Filtration](https://reader034.vdocuments.mx/reader034/viewer/2022042508/577ccd161a28ab9e788b75dc/html5/thumbnails/15.jpg)
PIERO M. ARMENANTENJIT
Analysis of Suspended SolidsRemoval During Cake Filtration
• As the suspension moves through the filtermedium (septum) the suspended solids arestopped by the filter septum forming a filter cakeon top of the filter septum
• As more solids suspension passes through thefilter the cake builds up providing most of thefiltering action for the incoming suspension
• Equations can be written to describe the removalof the particles in suspension by the filter (i.e.,the formation of the cake) and the pressure dropof the fluid as it passes through the cake
![Page 16: Cake Filtration](https://reader034.vdocuments.mx/reader034/viewer/2022042508/577ccd161a28ab9e788b75dc/html5/thumbnails/16.jpg)
PIERO M. ARMENANTENJIT
Important Variables in Cake Filtration• Available pressure drop across cake, ∆P (Pa)
• Area of filtration, A (m2)
• Specific resistances of cake, α (m/kg)
• Specific resistances of medium (septum), Rm (1/m)
• Fluid superficial velocity, us (m/s)
• Size of cake particles, Dp (m)
• Shape factor for particles, φs
• Type of solids in suspension
• Cake void fraction, ε (void volume/total bed volume)
• Time, t (s)
![Page 17: Cake Filtration](https://reader034.vdocuments.mx/reader034/viewer/2022042508/577ccd161a28ab9e788b75dc/html5/thumbnails/17.jpg)
PIERO M. ARMENANTENJIT
Important Variables in Cake Filtration• Cake thickness, L (m)
• Concentration of solids in wastewater, Xw (g/L)
• Residual concentration of solids in filtrate, XF (g/L)
• Mass fraction of solids in cake, XC’ (g/g)
• Cumulative volume of wastewater fed to filter, VW (L)
• Cumulative volume of filtrate generated, VF (L)
• Cumulative mass of wet cake, mc (g)
• Mass of solids in the cake per volume of filtrate, XS (g/L)
• Density of wastewater, ρw (g/L)
• Density of filtrate, ρf (g/L)
![Page 18: Cake Filtration](https://reader034.vdocuments.mx/reader034/viewer/2022042508/577ccd161a28ab9e788b75dc/html5/thumbnails/18.jpg)
PIERO M. ARMENANTENJIT
Approach to Cake Filtration as aBatch Process
• Cake filtration is intrinsically a batch process.Hence, it can be expected that as filtrationproceeds the cake will build up and thepressure drop across the cake will increase.
• Mathematical modeling of batch cake filtrationis based on the determination of the rate offormation of the cake and the calculation ofpressure drop at any given time.
![Page 19: Cake Filtration](https://reader034.vdocuments.mx/reader034/viewer/2022042508/577ccd161a28ab9e788b75dc/html5/thumbnails/19.jpg)
PIERO M. ARMENANTENJIT
Approach to Cake Filtration as aBatch Process (continued)
• Integral quantities (such as the cumulativevolume of filtrate produced during a timeinterval, or the mass of the cake generatedduring the same interval) can be calculated byintegration of the basic instantaneous massbalances. In these equations the pressuredrop is typically a function of time.
• Continuous filtration is often modeled as asuccession of batch processes carried outover infinitesimally small time intervals.
![Page 20: Cake Filtration](https://reader034.vdocuments.mx/reader034/viewer/2022042508/577ccd161a28ab9e788b75dc/html5/thumbnails/20.jpg)
PIERO M. ARMENANTENJIT
Mass Balance Around a FilterFor a filter operating in a batch mode thefollowing diagram can be drawn:
(Wet) Cake
WastewaterFiltrate
VW XW VF XF
mc X'c
![Page 21: Cake Filtration](https://reader034.vdocuments.mx/reader034/viewer/2022042508/577ccd161a28ab9e788b75dc/html5/thumbnails/21.jpg)
PIERO M. ARMENANTENJIT
Definition of Xs
Xs is defined as the mass of (dry) solids in thecake per volume of filtrate generated.
From this definition it is that:
XX m
Vsc c
F
= =mass of solids in cakevolume of filtrate
'
![Page 22: Cake Filtration](https://reader034.vdocuments.mx/reader034/viewer/2022042508/577ccd161a28ab9e788b75dc/html5/thumbnails/22.jpg)
PIERO M. ARMENANTENJIT
Relationship Between SolidConcentrations Around a Filter
Mass balances around the filter give:
X V X m X Vw W c c F F= +' (solids)
ρ ρw W c F FV m V= + (overall)
![Page 23: Cake Filtration](https://reader034.vdocuments.mx/reader034/viewer/2022042508/577ccd161a28ab9e788b75dc/html5/thumbnails/23.jpg)
PIERO M. ARMENANTENJIT
Relationship Between SolidConcentrations Around a Filter (cont.’d)
After an explicit expression for VF has beenobtained it can be substituted in the equationdefining Xs to get:
X XX X
X Xs cW F F W
W W c
= ⋅ −−
''
ρ ρρ
If the densities of the wastewater and the filtrateare the same, then:
ρ ρ ρW F= =
X XX X
X Xs cF W
W c
= ⋅ −−
ρρ
''
![Page 24: Cake Filtration](https://reader034.vdocuments.mx/reader034/viewer/2022042508/577ccd161a28ab9e788b75dc/html5/thumbnails/24.jpg)
PIERO M. ARMENANTENJIT
Relationship Between SolidConcentrations Around a Filter (cont.’d)
Special case: A common situation is that inwhich all the solids contained in the suspensionare removed by the filter and contribute to theformation of the cake. In other words, the filtratedoes not contain any solids. In such a case it is:
XF ≡ 0
and the expression for Xs becomes:
X XX
X Xs cW
c W
= ⋅−
ρρ
''
![Page 25: Cake Filtration](https://reader034.vdocuments.mx/reader034/viewer/2022042508/577ccd161a28ab9e788b75dc/html5/thumbnails/25.jpg)
PIERO M. ARMENANTENJIT
Relationship Between SolidConcentrations Around a Filter (cont.’d)
Note that, in general, Xs is different from XW
Only if:
ρ X Xc W'>>
it would then be that:
X XX
X XX
XX
Xs cW
c Wc
W
cW= ⋅
−≅ ⋅ ≅ρ
ρρ
ρ'
''
'
![Page 26: Cake Filtration](https://reader034.vdocuments.mx/reader034/viewer/2022042508/577ccd161a28ab9e788b75dc/html5/thumbnails/26.jpg)
PIERO M. ARMENANTENJIT
Cumulative Mass Balance for theSolids in the Cake
At a generic time t a cumulative mass balance forthe solids in the cake (i.e., the solids that havecontributed to the formation of the cake in thetime interval 0-t) gives:
solids accumulated in the cake during time t
solids removed from suspension during time t
=
![Page 27: Cake Filtration](https://reader034.vdocuments.mx/reader034/viewer/2022042508/577ccd161a28ab9e788b75dc/html5/thumbnails/27.jpg)
PIERO M. ARMENANTENJIT
Cumulative Mass Balance for theSolids in the Cake
The previous equation can be rewrittensymbolically, for a generic time t, as:
( ) ( )L A X L A X Vs W s F1− + =ε ρ ε
The first term represents the mass of solids in thesolid component of the cake at time t; the secondterm is the amount of solids still in suspension inthe water contained in the cake; and the thirdterm is the amount of solids removed from thefiltrate (and now held in the cake).
Remark: L, ε, and VF can all be functions of time.
![Page 28: Cake Filtration](https://reader034.vdocuments.mx/reader034/viewer/2022042508/577ccd161a28ab9e788b75dc/html5/thumbnails/28.jpg)
PIERO M. ARMENANTENJIT
Cumulative Mass Balance for theSolids in the Cake (continued)
The volume of water contained in the cake (εLA) istypically much smaller than the volume of filtrate,VF, produced during the time interval 0-t.Furthermore, Xs and Xw are of the same order ofmagnitude. Then, one can safely assume that:
( )X V X L As F W>> ε
Hence, the cumulative mass balance for thesolids in the cake becomes:
( )L A X Vs s F1− ≅ε ρ
![Page 29: Cake Filtration](https://reader034.vdocuments.mx/reader034/viewer/2022042508/577ccd161a28ab9e788b75dc/html5/thumbnails/29.jpg)
PIERO M. ARMENANTENJIT
Cake Thickness, L, as a Function ofVolume of Liquid Passed Through
the FilterThe previously derived mass balance equation forthe solids in the cake:
( )L A X Vs s F1− =ε ρ
can be rewritten as:
( )LX V
As F
s
=−1 ε ρ
where L, ε, and VF can all be functions of time.
![Page 30: Cake Filtration](https://reader034.vdocuments.mx/reader034/viewer/2022042508/577ccd161a28ab9e788b75dc/html5/thumbnails/30.jpg)
PIERO M. ARMENANTENJIT
Shape Factor of Particles in CakeThe particle shape factor, φp, is defined as:
φp = Surface area of sphere having same volume as particleSurface area of particle
i.e.,
φπ
πpp
p
p
sph p
D
DVA
VD A
sph
sph
= ⋅ =6 62
3
where Dsph is the diameter of a sphere having thesame volume as the particle.
![Page 31: Cake Filtration](https://reader034.vdocuments.mx/reader034/viewer/2022042508/577ccd161a28ab9e788b75dc/html5/thumbnails/31.jpg)
PIERO M. ARMENANTENJIT
Relationship Between Dp, Dsph, and φp
Since:
DVAp
p
p
=6
and:
φpsph
p
pDVA
= ⋅1 6
then:
D Dp p sph= φ
![Page 32: Cake Filtration](https://reader034.vdocuments.mx/reader034/viewer/2022042508/577ccd161a28ab9e788b75dc/html5/thumbnails/32.jpg)
PIERO M. ARMENANTENJIT
Approximate Relationship BetweenDp and Sieve Opening
The assumption is often made that:
D Dsph p≈
where Dp is the average size of the particleswhose size is between two sieve openings
D D Dp s s= 1 2
and where Ds1 and Ds2 are the sieve openings.Then:
D D Dp p sph p p= ≅φ φ
![Page 33: Cake Filtration](https://reader034.vdocuments.mx/reader034/viewer/2022042508/577ccd161a28ab9e788b75dc/html5/thumbnails/33.jpg)
PIERO M. ARMENANTENJIT
Pressure Drop During Cake Filtration
Filter Cake L(t)
Filter Medium
Filtrate
Suspension
∆Pc(t)
∆Pm
![Page 34: Cake Filtration](https://reader034.vdocuments.mx/reader034/viewer/2022042508/577ccd161a28ab9e788b75dc/html5/thumbnails/34.jpg)
PIERO M. ARMENANTENJIT
Pressure Drop During Cake FiltrationAt any time, t, the pressure drop experienced atthat time by a suspension passing through a filtercake supported by a filter medium (or septum) is:
( ) ( )∆ ∆ ∆P t P t Pc m= +where:∆P(t) = total pressure drop across filter
∆Pc(t) = pressure drop due to filter cake
∆Pm = pressure drop due to filter medium
Remark: during batch filtration the cake can beexpected to build up, and the pressure drop toincrease as time passes.
![Page 35: Cake Filtration](https://reader034.vdocuments.mx/reader034/viewer/2022042508/577ccd161a28ab9e788b75dc/html5/thumbnails/35.jpg)
PIERO M. ARMENANTENJIT
Pressure Drop Across Filter CakeSince the liquid passing through the filter cakemoves in laminar flow (because of the small poresof the cake and the slow fluid velocity) the Blake-Kozeny equation can be used (instead of the moregeneral Ergun equation) to describe thedependence of the pressure drop through thecake with the superficial velocity, us:
( )∆P
LD
ucp p
L s= −
150 1 2
32
Reε
ε ρ Blake-Kozeny equation
where ∆Pc is the pressure drop through the cake,and is, in general, a function of time.
![Page 36: Cake Filtration](https://reader034.vdocuments.mx/reader034/viewer/2022042508/577ccd161a28ab9e788b75dc/html5/thumbnails/36.jpg)
PIERO M. ARMENANTENJIT
Pressure Drop Across Filter CakesSubstituting the expression for Re in the Blake-Kozeny equation gives:
( )∆P
LD
ucp
s= −
1501 2
3 2µ εε
As before, ∆Pc is typically a function of time, sincethe cake thickness, L, the superficial velocity, us,and the void fraction, ε, can all change with time.
![Page 37: Cake Filtration](https://reader034.vdocuments.mx/reader034/viewer/2022042508/577ccd161a28ab9e788b75dc/html5/thumbnails/37.jpg)
PIERO M. ARMENANTENJIT
Pressure Drop Across Filter Cakes(Carman-Kozeny Equation)
For filter cakes the constant 150 may not beappropriate since the cake particles arecompressible.
Therefore the Blake-Kozeny equation is oftenrewritten to produce to so-called Carman-Kozenyequation:
( )∆P k
LD
ucp
s= −
1
2
3 2
1µ εε Carman-Kozeny equation
where: k1 = proportionality constant.
![Page 38: Cake Filtration](https://reader034.vdocuments.mx/reader034/viewer/2022042508/577ccd161a28ab9e788b75dc/html5/thumbnails/38.jpg)
PIERO M. ARMENANTENJIT
Superficial Velocity in Cake FiltrationAs before, the superficial (or approach) velocity isdefined as the velocity of the liquid as it flowsthrough a cross section equal to that of the tank(or filter vessel) in the absence of the cake. It isalso equal to the filtrate flow rate, QF, divided bythe total cross-sectional area normal to flow, i.e.:
uQA
dVd t As
F F= = 1
where:A = cross sectional area or empty filter vessel
VF = volume of filtrate passed through the cakeduring time t
![Page 39: Cake Filtration](https://reader034.vdocuments.mx/reader034/viewer/2022042508/577ccd161a28ab9e788b75dc/html5/thumbnails/39.jpg)
PIERO M. ARMENANTENJIT
Equation for Pressure Dropin the Cake
Combining together the equations:
( )∆P k
LD
ucp
s= −
1
2
3 2
1µ εε u
QA
dVd t As
F F= = 1
and: ( )LX V
As F
s
=−1 ε ρ
the following expression for ∆Pc is found:
( )( )∆P k
DX VA
dVd tc
s p
s F F=−
−
1
2
3 2 2
1 11ρ
εε ε
µ
![Page 40: Cake Filtration](https://reader034.vdocuments.mx/reader034/viewer/2022042508/577ccd161a28ab9e788b75dc/html5/thumbnails/40.jpg)
PIERO M. ARMENANTENJIT
Equation for Pressure Dropin the Cake
The previous equation can be re-arranged to givethe final equation for the pressure drop in thecake:
∆PX VA
dVd tc
s F F= α µ 2
where α = specific cake resistance to filtration, isgiven by:
αρ
εε
= −kDs p
12 3
1
∆PC, L, α, and VF can all be functions of time.
![Page 41: Cake Filtration](https://reader034.vdocuments.mx/reader034/viewer/2022042508/577ccd161a28ab9e788b75dc/html5/thumbnails/41.jpg)
PIERO M. ARMENANTENJIT
Equation for Pressure Dropin Filter Medium (Septum)
The pressure drop across the filter medium(septum) can also be expressed using theCarman-Kozeny equation that can be rewritten as:
( )∆P kLD
umm
m
m
pms=
−
2
2
3 2
1µ
εε
where the subscript “m” refers to the medium andthe superficial velocity is given by:
uQA
dVd t As
F F= = 1
![Page 42: Cake Filtration](https://reader034.vdocuments.mx/reader034/viewer/2022042508/577ccd161a28ab9e788b75dc/html5/thumbnails/42.jpg)
PIERO M. ARMENANTENJIT
Equation for Pressure Drop in FilterMedium (continued)
The resulting expression of the pressure drop inthe medium is:
∆P RA
dVd tm m
F= µ 1
where:
( )R k
LDm
m
m
m
pm
=−
2
2
3 2
1 εε
with Rm = specific resistance of medium tofiltration
![Page 43: Cake Filtration](https://reader034.vdocuments.mx/reader034/viewer/2022042508/577ccd161a28ab9e788b75dc/html5/thumbnails/43.jpg)
PIERO M. ARMENANTENJIT
Equation for Total Pressure DropDuring Cake Filtration
Recalling that the total pressure drop in a filter is:( ) ( )∆ ∆ ∆P t P t Pc m= +
it is:
( ) ( ) ( )∆P t t
X V tA
RA
dVd t
s Fm
F= +
α µ µ2
1
Since by definition it is: QF(t) = dVF/dt, then:
( ) ( )( ) ( )[ ]
dVd t
Q tA P t
t X V t ARF
Fs F m
= =+
2 ∆µ α
This is the main design equation for cake filters.
![Page 44: Cake Filtration](https://reader034.vdocuments.mx/reader034/viewer/2022042508/577ccd161a28ab9e788b75dc/html5/thumbnails/44.jpg)
PIERO M. ARMENANTENJIT
Specific Cake Resistance andCake Compressibility
From the expression for α:
αρ
εε
= −kDs p
12 3
1
one can incorrectly assume that the pressureacross the cake has no impact on specific cakeresistance. In fact, the void fraction ε for mostcakes can be significantly affected by pressure,since the cake is often compressible. Since thepressure drop changes with time the void fractionε can also be a function of time, at least inprinciple.
![Page 45: Cake Filtration](https://reader034.vdocuments.mx/reader034/viewer/2022042508/577ccd161a28ab9e788b75dc/html5/thumbnails/45.jpg)
PIERO M. ARMENANTENJIT
Specific Cake Resistance andCake Compressibility
In practice, it is convenient to carry outexperiments to determine:
• the specific cake resistance under no pressuredifference (no compression). Cake is built upby gravity filtering;
• the effect of pressure difference across thecake on the specific cake resistance. Cake isbuilt up first and then compressed to a knowpressure with a piston provided with a porousbottom. Filtrate is passed trough the cake.
![Page 46: Cake Filtration](https://reader034.vdocuments.mx/reader034/viewer/2022042508/577ccd161a28ab9e788b75dc/html5/thumbnails/46.jpg)
PIERO M. ARMENANTENJIT
Specific Cake Resistance andCake Compressibility
Piston
Suspension
Filtrate
Cake
CakeBuildup
CompressedCake
Suspension
Filtrate
Filtrate
PP
![Page 47: Cake Filtration](https://reader034.vdocuments.mx/reader034/viewer/2022042508/577ccd161a28ab9e788b75dc/html5/thumbnails/47.jpg)
PIERO M. ARMENANTENJIT
Specific Cake Resistance andCake Compressibility (continued)Possible results of cake compression experiment:
• cake is incompressible. Cake resistance, α isindependent of ∆P;
• cake is compressible. Cake resistance isexpressed as:
( )α α= osP∆
with: αo = empirical constants = coefficient of compressibility (typicalrange for most domestic sludges: 0.4-0.9; lime sludges: 1.05; sand: 0).
![Page 48: Cake Filtration](https://reader034.vdocuments.mx/reader034/viewer/2022042508/577ccd161a28ab9e788b75dc/html5/thumbnails/48.jpg)
PIERO M. ARMENANTENJIT
Specific Cake ResistanceTypical values of the specific cake resistance, α,are in the following ranges:
• 1013-1015 m/kg for raw sludges;
• 1011-1012 m/kg for well conditioned sludges.
![Page 49: Cake Filtration](https://reader034.vdocuments.mx/reader034/viewer/2022042508/577ccd161a28ab9e788b75dc/html5/thumbnails/49.jpg)
PIERO M. ARMENANTENJIT
Typical Specific Cake Resistancewith Chemical Conditioning
0 0.5 1 1.5 2 2.5
Chemical Conditioner (%)
1E+010
1E+011
1E+012
1E+013
1E+014
Spe
cific
Cak
e R
esis
tanc
e, α
(m/k
g)
After Vesilind, Treatment and Disposal of Wastewater Sludges, 1979, p.156.
![Page 50: Cake Filtration](https://reader034.vdocuments.mx/reader034/viewer/2022042508/577ccd161a28ab9e788b75dc/html5/thumbnails/50.jpg)
PIERO M. ARMENANTENJIT
Compressibility of Sludges as Measuredby the Specific Resistance Test
1 10 100
Vacuum Pressure (kPa)
1E+010
1E+011
1E+012
Spe
cific
Cak
e R
esis
tanc
e, α
(m/k
g)
After Vesilind, Treatment and Disposal of Wastewater Sludges, 1979, p.156.
![Page 51: Cake Filtration](https://reader034.vdocuments.mx/reader034/viewer/2022042508/577ccd161a28ab9e788b75dc/html5/thumbnails/51.jpg)
PIERO M. ARMENANTENJIT
Batch Filtration OperationsBatch cake filtration is typically carried out underone of the following conditions:
• Constant filtrate flow rate. Since the pressuredrop across the filter increases as a result ofcake buildup this condition implies that theupstream pressure must be increased withtime.
• Constant pressure drop across the filter. Thiscondition implies that the filtrate flow ratedeclines as the cake builds up.
• Variable flow rate and variable pressure drop.
![Page 52: Cake Filtration](https://reader034.vdocuments.mx/reader034/viewer/2022042508/577ccd161a28ab9e788b75dc/html5/thumbnails/52.jpg)
PIERO M. ARMENANTENJIT
Batch Cake Filtration atConstant Filtrate Flow Rate
In some cases cake filtration is carried out usinga positive displacement pump. This results in aconstant flow rate process. Then:
QF = constantRecalling the design equation for cake filters it is:
( )( ) ( )[ ]
dVd t
QA P t
t X V t ARF
Fs F m
= =+
2 ∆µ α
Important: although QF = dVF/dt is a constant, VF(the total filtrate at time t) is not. In fact, it is:
dV Qdt V Q tF F F= ⇒ =
![Page 53: Cake Filtration](https://reader034.vdocuments.mx/reader034/viewer/2022042508/577ccd161a28ab9e788b75dc/html5/thumbnails/53.jpg)
PIERO M. ARMENANTENJIT
Batch Cake Filtration atConstant Filtrate Flow Rate (cont' d)
Assuming that the cake is not compressible (i.e.,α is independent of ∆P) the pressure buildupwhile operating at constant filtrate flow rate (i.e.,constant QF) is given by:
( ) ( )[ ]∆P tX V t AR
AQs F m
F=+µ α
2
i.e., recalling that VF = QF·t:
( )∆P tX QA
tR Q
As F m F= +µα µ2
2 ( )for constantQF =
![Page 54: Cake Filtration](https://reader034.vdocuments.mx/reader034/viewer/2022042508/577ccd161a28ab9e788b75dc/html5/thumbnails/54.jpg)
PIERO M. ARMENANTENJIT
Applications of Batch Cake Filtrationat Constant Filtrate Flow Rate
• Batch cake filtration at constant filtrate flowrate is used primarily in sludge dewatering;
• The type of filters that utilizes filtration methodis the filter press;
• Positive displacement pumps are used to forcethe suspension through the filter;
• Gauge pressures up to 225 psi (15 atm) areused.
![Page 55: Cake Filtration](https://reader034.vdocuments.mx/reader034/viewer/2022042508/577ccd161a28ab9e788b75dc/html5/thumbnails/55.jpg)
PIERO M. ARMENANTENJIT
Batch Cake Filtration at Constant ∆PIf the pressure across the filter is constant thegeneral filtration equation:
( ) ( )( ) ( )[ ]
dVd t
Q tA P t
t X V t ARF
Fs F m
= =+
2 ∆µ α
becomes:
( ) ( )[ ]dVd t
Q tA P
X V t ARF
Fs F m
= =+
2 ∆µ α
Note that the coefficient α is constant (but notnecessarily equal to αo) even if the cake iscompressible, since ∆P = constant.
![Page 56: Cake Filtration](https://reader034.vdocuments.mx/reader034/viewer/2022042508/577ccd161a28ab9e788b75dc/html5/thumbnails/56.jpg)
PIERO M. ARMENANTENJIT
Batch Cake Filtration atConstant ∆P (continued)
Since ∆P = constant, the previous equation can beintegrated by separating variables:
( )µ α X V AR
A PdV d ts F m
F
Vt
F '' '
+⌠
⌡ = ∫2
00∆
Integration of this equation yields:
( ) ( )µα µXA P
V tR
A PV t ts
Fm
F2 22
∆ ∆+ = ( )for constant∆P =
![Page 57: Cake Filtration](https://reader034.vdocuments.mx/reader034/viewer/2022042508/577ccd161a28ab9e788b75dc/html5/thumbnails/57.jpg)
PIERO M. ARMENANTENJIT
Batch Cake Filtration atConstant ∆P (continued)
The previous equation can be rewritten as:
( ) ( )η γV t V t tF F2 + =
where ∆P is constant, and the parameters η and γare given by the equations:
η µα= XA P
s
2 2 ∆ and γ µ= RA P
m
∆
![Page 58: Cake Filtration](https://reader034.vdocuments.mx/reader034/viewer/2022042508/577ccd161a28ab9e788b75dc/html5/thumbnails/58.jpg)
PIERO M. ARMENANTENJIT
Batch Cake Filtration at Constant ∆P:Determination of Filtration Parameters
The determination of η and γ from batchexperiments conducted at constant ∆P can bemade by rearranging the equation:
( ) ( )µα µXA P
V tR
A PV t ts
Fm
F2 22
∆ ∆+ =
as: ( ) ( )tV t
XA P
V tR
A PF
sF
m= +µα µ2 2 ∆ ∆
with: slope and intercept= = = =η µα γ µXA P
RA P
s m
2 2 ∆ ∆and t/VF = y-coordinate and VF = x-coordinate
![Page 59: Cake Filtration](https://reader034.vdocuments.mx/reader034/viewer/2022042508/577ccd161a28ab9e788b75dc/html5/thumbnails/59.jpg)
PIERO M. ARMENANTENJIT
Determination of the Specific Cake ResistanceThrough Batch Filtration Experiments atConstant ∆P: Buchner Funnel Apparatus
Buchner Filter
Vacuum Gauge
Graduated Cylinder
To Vacuum
![Page 60: Cake Filtration](https://reader034.vdocuments.mx/reader034/viewer/2022042508/577ccd161a28ab9e788b75dc/html5/thumbnails/60.jpg)
PIERO M. ARMENANTENJIT
Determination of the Specific Cake ResistanceThrough Batch Filtration Experiments at
Constant ∆P: Filter Leaf Apparatus
VacuumGauge
GraduatedCylinder
To Vacuum
FilterLeaf
![Page 61: Cake Filtration](https://reader034.vdocuments.mx/reader034/viewer/2022042508/577ccd161a28ab9e788b75dc/html5/thumbnails/61.jpg)
PIERO M. ARMENANTENJIT
Example of ExperimentalDetermination of Filtration Constants
0 0.001 0.002 0.003 0.004 0.005 0.006
Cumulative Filtrate Volume, VF (m3)
0
5,000
10,000
15,000
20,000
25,000
t/VF
(s/m
3 )
ExperimentalRegression
Intercept = 6840.95 s/m3
Slope = 2.85378E+006 s/m6
![Page 62: Cake Filtration](https://reader034.vdocuments.mx/reader034/viewer/2022042508/577ccd161a28ab9e788b75dc/html5/thumbnails/62.jpg)
PIERO M. ARMENANTENJIT
Batch Cake Filtration at Constant ∆P:Approximate Equations
If the resistance of the filter medium, Rm, is verysmall in comparison to the cake resistance, α, thebatch filtration equation can be rewritten as:
( )tX
A PV ts
s≅ µα2 2
2
∆ ⇒ ( )V t
A PX
tt
FF
≅ =2 2 ∆µα η
i.e.:
( )Q tdVd t
A PX t tF
F
F
= ≅ =12
2 12
12 ∆µα η
Note that QF(t)→ 0 for t→ ∞ .
![Page 63: Cake Filtration](https://reader034.vdocuments.mx/reader034/viewer/2022042508/577ccd161a28ab9e788b75dc/html5/thumbnails/63.jpg)
PIERO M. ARMENANTENJIT
Batch Cake Filtration at Constant ∆P:Plot of Approximate Expression
for VF(t)
Time, t
Cum
ulat
ive
Filtr
ate
Vol
ume,
VF
![Page 64: Cake Filtration](https://reader034.vdocuments.mx/reader034/viewer/2022042508/577ccd161a28ab9e788b75dc/html5/thumbnails/64.jpg)
PIERO M. ARMENANTENJIT
Applications of Batch Cake Filtrationat Constant ∆P
• Batch cake filtration at constant ∆P is usedprimarily in sludge dewatering;
• The types of filters using this filtration methodinclude:
- Filter presses
- Belt filter presses
![Page 65: Cake Filtration](https://reader034.vdocuments.mx/reader034/viewer/2022042508/577ccd161a28ab9e788b75dc/html5/thumbnails/65.jpg)
PIERO M. ARMENANTENJIT
Continuous Cake Filtration atConstant ∆P: Rotary Vacuum Filters
![Page 66: Cake Filtration](https://reader034.vdocuments.mx/reader034/viewer/2022042508/577ccd161a28ab9e788b75dc/html5/thumbnails/66.jpg)
PIERO M. ARMENANTENJIT
Continuous Cake Filtration atConstant ∆P: Rotary Vacuum Filter
CakeDewatering
CakeDischarge
Suspension
Solids
Cake
Blade
CakeFormation
θ
![Page 67: Cake Filtration](https://reader034.vdocuments.mx/reader034/viewer/2022042508/577ccd161a28ab9e788b75dc/html5/thumbnails/67.jpg)
PIERO M. ARMENANTENJIT
Continuous Cake Filtration atConstant ∆P: Rotary Vacuum Filters
• In continuous filtration operations, such asthose involving vacuum rotary filters, eachfilter element undergoes a batch cakefiltration, followed by a cake dewatering phase,and a cake discharge phase with each rotation.
• The filter cake is formed under a constant ∆Pdriving force generated by a vacuum.
• The filter cake is formed only during the timeperiod when the filter surface is immersed inthe suspension.
![Page 68: Cake Filtration](https://reader034.vdocuments.mx/reader034/viewer/2022042508/577ccd161a28ab9e788b75dc/html5/thumbnails/68.jpg)
PIERO M. ARMENANTENJIT
Continuous Cake Filtration atConstant ∆P: Cycle Time
The immersion period in each cycle (i.e., for eachfull rotation of the filter drum) is given by:
t f t fk c k= = = =22
2πω
θπ
πω
θω
where: tc = cycle time (time for one full rotation)fk = fraction of cycle time available for cake
formation = fraction submergence of drumsurface
θ = angle comprising the sector immersed insuspension (rad)
ω = rotational (angular) velocity (rad/s)
![Page 69: Cake Filtration](https://reader034.vdocuments.mx/reader034/viewer/2022042508/577ccd161a28ab9e788b75dc/html5/thumbnails/69.jpg)
PIERO M. ARMENANTENJIT
Continuous Cake Filtration atConstant ∆P: Filtration EquationDuring the time period t = fk tc the filter cake in arotary filter is formed just as in a batch operation.The (batch) filtration equation for part of thecontinuous process is:
( ) ( )µα µXA P
V tR
A PV t t f ts
Fm
F k c2 22
∆ ∆+ = =
since ∆P is constant.
![Page 70: Cake Filtration](https://reader034.vdocuments.mx/reader034/viewer/2022042508/577ccd161a28ab9e788b75dc/html5/thumbnails/70.jpg)
PIERO M. ARMENANTENJIT
Continuous Cake Filtration atConstant ∆P: Filtrate Generated
During a CycleThe previous quadratic expression is an equationin VF that can be solved for VF and rearranged togive:
( )V tAX
R Rf t X P
Fs
m mk c s= − + +
α
αµ
2 2 ∆
![Page 71: Cake Filtration](https://reader034.vdocuments.mx/reader034/viewer/2022042508/577ccd161a28ab9e788b75dc/html5/thumbnails/71.jpg)
PIERO M. ARMENANTENJIT
Continuous Cake Filtration atConstant ∆P: Filtrate Flux
The previous equation can be rearranged to give:
Filtrate flux = =
= = − + +
QA
VAt X
Rt
Rt
f X Pt
F
F
c s
m
c
m
c
k s
c
1 22
2ααµ
∆
which predicts the filtrate flux, i.e., the amount offiltrate VF produced per unit filter area during acycle (or full rotation) lasting a time interval tc.QF = average filtrate flow rate during the interval tc
![Page 72: Cake Filtration](https://reader034.vdocuments.mx/reader034/viewer/2022042508/577ccd161a28ab9e788b75dc/html5/thumbnails/72.jpg)
PIERO M. ARMENANTENJIT
Continuous Cake Filtration atConstant ∆P: Filter Loading
This equation can be rewritten to give the amountof filter solids loading, Gk, produced during thesame cycle:
G V XAt
Rt
Rt
f X Ptk
F s
c
m
c
m
c
k s
c
= = − + +
1 22
2ααµ
∆
where:
Gk = filter solids loading (kg/m2 s) = amount ofsolids filtered per unit filter area over a timeinterval tc.
![Page 73: Cake Filtration](https://reader034.vdocuments.mx/reader034/viewer/2022042508/577ccd161a28ab9e788b75dc/html5/thumbnails/73.jpg)
PIERO M. ARMENANTENJIT
Continuous Cake Filtration atConstant ∆P: Continuous Operation
Since a continuous process is nothing more thana sequence of cycles, each one lasting tc, then:
Filtrate flux = = − + +
QA X
Rt
Rt
f X Pt
F
s
m
c
m
c
k s
c
1 22
2ααµ
∆
The equation:
G V XAt
Rt
Rt
f X Ptk
F s
c
m
c
m
c
k s
c
= = − + +
1 22
2ααµ
∆
can be used to describe the continuous operationof a rotary (vacuum) filter.
![Page 74: Cake Filtration](https://reader034.vdocuments.mx/reader034/viewer/2022042508/577ccd161a28ab9e788b75dc/html5/thumbnails/74.jpg)
PIERO M. ARMENANTENJIT
Continuous Cake Filtration atConstant ∆P: Rotational VelocityIf the angular velocity is expressed in rpm, i.e.:
N = 602ωπ
then the cycle time and the rotational (angular)velocities (in rad/s or rpm) are related by:
tNc = =2 60π
ωwhere: ω = rotational (angular) velocity in rad/stc = cycle time (to complete a rotation) in secondsN = rotational (angular) velocity in rpm (rotations
per minute).
![Page 75: Cake Filtration](https://reader034.vdocuments.mx/reader034/viewer/2022042508/577ccd161a28ab9e788b75dc/html5/thumbnails/75.jpg)
PIERO M. ARMENANTENJIT
Continuous Cake Filtration atConstant ∆P: Continuous Filtrate Flux
The continuous filtrate flux can be convenientlyexpressed in terms of the agitation velocity:
( )Filtrate flux = = − + +
QA X
R R f X PF
s
m m k s12 2
2 2
2αω
πω
πω α
µπ∆
( )Filtrate flux = = − + +
QA X
R N R N f N X PF
s
m m k s160 60
260
2 2
2αα
µ∆
where: tc = cycle time (to complete a rotation), in sN = rotational (angular) velocity in rpm (rotations
per minute).
![Page 76: Cake Filtration](https://reader034.vdocuments.mx/reader034/viewer/2022042508/577ccd161a28ab9e788b75dc/html5/thumbnails/76.jpg)
PIERO M. ARMENANTENJIT
Continuous Cake Filtration atConstant ∆P: Simplified EquationsIf the resistance of the filter medium, Rm, is verysmall then the equations for continuous filtrationcan be simplified and rewritten as:
Filtrate flux = = ≅QA
VAt
f PX t
F F
c
k
s c
2 ∆µα
G V XAt
f X Ptk
F s
c
k s
c
= ≅ 2 ∆µα
These equations are especially useful tounderstand the relationships between the variousvariables affecting a continuous filtration process.
![Page 77: Cake Filtration](https://reader034.vdocuments.mx/reader034/viewer/2022042508/577ccd161a28ab9e788b75dc/html5/thumbnails/77.jpg)
PIERO M. ARMENANTENJIT
Continuous Cake Filtration atConstant ∆P: Simplified Equations
If the resistance of the filter medium, Rm, is verysmall the equations for continuous filtration can beexpressed as a function of the rotational velocity as:
Filtrate flux = = ≅ ≅QA
VAt
f PX
F F
c
k
s
∆ ωπµα
≅ 260
f P NX
k
s
∆µα
where: tc = cycle time (to complete a rotation) in sN = rotational (angular) velocity in rpm (rotations
per minute).
![Page 78: Cake Filtration](https://reader034.vdocuments.mx/reader034/viewer/2022042508/577ccd161a28ab9e788b75dc/html5/thumbnails/78.jpg)
PIERO M. ARMENANTENJIT
Applications of Continuous CakeFiltration at Constant ∆P
• Continuous cake filtration at constant ∆P is themost widely used method of sludgedewatering;
• The types of filters using this filtration methodinclude:
- Rotary vacuum belt filter
- Rotary vacuum precoat filters
- Rotary vacuum drum filters
- Rotary vacuum disc filters
![Page 79: Cake Filtration](https://reader034.vdocuments.mx/reader034/viewer/2022042508/577ccd161a28ab9e788b75dc/html5/thumbnails/79.jpg)
PIERO M. ARMENANTENJIT
Comparison of Different Types of FiltersRotary
Drum BeltFilters
Belt FilterPresses
FilterPresses
GranularDeep-Bed
Filters
Size 1-70 m2 1-2 m beltwidth
0.02-16 m3 0.2-10 m2
Solids inFeed (%)
2-5 2-8 -- --
Solids inCake (%)
15-20 15-25 28-40
SolidsLoading
10 kg/m2 h 190-270kg/m h
-- 0.12-0.5m3/m2 min
After Freeman, Standard Handbook of Hazardous Waste Treatment andDisposal, 1989, p. 7.12
![Page 80: Cake Filtration](https://reader034.vdocuments.mx/reader034/viewer/2022042508/577ccd161a28ab9e788b75dc/html5/thumbnails/80.jpg)
PIERO M. ARMENANTENJIT
Design Information for PressureCake Filters
Cycle time: of the order of hours
Solids loading: 0.2-2 lb/ft·h
Solids in cake: up to 50%
Remark: although solids loading in pressurefilters (e.g., filter presses) is typically smaller thanthat of vacuum filters the percentage of solids inthe cake is typically higher. This is the result ofthe higher pressure that can be used in theoperation of pressure filters (as opposed to amaximum of 1 atm in vacuum filters).
![Page 81: Cake Filtration](https://reader034.vdocuments.mx/reader034/viewer/2022042508/577ccd161a28ab9e788b75dc/html5/thumbnails/81.jpg)
PIERO M. ARMENANTENJIT
Design Information for Rotary Vacuum FiltersDiameter: up to 5 mLength: up to 6 mVacuum levels: typically 20 in. Hg (68 kPa)Submergence: 15-25% of drum area
PrimarySludges
Waste-activatedsludges
Solids loading 20-60 kg/m2·h(4-12 lb/ft2·h)
5-20 kg/m2·h(1-4 lb/ft2·h)
Solids in cake 25-40%(typically 20-25%)
10-15%
After Sundstrom and Klei, Wastewater Treatment, 1979, p. 234.
![Page 82: Cake Filtration](https://reader034.vdocuments.mx/reader034/viewer/2022042508/577ccd161a28ab9e788b75dc/html5/thumbnails/82.jpg)
PIERO M. ARMENANTENJIT
Operation of Rotary Vacuum Filters• Solids loading increases with increasing drum
submergence, drum rotational speed, pressuredifference across cake, solids concentration infeed.
• Percentage of solids in cake decreases withincreasing drum submergence, and drumrotational speed.
• Rotary vacuum filters can be used to dewatersludges from activated sludge plants(biological sludges), chemical sludges, andsludges from precipitation operations.
![Page 83: Cake Filtration](https://reader034.vdocuments.mx/reader034/viewer/2022042508/577ccd161a28ab9e788b75dc/html5/thumbnails/83.jpg)
PIERO M. ARMENANTENJIT
Additional Information and Exampleson Cake Filtration
Additional information and examples on can be found inthe following references:
• Sundstrom, D. W. and Klei, H. E., 1979, WastewaterTreatment, Prentice Hall, Englewood Cliffs, NJ, p.229-234.
• Geankoplis, C. J., Transport Processes and UnitOperations, 3rd Edition, 1993, Allyn and Bacon,Boston, pp. 800-815.
• Freeman, H. M. (ed.), 1989, Standard Handbook ofHazardous Waste Treatment and Disposal, McGraw-Hill, New York, pp. 7.3-7.19.
![Page 84: Cake Filtration](https://reader034.vdocuments.mx/reader034/viewer/2022042508/577ccd161a28ab9e788b75dc/html5/thumbnails/84.jpg)
PIERO M. ARMENANTENJIT
Additional Information and Exampleson Cake Filtration
• Haas, C. N. and Vamos, R. J., 1995, Hazardous andIndustrial Waste Treatment, Prentice Hall, EnglewoodCliffs, NJ, pp. 75-78.
• Wentz, C. W., 1995, Hazardous Waste Management,Second Edition, McGraw-Hill, New York. pp. 196-200.
• Vesilind, P. A., 1979, Treatment and Disposal ofWastewater Sludges, Ann Arbor Science, Ann Arbor,MI, pp. 140-161.
![Page 85: Cake Filtration](https://reader034.vdocuments.mx/reader034/viewer/2022042508/577ccd161a28ab9e788b75dc/html5/thumbnails/85.jpg)
PIERO M. ARMENANTENJIT