8 Solid/Liquid Separation: Scale-up of Industrial Equipment
Figure 1.4 shows a vacuum test filter apparatus system that can be used
to obtain the data. The drainage characteristics of the normal ceramic
perforate bottom Buchner funnel are poor and make it unsuitable for
these tests; it is preferable to use a filter funnel with a porous sintered
bottom, or ideally a top-fed leaf assembly equipped with a filter cloth.
To simplify subsequent analysis of the data it is usual to operate the
apparatus at a constant pressure difference (or vacuum). An equivalent
pressure driven apparatus is easily devised for obtaining data using
higher pressure differences across the filter. More sophisticated vacuum
or pressure leaf apparati can be used if a well equipped separations
laboratory is already available. However, the use of complicated
equipment at the early stage of slurry assessment is probably not
justified, but it is wise to use the same type of filter medium in the tests
as is used at full scale so that any medium effects are incorporated into
the experimental results.
75mm diameter
filter funnel;
filter cloth
Vacuum gauges
Water
vacuum
pump
250 ml
measuring
Vacuum cylinder
reservoir
Figure 1.4 Test apparatus showing a top-fed leaf filter for obtaining
vacuum filtration data
The general procedure to obtain the filtration data is as follows:
1. Determine the solids concentration in the slurry sample before
carrying out the filter test(s).
2. With the needle valve A in the vacuum line closed, adjust the bleed
needle valve B to give the required level of (constant) vacuum.
1 - Solid/liquid separation equipment selection 9
This is likely to be in the region of 50 to 70 kPa (380 to 530 mm
Hg). In the wastewater treatment industry the standard used is 49
kPa(386mmHg).
3. Pour the well stirred slurry sample into the filter funnel containing
the filter cloth and open the needle valve A in the vacuum line so
that a pre-set vacuum level is achieved as rapidly as possible in the
graduated cylinder. It is better not to pour the sample directly onto
the cloth, but to feed it onto a perforated plate located 2 to 3 cm
above the cloth. The plate acts as a distributor to (i) prevent solids
approaching the pores in the cloth at an unrealistically high velocity
and thereby causing unexpected plugging of the pores, and (ii)
spread the feed over the full area of the funnel, enabling formation
of a cake of more uniform thickness.
4. Monitor the filtration test by recording the filtration rate by
measuring the volume of filtrate collected at various time intervals.
The intervals between recording the measured volumes need not be
constant but may be increased progressively to compensate for the
gradual drop in filtrate flow rate.
5. If the cake form rate is too slow (e.g. of the order of cm h~^) it may
be desirable to add flocculants or pretreatment chemicals to the
slurry and repeat the filtration tests as appropriate. For slow filtering
slurries it may be necessary to use smaller measuring cylinders.
6. When ten or more sets of volume-time readings have been
obtained fully open both valves to break the vacuum. There should
be some surplus unfiltered slurry visible on top of the cake at this
stage. If there is not, it is likely that the cake will have started to
deliquor and subsequent cake moisture measurements will be
erroneously low, whilst leaving the surplus on the cake will lead to
incorrectly high moisture measurements.
7. Pipette the excess slurry from the surface of the filter cake.
8. If possible, measure the thickness of the filter cake.
9. Remove as much of the cake as possible from the filter, weigh it,
dry it and reweigh it. From these measurements calculate the ratio
of the mass of wet cake to the mass of wet cake.
1.2.2.1 Example calculations from constant pressure test data
The following data were obtained from a leaf filter with an area of 45
cm^ using a pressure difference of 70 kPa (data taken from Wakeman
and Tarleton, 2005). The ratio of the mass of wet cake to mass of dry
10 Solid/Liquid Separation: Scale-up of Industrial Equipment
cake was measured and found to be 1.34, the densities of the solid and
liquid phases are 2640 and 1000 kg m~^ respectively, and the solids
concentration in the feed slurry is 132 kg m~^.
Filtration time, t Filtrate volume, V Cake thickness^
(s) (ml) (cm)
0 0 0
170 141 0.324
275 200 0.460
340 230 0.529
390 252 0.580
457 285 0.656
527 320 0.736
589 341 0.784
660 370 0.851
^Calculated data.
Estimate the cake formation rate from these data.
The mass fraction of solids in the slurry (s) is obtained using:
c'
s = •
f f \
(1)
c +p, 1 -
Ps)
where c' is the solids concentration in the feed (kg m~^), and p/ and p^
are the densities of the liquid and solids respectively (kg m^^).
132
s=- = 0.122
132
132 + 1000 1-
2640
The filtrate volume collected (V) is related to the thickness of cake
formed (L) by:
V= AL (2)
^[AKV-I)+A]
where m^^ is the ratio of the mass of wet cake to the mass of dry cake
and A is the filtration area of the test filter.
1 - Solid/liquid separation equipment selection 11
V - 2640(1-1.34x0.122) ^^^ _ ^^^ 3^
0.122[2640(1.34-1) + 1000]
i.e. Z = 0.0023 F cm
The cake thicknesses shown in the last column in the above table are
calculated from this expression. The rate of cake formation decreases
with time in a constant pressure filtration; in this case it decreases from
0.0019 cm s~^ over the period 0 to 170 s down to 0.0013 cm s~^ over
the period 0 to 660 s; i.e. the cake formation rate is on the order of
0.001 c m s - ^
1.3 Initial selection procedures
An initial list of equipment that could satisfy a particular duty require-
ment is drawn up in the following way (this follows the first few boxes
in the flowchart on Figure 1.1). The first step is a preliminary
specification of the separator duty to define three characterising letter
codes. The second and third steps involve the fairly rudimentary bench
scale tests described in the preceding section. The analysis of the test
data yields overall settling and cake formation rates as well as four
more characterising letter codes. Although it is not necessary to collect
additional information, it is recommended that as much data as possible
are obtained from the tests as these can aid later refinement of an
equipment list. The characterising letter codes defined for required
duty, settling and filtration are then used in conjunction with selection
codings to give a list of equipment that is potentially suited to the
required separation objective(s).
1.3.1 Specification of duty
The first step in defining a selection problem is to specify the general
requirements of the process. An initial specification can be quite
limited and is essentially confined to the process scale, mode of
operation, and overall objective of the separation. These objectives and
unavoidable restrictions can be specified before any experiments are
undertaken. Other specifications, such as the need for filter sterility or
the possibility of toxic or flammable hazards, can be considered at a
later stage in the selection process.
The principles involved in duty specification are shown in Figure 1.5.
Each specification is identified by a characteristic letter, so that a
12 Solid/Liquid Separation: Scale-up of Industrial Equipment
group of letters specify the nature of a problem. For example, a large
scale batch operated system for the recovery of untreated solids would
be coded as adg.
Duty
specification Large >50m^h"'
W Scale Medium 5-50 m^h"'
3u-l
Small <5m'h
Batch
Operation
M Continuous
Objective Clarification
Untreated
M Solids recovery Washed
Dewatered
(dried) in situ
Figure 1.5 Coding the duty specification (Purchas and Wakeman, 1986)
1.3.2 Specification of settling characteristics
The results from the procedure in Section 1.2.1 are coded using a
second set of characteristic letters as shown in Figure 1.6. For
example, a slurry which settles at 3 cm s~^ to yield a clear liquid and a
sediment whose volume is 15% of the initial slurry volume would be
coded BEG,
1.3.3 Specification of filtration characteristics
The results from the test procedure in Section 1.2.2 are coded using a
third set of characteristic letters, as shown in Figure 1.7. For example,
a slurry which forms a cake at the rate of the order of cm min~^ is
1 - Solid/liquid separation equipment selection 13
W Settling characteristics
Modify by pretreatment
<
Rate (cm s') Clarity Proportion
of sludge (%v/v)
Poor Good
H Low < 0.1 Medium High > 5 Low < 2 Medium High > 20
0.1-5 2-20
A B C F G H
F i g u r e 1.6 Coding the slurry settling characteristics (Purchas and Wakeman, 1986)
r r.1,^ ^
L^aivc giuwill laic
Modify by
pretreatment
i
^ ^r
^r ^r ^r ^r
Negligible Slow Medium Fast
. -1 -1
cake cm h cm mm cm s
I K
F i g u r e 1.7 Coding the slurry filtration characteristics (Purchas and
Wakennan, 1986)
coded K. Combining this with the settling characteristics (e.g. BEG)
gives a total preliminary description of the separation characteristics of
the slurry (e.g. BEG, K). If the proposed duty is simply to thicken a
slurry then it is not necessary to carry out a filtration test. However, for
a total separation of the solid from the liquid (as obtained in a filter, for
example) both settling and filtration tests need to be performed.
