Centrifugation
9,096 views
74 slides
Nov 15, 2013
Slide 1 of 74
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
About This Presentation
SYNOPSIS
The principles underlying centrifugal separation of particulate species are briefly considered, and the main types of separator available are noted. The procedures available for scale-up from laboratory or semi-technical data are then discussed in detail with particular reference to perhaps...
Slide Content
41,
{af C?PT.
NO 1 CATALYST, PROCESS TECHNOLOGY
CONSULTANCY
CENTRIFUGATION
Process Information Disclaimer
believed
{af C?PT.
NO 1 CATALYST, PROCESS TECHNOLOGY
CONSULTANCY
CENTRIFUGATION
Process Information Disclaimer
believed
INTRODUCTION
Centriluge and How Does It Work?
Operating Regimes - What Limits the Machine's Output?
of separation
Characteristics
of Separation
(solids-imitation)
apabilty (Hydraulic limitation)
ling (Flux limitation)
USEFUL LABORATORY TESTS
Background
24.1 Basic Suspension Proper
1.2 Settling Kinetics
4.3. Network Strength N
241.4 Floc Stability Ass
Test Techniques
Centriluge and How Does It Work?
Operating Regimes - What Limits the Machine's Output?
of separation
Characteristics
of Separation
(solids-imitation)
apabilty (Hydraulic limitation)
ling (Flux limitation)
USEFUL LABORATORY TESTS
Background
24.1 Basic Suspension Proper
1.2 Settling Kinetics
4.3. Network Strength N
241.4 Floc Stability Ass
Test Techniques
Scale-Up Factor
Meth
Semi-Empirical Approach I!
Ab Initio Approach
mation of N
Background.
sary Parameters and the Problems
) Factor
sept of Ultimate Thicknes
e-Up When Hindered is the Limiting Process
2 Gravity Thickeners — Behavior
Batch Thickeners
1 _ Continuous Thickeners
Gravity Thickeners — Interpretation of Be
plication to Centrifuges
Batch Flux Curve
Meth
Semi-Empirical Approach I!
Ab Initio Approach
mation of N
Background.
sary Parameters and the Problems
) Factor
sept of Ultimate Thicknes
e-Up When Hindered is the Limiting Process
2 Gravity Thickeners — Behavior
Batch Thickeners
1 _ Continuous Thickeners
Gravity Thickeners — Interpretation of Be
plication to Centrifuges
Batch Flux Curve
3: Prediction of Ultimate Thickness for Solids Limited Syste
Steps in Semi-Empirical Scale-Up for Solid
Limited System
Steps in Ab Initio Scale-Up for Solids Limited Sy
Steps in Semi-Empirical Scale-Up for System Limited
by Hindered Setting Si
Steps in Ab Initio G
Hindered Setting
Worked Examples
Solids-Limited Scale-Up: Background to Examples
olids-Limited Examples: Semi-Empirical and Ab Initio Methods
Semi-Empirical Approach to Solids Limited Operation
(Gimpified method)
52.2 Standard Semi-Empirical Approac!
523 Ab Initio Approach to Solids Limitation
Calculation of Thickening Ultimate Limit
Example: "Structural" limit for Thickening in “Pruteen” Centrifugation
Scales-Up for Clarification Limited Systems
54.1 Ab Initio Clarification Calculation for P2 Type Single Cell Protein
5.4.2 (Ab Initio) Clarification Limit for Unflocculated Bacterial Cells
54.3 Semi£s 'Scale-Up for Clarification
Limited Sy
Ab Initio Calculation for Hindered Setting Li
Steps in Semi-Empirical Scale-Up for Solid
Limited System
Steps in Ab Initio Scale-Up for Solids Limited Sy
Steps in Semi-Empirical Scale-Up for System Limited
by Hindered Setting Si
Steps in Ab Initio G
Hindered Setting
Worked Examples
Solids-Limited Scale-Up: Background to Examples
olids-Limited Examples: Semi-Empirical and Ab Initio Methods
Semi-Empirical Approach to Solids Limited Operation
(Gimpified method)
52.2 Standard Semi-Empirical Approac!
523 Ab Initio Approach to Solids Limitation
Calculation of Thickening Ultimate Limit
Example: "Structural" limit for Thickening in “Pruteen” Centrifugation
Scales-Up for Clarification Limited Systems
54.1 Ab Initio Clarification Calculation for P2 Type Single Cell Protein
5.4.2 (Ab Initio) Clarification Limit for Unflocculated Bacterial Cells
54.3 Semi£s 'Scale-Up for Clarification
Limited Sy
Ab Initio Calculation for Hindered Setting Li
Starting with the basic concepts behind their design, discussion follows to explain
may limit centrifuge performance. I is shown how a few simple,
1s can give a valuable insight into the design and operation o
full-scale industrial machines. Various scale-up theories are presented and
orked samples given, using data drawn from GBHE Agricultural Di
experience in biological separation:
1 INTRODUCTION
1.1. Scope of the Section
Centrifugal separation of solids is governed by the action of a centrifugal field in
either enhancing setting of particles or promoting deliquoring by filtration. The
former effect is the more important for fine particie processing (species < 10 um
in size > and thus we have chosen to focus attention on such operations in this
ction.
Fitration in general is considered in GBHE SPG PEG 300 Filtration.
ion is centered upon the scale-up and operation of (continuous)
-nozzle machines: a significant amount of data is available for this very
important class of centrifuges to demonstrate the strength and limitations of
scale-up procedures. It is considered that many of the concepts described can be
readily extended to batch or intermittently discharging machines, and also to.
Centrituges of radically diferent design, such as decanters. However at the
rent time there are insufficient working data to warrant detailed discussion of
may limit centrifuge performance. I is shown how a few simple,
1s can give a valuable insight into the design and operation o
full-scale industrial machines. Various scale-up theories are presented and
orked samples given, using data drawn from GBHE Agricultural Di
experience in biological separation:
1 INTRODUCTION
1.1. Scope of the Section
Centrifugal separation of solids is governed by the action of a centrifugal field in
either enhancing setting of particles or promoting deliquoring by filtration. The
former effect is the more important for fine particie processing (species < 10 um
in size > and thus we have chosen to focus attention on such operations in this
ction.
Fitration in general is considered in GBHE SPG PEG 300 Filtration.
ion is centered upon the scale-up and operation of (continuous)
-nozzle machines: a significant amount of data is available for this very
important class of centrifuges to demonstrate the strength and limitations of
scale-up procedures. It is considered that many of the concepts described can be
readily extended to batch or intermittently discharging machines, and also to.
Centrituges of radically diferent design, such as decanters. However at the
rent time there are insufficient working data to warrant detailed discussion of
In “centrifugal fiters” the phenomenon which gives concentration of the solids is
a fitration mechar are trapped by a filer medium of some
description a erated by the field. Estimates of
the performance of such operations are made by appropriate modification of
ory, based on Darcy's Law ions 3.2 and 3.5)
allow for the “g" force. Procedure
as [1], [7] and [10]. Though f ese predictive techniques are
ni avaable. As a rl cntfuga ers ae not
ecies are =: 100 um or gre
ofthe this manual and wll ol be further considered
In centrifuges based upon a settling principle, it is migration of the solid particles
through the liquid phase that is enhanced by the centrifugal field. As will be seen
later, this may involve either sedimentation of single particles in a Stoke's Law
type of “ree fall” or consolidation of a structured (thickened> sludge. As noted
above, settling rather than filtering centrifuges are appropriate for fine species
particles down to the order of 0.01 um can be dealt with in appropriate
Circumstances, Machines come in very many designs — disc nozzles, decanters
and so on - all of which have their particular operating regimes. Details of the
principal kinds of setting centrifuge are given in reference [10]. It should also be
Noted that centrifuges may have intermittent or continuous solids discharge or
may be operated in batch mode.
Figure 1 represents schematically the continuous disc-nozzle centrifuge which ls
in fine particle Feed Is introduced into the
a fitration mechar are trapped by a filer medium of some
description a erated by the field. Estimates of
the performance of such operations are made by appropriate modification of
ory, based on Darcy's Law ions 3.2 and 3.5)
allow for the “g" force. Procedure
as [1], [7] and [10]. Though f ese predictive techniques are
ni avaable. As a rl cntfuga ers ae not
ecies are =: 100 um or gre
ofthe this manual and wll ol be further considered
In centrifuges based upon a settling principle, it is migration of the solid particles
through the liquid phase that is enhanced by the centrifugal field. As will be seen
later, this may involve either sedimentation of single particles in a Stoke's Law
type of “ree fall” or consolidation of a structured (thickened> sludge. As noted
above, settling rather than filtering centrifuges are appropriate for fine species
particles down to the order of 0.01 um can be dealt with in appropriate
Circumstances, Machines come in very many designs — disc nozzles, decanters
and so on - all of which have their particular operating regimes. Details of the
principal kinds of setting centrifuge are given in reference [10]. It should also be
Noted that centrifuges may have intermittent or continuous solids discharge or
may be operated in batch mode.
Figure 1 represents schematically the continuous disc-nozzle centrifuge which ls
in fine particle Feed Is introduced into the
Figure 1
In this section we will focus our discussion on this kind of process. We wil
consider the way that the output of a given centrifuge is limited by one of three
main mechanisms, viz:
(0. Falure to capture the fine panicles contained in the otherwise clear
contrate.
(} Falure to give sufficient time for the concentrated stream to adequately
thicken,
(ii) Flux limitation caused by particle;parlcle interaction
The first of these is likely to be the limiting mechanism when centrifuging weak
suspensions of discrete, compact particles of small diameter. The second oc
ing to obtain a thick concentrate from an open, flocculated sı
In this section we will focus our discussion on this kind of process. We wil
consider the way that the output of a given centrifuge is limited by one of three
main mechanisms, viz:
(0. Falure to capture the fine panicles contained in the otherwise clear
contrate.
(} Falure to give sufficient time for the concentrated stream to adequately
thicken,
(ii) Flux limitation caused by particle;parlcle interaction
The first of these is likely to be the limiting mechanism when centrifuging weak
suspensions of discrete, compact particles of small diameter. The second oc
ing to obtain a thick concentrate from an open, flocculated sı
be obtained on any commer rifuge. An insight into the limiting mechanism
will also allow a quantitative nent of the benefit of modifying proces:
ditions (e. . feed concentration > or machine parameters (e.g. number of
discs) without the need for time consuming, expensive trials. An informed
judgment on the suitability of centrifugation for a given du
obtained,
Currently, the battery of scale-up techniques available for disc-nozzle machin
have not been properly extended to, for example, deca robable that
the concepts of Section 3.4.3 are applicable in sı
obviously inadequate, In particular, decanters are perhaps at best ac
ening to high solids (when their method of solids discharge is more
appropriate than that of the disc-nozzle design). In such a regime it will be
probable that the process Is solids rather than clarification limited as is assumed
inthe E approach.
13 _ What isa Disc Centrifuge and How Does It Work?
sider a simple clarifying tank in which a single, dense particle is setting
ut Under the force of gravity: Figure 2
will also allow a quantitative nent of the benefit of modifying proces:
ditions (e. . feed concentration > or machine parameters (e.g. number of
discs) without the need for time consuming, expensive trials. An informed
judgment on the suitability of centrifugation for a given du
obtained,
Currently, the battery of scale-up techniques available for disc-nozzle machin
have not been properly extended to, for example, deca robable that
the concepts of Section 3.4.3 are applicable in sı
obviously inadequate, In particular, decanters are perhaps at best ac
ening to high solids (when their method of solids discharge is more
appropriate than that of the disc-nozzle design). In such a regime it will be
probable that the process Is solids rather than clarification limited as is assumed
inthe E approach.
13 _ What isa Disc Centrifuge and How Does It Work?
sider a simple clarifying tank in which a single, dense particle is setting
ut Under the force of gravity: Figure 2
le up > vb
Therefore the maximum feed rate that the tank can clarfy is given by
= usbl
Le. Max Liquid Rate = Tank Surface Area x Settling Velocity under gravity
The performance of the tank can be improved in one of two ways:
By increasing the gravitational field. An n-fold increase will give an n-fold
Increase in settling velocity and hence allow an n-fold increase In
throughput.
By reducing th e height the particle has to fall, e.g. by paritioning the
tank as shown below: Figure 3
Therefore the maximum feed rate that the tank can clarfy is given by
= usbl
Le. Max Liquid Rate = Tank Surface Area x Settling Velocity under gravity
The performance of the tank can be improved in one of two ways:
By increasing the gravitational field. An n-fold increase will give an n-fold
Increase in settling velocity and hence allow an n-fold increase In
throughput.
By reducing th e height the particle has to fall, e.g. by paritioning the
tank as shown below: Figure 3
Biawewes
Field
LL.
ho tt
Feed
Figure 4
This is not a practical arrangement because there is no convenient means of
separating solid and liquid flows. In practice, discs are Inclined to the
gravitational field with the feed entering through holes in the discs. The solids
migrate along the underside of the discs to the bow periphery whilst the clarified
supernatant flows inwards.
Field
LL.
ho tt
Feed
Figure 4
This is not a practical arrangement because there is no convenient means of
separating solid and liquid flows. In practice, discs are Inclined to the
gravitational field with the feed entering through holes in the discs. The solids
migrate along the underside of the discs to the bow periphery whilst the clarified
supernatant flows inwards.
Figure 5
Liquid inventory in the centrifuge is kept constant by a weir arrangement as
below, the liquid finding Its own level with respect to the applied
tational field. The thickened solids discharge through a number of nozzles.
at the periphery of the bow, the discharge rate being controlled (for some
machines) by varying the nozzle diameter; less sophisticated machines may no
have on-line control but different sizes of nozzles will generally be available.
‘will also be noted that the feed enters the machine axially, typically from above.
Ribbed rotating surfaces in the feed zone help accelerate the Incoming liquid to
velocity. Any necessary further acceleration wil be imparted by the rotating
disc-stack
As a further variation, the solids may be transferred through tubes from the bow!
apex back towards the centre of the machine before reaching a nozzle and being
discharged. The reduced hydrostatic pressure here (smaller radius of rotation
and lower quid depth > mean tha larger nozzles can be used forth
Liquid inventory in the centrifuge is kept constant by a weir arrangement as
below, the liquid finding Its own level with respect to the applied
tational field. The thickened solids discharge through a number of nozzles.
at the periphery of the bow, the discharge rate being controlled (for some
machines) by varying the nozzle diameter; less sophisticated machines may no
have on-line control but different sizes of nozzles will generally be available.
‘will also be noted that the feed enters the machine axially, typically from above.
Ribbed rotating surfaces in the feed zone help accelerate the Incoming liquid to
velocity. Any necessary further acceleration wil be imparted by the rotating
disc-stack
As a further variation, the solids may be transferred through tubes from the bow!
apex back towards the centre of the machine before reaching a nozzle and being
discharged. The reduced hydrostatic pressure here (smaller radius of rotation
and lower quid depth > mean tha larger nozzles can be used forth
Figure 6
14 Operating Regimes - What Limits the Machine's Output?
1.4.1 The Mechanics of separation
Despite engineering complexities (see Figure 6) the mechanics of separation in a
disc-nozzle centrifuge are straightforward. Feed Introduced to the disc stack Is
conveyed to the periphery by the strong centrifugal field consequent on the
rotation of the disc stack and bowl. The thickened solids are discharged through
nozzles whilst the less dense liquid is forced out at the weir by hydrostatic
pressure. The specific capacity of a centrifuge to concentrate a particular
14 Operating Regimes - What Limits the Machine's Output?
1.4.1 The Mechanics of separation
Despite engineering complexities (see Figure 6) the mechanics of separation in a
disc-nozzle centrifuge are straightforward. Feed Introduced to the disc stack Is
conveyed to the periphery by the strong centrifugal field consequent on the
rotation of the disc stack and bowl. The thickened solids are discharged through
nozzles whilst the less dense liquid is forced out at the weir by hydrostatic
pressure. The specific capacity of a centrifuge to concentrate a particular
because the pressure drop across the nozzle is determined solely by the liquid
depth and centrifugal field In the machine,
Increasing either the feed rate or feed concentration will increase the quantity
lids to be discharged in this constant volume, and hence increase the thicks
ncentration. This represents one of those rare occasions when driving a piece
‘equipment harder will improve its performance
Eventually, however, the machine's separation capacity will be exceeded (for one
of a number of reasons to be discussed later) and solids will be lost into the
trate. The point at which this led the breakthrough point
Effect of Nozzle Diameter
es) will allow a greater
te of thicks to pass. I is now possible to reach a new breakthrough
Point where the solids throughput is higher. Generally, however, the solids
Concentration will have been reduced (see later,
The locus of breakthrough points which can be achieved by changing the noz:
diameter Is termed the breakthrough curve. By using a variable diameter nc
itis possible to operate with clear centrate at any point on or below this curv
The above points are sketched below, where thicks concentration is plotted
against solids throughput. The latter represents a measure of both feed rate and
concentration,
depth and centrifugal field In the machine,
Increasing either the feed rate or feed concentration will increase the quantity
lids to be discharged in this constant volume, and hence increase the thicks
ncentration. This represents one of those rare occasions when driving a piece
‘equipment harder will improve its performance
Eventually, however, the machine's separation capacity will be exceeded (for one
of a number of reasons to be discussed later) and solids will be lost into the
trate. The point at which this led the breakthrough point
Effect of Nozzle Diameter
es) will allow a greater
te of thicks to pass. I is now possible to reach a new breakthrough
Point where the solids throughput is higher. Generally, however, the solids
Concentration will have been reduced (see later,
The locus of breakthrough points which can be achieved by changing the noz:
diameter Is termed the breakthrough curve. By using a variable diameter nc
itis possible to operate with clear centrate at any point on or below this curv
The above points are sketched below, where thicks concentration is plotted
against solids throughput. The latter represents a measure of both feed rate and
concentration,
Figure 7
1.4.2.3 Theory of Separation.
The detailed hydr high-speed centrifuges are horrendous. Even
simplified” treatments are often mathematically and conceptually comple)
despite the gross assumptions that are usually made. Only concepts direct
relevant to separation power will therefore be discussed,
For the purposes of many applications the separating power of a centrifuge may
be defined as the rate and concentration of thicks attainable at a given solids
throughput while maintaining a nominally clear centrate.
This immediately suggests two separate mechanisms which could limit centrifuge
performance, ie. the ability to thicken concentrate and the ability to clarify
1.4.2.3 Theory of Separation.
The detailed hydr high-speed centrifuges are horrendous. Even
simplified” treatments are often mathematically and conceptually comple)
despite the gross assumptions that are usually made. Only concepts direct
relevant to separation power will therefore be discussed,
For the purposes of many applications the separating power of a centrifuge may
be defined as the rate and concentration of thicks attainable at a given solids
throughput while maintaining a nominally clear centrate.
This immediately suggests two separate mechanisms which could limit centrifuge
performance, ie. the ability to thicken concentrate and the ability to clarify
Figure 8
The gravitational field increases with the square of the rotational
proportional to the radial distance from the axis of the machine. The time
available for compaction wil be roughly proportional o the volume of the
peripheral space and is inversely proportional to throughput.
Two points then follow
The gravitational field increases with the square of the rotational
proportional to the radial distance from the axis of the machine. The time
available for compaction wil be roughly proportional o the volume of the
peripheral space and is inversely proportional to throughput.
Two points then follow
unimportant for solids-Imited applications;
the dis rification rather than residence times for
solids compaction. If a machine is thought to be solids limited, there Is therefore
litle point In installing a pre-thickening stage or modifying the discs.
1.6 Clarification capability (Hydraulic limitation)
The following sketch illustrates the paths followed by particles in the disc
Solids slide outwards
on underside
the dis rification rather than residence times for
solids compaction. If a machine is thought to be solids limited, there Is therefore
litle point In installing a pre-thickening stage or modifying the discs.
1.6 Clarification capability (Hydraulic limitation)
The following sketch illustrates the paths followed by particles in the disc
Solids slide outwards
on underside
ack configuration is similarly important: doubling the number of di
roughly halve the distance each particle has to travel before being captured. This
will theoretically double machine capacity, although other effects, e.g. disc
thickness, flow maldistribution, etc, may be significant
Clarification limitation can be readily identified If different feed concentrations are
fed to a machine because a whole family of breakthrough curves will be
obtained, one for each feed. This follows from a simple mass balance If It Is
recognized that breakthrough will always occur at a constant centrate rate.
A comparison between solids and clarification limitation Is sho
roughly halve the distance each particle has to travel before being captured. This
will theoretically double machine capacity, although other effects, e.g. disc
thickness, flow maldistribution, etc, may be significant
Clarification limitation can be readily identified If different feed concentrations are
fed to a machine because a whole family of breakthrough curves will be
obtained, one for each feed. This follows from a simple mass balance If It Is
recognized that breakthrough will always occur at a constant centrate rate.
A comparison between solids and clarification limitation Is sho
limitation is less obvious than the previous two but may b
importa ll known effect in gravitational thickener
liquid
‘generally proportional to the applied gravitational field but
as the solids concentration builds up. This latter phenomenon is called
ling”. I results from increasingly significant effects of hydrodynamic
interaction between setting species ge distance between them
reduced by rising solids content.
For a constant volumetric flowrate of fluid the solids flux must, in the absence of
particle-particle Interactions, be proportional to solids concentration. If hindered
settling is taken into account, however, It can be shown that there is an absolute
maximum to the solids flux that can be obtained,
The effect is typified in gravity settlers by a critical concentration which fils the
settler; solids In excess of the maximum flux are lost to the clarified liquor. The
tical concentration reached depends on the feed concentration but will typically
be before compression effects come into play,
There are two possible places where a maximum solids flux may limit centrifuge
performance:
(1) Inthe di ids are setting out of the clarifying centrate.
importa ll known effect in gravitational thickener
liquid
‘generally proportional to the applied gravitational field but
as the solids concentration builds up. This latter phenomenon is called
ling”. I results from increasingly significant effects of hydrodynamic
interaction between setting species ge distance between them
reduced by rising solids content.
For a constant volumetric flowrate of fluid the solids flux must, in the absence of
particle-particle Interactions, be proportional to solids concentration. If hindered
settling is taken into account, however, It can be shown that there is an absolute
maximum to the solids flux that can be obtained,
The effect is typified in gravity settlers by a critical concentration which fils the
settler; solids In excess of the maximum flux are lost to the clarified liquor. The
tical concentration reached depends on the feed concentration but will typically
be before compression effects come into play,
There are two possible places where a maximum solids flux may limit centrifuge
performance:
(1) Inthe di ids are setting out of the clarifying centrate.
Concentration. Such information should be yielded by a strobe centrifuge te
1.8 Comments
(1) The underlying principles of centrifuge operation are relatively easy to
comprehend even I the mathematics are frightening
(2) A few strobe centrifuge and pulse shearometer tests of material on a
laboratory sele can theoretical vid al the formation required to
predict centrtuge performance.
interpretation and provide a us
used.
the previous discus
ortex formation, counter-current shear, etc). A
model is therefore a remote possibilty. However,
the general principles discussed should hold to some extent and an
estimate of likely centrifuge limitations should be possible.
In subsequent sections we detail useful laboratory tests for prediction of
Centrifuge performance; develop scale-up theories; and describe and exemplity
the procedures necessary for their application
1.8 Comments
(1) The underlying principles of centrifuge operation are relatively easy to
comprehend even I the mathematics are frightening
(2) A few strobe centrifuge and pulse shearometer tests of material on a
laboratory sele can theoretical vid al the formation required to
predict centrtuge performance.
interpretation and provide a us
used.
the previous discus
ortex formation, counter-current shear, etc). A
model is therefore a remote possibilty. However,
the general principles discussed should hold to some extent and an
estimate of likely centrifuge limitations should be possible.
In subsequent sections we detail useful laboratory tests for prediction of
Centrifuge performance; develop scale-up theories; and describe and exemplity
the procedures necessary for their application
in the laboratory and estimate behavior o
route Is, as a rule, less expensive and time-consuming and it giv
capacity unattainable by the empirical approach. Also, when novel so
aration problems appear provides ihe oy method of obaning oder
etely "hit or
We would caution, however, that
because of ihe
scale assessment of suspen
to test predictor
The laboratory tests used to characterize the centrifugation of a suspension are
le in concept. They are best understood
reference to the basic phenomena underiying all separations dependent upon
lids setting, whether accelerated or not. Figures 11 and 12 ilustrate these
effects. The first of the diagrams shows typical setting curves under normal
gravity and under enhanced "g’. In the Initial stages the sedimentation graph
tends to be linear, then setting slows as particle interaction or "crowding?
occurs. Eventually, solids content reaches a limiting value as the forces
[promoting consolidation are balanced by the internal strength of the sediment.
route Is, as a rule, less expensive and time-consuming and it giv
capacity unattainable by the empirical approach. Also, when novel so
aration problems appear provides ihe oy method of obaning oder
etely "hit or
We would caution, however, that
because of ihe
scale assessment of suspen
to test predictor
The laboratory tests used to characterize the centrifugation of a suspension are
le in concept. They are best understood
reference to the basic phenomena underiying all separations dependent upon
lids setting, whether accelerated or not. Figures 11 and 12 ilustrate these
effects. The first of the diagrams shows typical setting curves under normal
gravity and under enhanced "g’. In the Initial stages the sedimentation graph
tends to be linear, then setting slows as particle interaction or "crowding?
occurs. Eventually, solids content reaches a limiting value as the forces
[promoting consolidation are balanced by the internal strength of the sediment.
sonia
force (beyond this pint onsclidation
= are un alow to be We
Figure 11
Typical setting curves for a suspension sedimenting under
er | Consol
mente! [compression
Solids Content
force (beyond this pint onsclidation
= are un alow to be We
Figure 11
Typical setting curves for a suspension sedimenting under
er | Consol
mente! [compression
Solids Content
is colloidally stable, ie. there are repulsive interactions between the
Here the ultimate solids content achievable is when the particles form a
to "ng" is two-fold
sualy nearly with n In the early stages) because o {the greater
ration, and the "structural limit is pushed up as greater
ids content> is required to resist the incre:
Impression. The setting curves are dependent upon suspens ristic
(e.g. degree/strength of flocculation) and the physical set-up of the experiment
used to derive the data < e.g. the number of "g" employed; height of
diment used). In principle, a family of such graphs contain u the inf
a particular suspension needed for prediction o its large-scale centrifuge
behavior. Initial setting rates allow calculation of throughput in the clarification
regime, whilst those for later stages of consolidation are needed for estimation of
ntrifuge requiremel er solids.imitation conditions. The behavior of the
tructural limit” 9" also enables calculation of the maximum
solids that could ever be attained from a particular size/speed of machine,
irrespective of the manipulation performed to increase residence time available
for consolidation.
In practice, it Is useful to have a bit more data than simply the raw settling curves.
for a sequence of accelerating conditions, There are a number of reasons for
this, perhaps the most important being the difficulty in identifying transitions from,
‘say, "free fall” of particles to a hindered setting or compression regime (Section
33.1). However, this problem is readily circumvented by measuring the shear
ure of strength of any cohering
Here the ultimate solids content achievable is when the particles form a
to "ng" is two-fold
sualy nearly with n In the early stages) because o {the greater
ration, and the "structural limit is pushed up as greater
ids content> is required to resist the incre:
Impression. The setting curves are dependent upon suspens ristic
(e.g. degree/strength of flocculation) and the physical set-up of the experiment
used to derive the data < e.g. the number of "g" employed; height of
diment used). In principle, a family of such graphs contain u the inf
a particular suspension needed for prediction o its large-scale centrifuge
behavior. Initial setting rates allow calculation of throughput in the clarification
regime, whilst those for later stages of consolidation are needed for estimation of
ntrifuge requiremel er solids.imitation conditions. The behavior of the
tructural limit” 9" also enables calculation of the maximum
solids that could ever be attained from a particular size/speed of machine,
irrespective of the manipulation performed to increase residence time available
for consolidation.
In practice, it Is useful to have a bit more data than simply the raw settling curves.
for a sequence of accelerating conditions, There are a number of reasons for
this, perhaps the most important being the difficulty in identifying transitions from,
‘say, "free fall” of particles to a hindered setting or compression regime (Section
33.1). However, this problem is readily circumvented by measuring the shear
ure of strength of any cohering
mé and the
‘material starts to exhibit substantial s. It should be
noted that the 'rheologi
hine will be greater than that for a dk
31 of solids discharge.
which the “structural limit” Is
tifuge to thicken
impractical long. Thus if calculations sugges
is just capable of thickening to the required degree,
sub Ke not prove
In Figure 11 we have indicated the kind of "practical" thick
seen in operation
24.1 Basic Suspension Properties
Before carrying out any other measurement til to obtain
certain basic data concerning the feed, viz the particle size, the density of
the solids, the state of aggregation of the species. Simple procedures,
such as optical microscopy, are often entirely appropriate to the purpose.
Further details on methods which can be applied are given In reference
[19]. For biological particles such as bacterial cells the density may often
be accurately measured via a silica suspension density gradient (2
24.2 Settling Kinetics
‘material starts to exhibit substantial s. It should be
noted that the 'rheologi
hine will be greater than that for a dk
31 of solids discharge.
which the “structural limit” Is
tifuge to thicken
impractical long. Thus if calculations sugges
is just capable of thickening to the required degree,
sub Ke not prove
In Figure 11 we have indicated the kind of "practical" thick
seen in operation
24.1 Basic Suspension Properties
Before carrying out any other measurement til to obtain
certain basic data concerning the feed, viz the particle size, the density of
the solids, the state of aggregation of the species. Simple procedures,
such as optical microscopy, are often entirely appropriate to the purpose.
Further details on methods which can be applied are given In reference
[19]. For biological particles such as bacterial cells the density may often
be accurately measured via a silica suspension density gradient (2
24.2 Settling Kinetics
‚dimentation. Alternati
imated from first pr
the centrifugation experiment
em the larger particles will often have formed cakes at
of the tube before the smaller ones have settled appreciably
These systems require the appropriate sedimentation velocity of a given
clarification criterion: e.g. ifthe separation criterion Is 100% clarity of
centrate then the sedimentation
appropriate quantity for on (see Example (d)
345)
Further details on the precautior ved in the measurement
and interpretation of setting kinetics are given in reference [19].
Though strobe centrifuge m rovide the best approach to
characterizing settling kinetics, we would note that information obtained
from even a basic bench centrifuge is better than nothing. Indeed this
appears to be the procedure followed by centrifuge manufacturers under
h names as "spin test" [6]
241.3. Network Strength Measurements
imated from first pr
the centrifugation experiment
em the larger particles will often have formed cakes at
of the tube before the smaller ones have settled appreciably
These systems require the appropriate sedimentation velocity of a given
clarification criterion: e.g. ifthe separation criterion Is 100% clarity of
centrate then the sedimentation
appropriate quantity for on (see Example (d)
345)
Further details on the precautior ved in the measurement
and interpretation of setting kinetics are given in reference [19].
Though strobe centrifuge m rovide the best approach to
characterizing settling kinetics, we would note that information obtained
from even a basic bench centrifuge is better than nothing. Indeed this
appears to be the procedure followed by centrifuge manufacturers under
h names as "spin test" [6]
241.3. Network Strength Measurements
stabilty are rat in diagnostic
rough guide to actual behavior In a lar
for this he difficulty i
and elongational stresses prevalent in a
‘apparatus cannot be designed beca!
1, some assessment can be
uspension towards breakup, albeit
a rather empirical nature. The standard approach involves measuring a
quantity dependent on Floc size, subjecting the suspension to a
prescribed amount of agitation from some kind of mixer, and then re-
measuring the property. The degree to which the two values differ is an
in) stability
Settling rate (as determined by, say, the strobe centrifuge) ls a good Floc-
size-dependent characteristic to observe, though fitration times are also
employed [8.9]. (Fitration is generally faster the larger the Flocs. If
ignficant loe breakup occurs, fitration rates slow dramatically due both
to reduced cake permeability and to binding effects. ) The problems arise
with respect to prescription of the appropriate degree of agitation for the
reasons mentioned above. Commercial apparatus (ex Triton Electronics
Limited) ls available for the task (see reference (81, p179] but a simple
gitator in a beaker probably serves as well. Clearly the method cannot
work on an absolute basis but satisfactory results can be obtained by
comparison.
In addition a relatively new instrument produced by Rank Brothers, the
Photometric Dispersion Analyzer, allows the correlation of Floc size and
rough guide to actual behavior In a lar
for this he difficulty i
and elongational stresses prevalent in a
‘apparatus cannot be designed beca!
1, some assessment can be
uspension towards breakup, albeit
a rather empirical nature. The standard approach involves measuring a
quantity dependent on Floc size, subjecting the suspension to a
prescribed amount of agitation from some kind of mixer, and then re-
measuring the property. The degree to which the two values differ is an
in) stability
Settling rate (as determined by, say, the strobe centrifuge) ls a good Floc-
size-dependent characteristic to observe, though fitration times are also
employed [8.9]. (Fitration is generally faster the larger the Flocs. If
ignficant loe breakup occurs, fitration rates slow dramatically due both
to reduced cake permeability and to binding effects. ) The problems arise
with respect to prescription of the appropriate degree of agitation for the
reasons mentioned above. Commercial apparatus (ex Triton Electronics
Limited) ls available for the task (see reference (81, p179] but a simple
gitator in a beaker probably serves as well. Clearly the method cannot
work on an absolute basis but satisfactory results can be obtained by
comparison.
In addition a relatively new instrument produced by Rank Brothers, the
Photometric Dispersion Analyzer, allows the correlation of Floc size and
Centrfuging biological suspensions due to their buoy
proved. Simple sedimentation tests also have their place - settling usually
stops under “ig” Just where network strength becomes significant. Thus the
suspension concentration process up to this point wil probably be
clarification limited.
22 Test Techniques
Laboratory test methods resolve themselves into four main classes:
(1) Basic suspension properties
Procedures for measuring seing kinetics
Network strength characterization methods;
Techniques for assessing the likelihood of aggregate breakup
in the machine.
proved. Simple sedimentation tests also have their place - settling usually
stops under “ig” Just where network strength becomes significant. Thus the
suspension concentration process up to this point wil probably be
clarification limited.
22 Test Techniques
Laboratory test methods resolve themselves into four main classes:
(1) Basic suspension properties
Procedures for measuring seing kinetics
Network strength characterization methods;
Techniques for assessing the likelihood of aggregate breakup
in the machine.
3.1.1. Qualitative
Referring back to Section 3.4.1(c), if the solid particle is unable to reach
tling tank before leaving, then it wll pass out in
supernatant stream, ie full clarification of the liquid overflow will not be
was shown that:
ity under one
The clafication capacity of a disc centrifuge can similarly be quantified
and is given the symbol
The E value of a centrifuge represents the area of a simple gravity setting
tank (m?) which would be required to give the same separation capacity
Quantitative
Calculation of the E value of a known centrifuge is reasonabl
straightforward and the standard derivation is outlined below:
For a single disc
Referring back to Section 3.4.1(c), if the solid particle is unable to reach
tling tank before leaving, then it wll pass out in
supernatant stream, ie full clarification of the liquid overflow will not be
was shown that:
ity under one
The clafication capacity of a disc centrifuge can similarly be quantified
and is given the symbol
The E value of a centrifuge represents the area of a simple gravity setting
tank (m?) which would be required to give the same separation capacity
Quantitative
Calculation of the E value of a known centrifuge is reasonabl
straightforward and the standard derivation is outlined below:
For a single disc
Inclined dise
Figure 13
Area of disc between r and r+ 5 ris given by
SA = Zar 6r
sin 8
w (rads/s)
Figure 13
Area of disc between r and r+ 5 ris given by
SA = Zar 6r
sin 8
w (rads/s)
3.1.3 Validity
Although Westfalia Separator ‘equation for 3, Al
an empirically “corrected” version of E, called KQ, where
wie cot 8 irre -
Centrfuge capacities may be compared by means of their relative KO
interpretation.
The use of either E or KQ is widespread and indeed often considered to
be the only scale-up parameter of importance. It is, however, only
‘meaningful when clarification isthe limiting process. It has no relevance to
solids-imited applications,
Other mechanical factors may also reduce the applicability of the above
formula, e.g. 100 narrow disc spacing or Inadequate feed acceleration
Other factors that should ideally be considered include the significant
Coriolis forces present, and the inevitable in actual terminal
locites.
Although Westfalia Separator ‘equation for 3, Al
an empirically “corrected” version of E, called KQ, where
wie cot 8 irre -
Centrfuge capacities may be compared by means of their relative KO
interpretation.
The use of either E or KQ is widespread and indeed often considered to
be the only scale-up parameter of importance. It is, however, only
‘meaningful when clarification isthe limiting process. It has no relevance to
solids-imited applications,
Other mechanical factors may also reduce the applicability of the above
formula, e.g. 100 narrow disc spacing or Inadequate feed acceleration
Other factors that should ideally be considered include the significant
Coriolis forces present, and the inevitable in actual terminal
locites.
Similarly, a centrituge of equivalent setting area E (m?) could cope with a
liquid throughput (Le.
ugh (m/s)
A more empirical approach may be preferable if data trom e.
technical centrifuges are available. If the performance of a sm:
is known the likely throughput on a large machine may be pred
the ratio of the respective E values for Ihe two:
n Machine
lt mustbe emphasized that this empirical approach should be used whenever
possible. To scale from the behavior of a single particle in a spinning test-tube to
mmercial scale machine may be a useful scouting technique but
accurate,
liquid throughput (Le.
ugh (m/s)
A more empirical approach may be preferable if data trom e.
technical centrifuges are available. If the performance of a sm:
is known the likely throughput on a large machine may be pred
the ratio of the respective E values for Ihe two:
n Machine
lt mustbe emphasized that this empirical approach should be used whenever
possible. To scale from the behavior of a single particle in a spinning test-tube to
mmercial scale machine may be a useful scouting technique but
accurate,
In this case the principal quantity of importance is, the residence time
sential to give thickening to the required solids content, under the given
centrifugal field, This a of the material and thus the
suspension properties require characterization before accurate s
estimates can be made.
machin:
To reach a given thickness, a minimum residence time t, must be allowed,
For a given radius, this then limits th tra
stream to be
“a3
Where Vis the volume of solids held in the machine, The calculation
oft and V willbe discussed shorty
32.2 Scale-Up Factor
Ideally, a simple scale-up factor analogous to the E value in clarification.
theory is desirable. Such a factor, F, is described be
sential to give thickening to the required solids content, under the given
centrifugal field, This a of the material and thus the
suspension properties require characterization before accurate s
estimates can be made.
machin:
To reach a given thickness, a minimum residence time t, must be allowed,
For a given radius, this then limits th tra
stream to be
“a3
Where Vis the volume of solids held in the machine, The calculation
oft and V willbe discussed shorty
32.2 Scale-Up Factor
Ideally, a simple scale-up factor analogous to the E value in clarification.
theory is desirable. Such a factor, F, is described be
» is the number of gravities pulled by the machine
periphery (= 9.81 wer);
fan is the mean mass fraction of the consolidating slurry in
the machine
Ha is the depth of consolidating slurry
ap is the density difference between solids and liquid.
The time required for a given degree of consolidation will clearly depend
on the concentrating pressure available. It is reasonable to assume that
(15)
The effective capacities of two centrituges can now be directly compared if
their volume and concentrating pressures are known, viz
GU
The R.HSS. of the equation is thus the required scaling factor, F.
32.3 Three Methods of Application
3.2.3.1 Semi-Empirical Approach |
If data on a given material is available on the semi-technical scale,
then the performance of a full-scale centrifuge can be predicted
periphery (= 9.81 wer);
fan is the mean mass fraction of the consolidating slurry in
the machine
Ha is the depth of consolidating slurry
ap is the density difference between solids and liquid.
The time required for a given degree of consolidation will clearly depend
on the concentrating pressure available. It is reasonable to assume that
(15)
The effective capacities of two centrituges can now be directly compared if
their volume and concentrating pressures are known, viz
GU
The R.HSS. of the equation is thus the required scaling factor, F.
32.3 Three Methods of Application
3.2.3.1 Semi-Empirical Approach |
If data on a given material is available on the semi-technical scale,
then the performance of a full-scale centrifuge can be predicted
up data obtained on material A to predict how
material B will behave in the larger centrifuge.
3.2.3.3 Ab Initio Approach
Insight into the compaction characteristics of a material can be
gained quickly from small quantities of feed Wg a strobe
entifuge, where the time required to achieve a given compaction.
under a known concentrating pressure can be found. Knowing the
concentrating pres a larger centrifuge, the
maximum allowable no: arge rate can be readily
calculat
Estimation of Necessary Parameters and the Problems
Encountered
3.2.4.1 Relationship between i and P
The value of y in equation (15) is difficult to establish a priori
Fortunately, the concentrating press.
material B will behave in the larger centrifuge.
3.2.3.3 Ab Initio Approach
Insight into the compaction characteristics of a material can be
gained quickly from small quantities of feed Wg a strobe
entifuge, where the time required to achieve a given compaction.
under a known concentrating pressure can be found. Knowing the
concentrating pres a larger centrifuge, the
maximum allowable no: arge rate can be readily
calculat
Estimation of Necessary Parameters and the Problems
Encountered
3.2.4.1 Relationship between i and P
The value of y in equation (15) is difficult to establish a priori
Fortunately, the concentrating press.
1 total volume which might be filed
by thickened slurry
(including, for example, the volume occluded by th
‘can be evaluated by simple geometry
of the centrifuge. As a first approximati
from the kno\
ion it appe:
take the maximum allowable inventory of thickened slurry as,
half this volume. (In the semi-empirical
sumption will largely cancel but the
potentially more sensitive to the basi
3.2.4.3 Pressure (or “fi
of this factor are based po!
y
‘Ab initio approa
at this part of the calculation.)
n equation (14). For the
entrfuge the quantities to be used are sel evident and are
detailed in the worked exampl
machine some explanation is needed.
later) but for the continuous
Ho, the depth of th idating sediment, is taken as the
distance between the disc stack periphery and the nozzle.
Ing" is some mean of the centrifugal fi
at the machine wall. For the present e»
Id at the disc periphery and
xample we employed an
arithmetic average of the two values in question.
©... nis taken as the solids content where the system starts to
display a perceptible cohesi
solids in the case of flocculated "Prute
because thickening to the so-
en"
by thickened slurry
(including, for example, the volume occluded by th
‘can be evaluated by simple geometry
of the centrifuge. As a first approximati
from the kno\
ion it appe:
take the maximum allowable inventory of thickened slurry as,
half this volume. (In the semi-empirical
sumption will largely cancel but the
potentially more sensitive to the basi
3.2.4.3 Pressure (or “fi
of this factor are based po!
y
‘Ab initio approa
at this part of the calculation.)
n equation (14). For the
entrfuge the quantities to be used are sel evident and are
detailed in the worked exampl
machine some explanation is needed.
later) but for the continuous
Ho, the depth of th idating sediment, is taken as the
distance between the disc stack periphery and the nozzle.
Ing" is some mean of the centrifugal fi
at the machine wall. For the present e»
Id at the disc periphery and
xample we employed an
arithmetic average of the two values in question.
©... nis taken as the solids content where the system starts to
display a perceptible cohesi
solids in the case of flocculated "Prute
because thickening to the so-
en"
speed centrifuge (see Si ofthis manual or
Appendix A, reference [19].
The calculated ultima allows one to rapidly decide
‘whether a desired concentration can sen ieved by
centrifugation.
3.3Scale-Up When Hindered is the Limiting Process
33.1 Background
The terminal velocity of a swarm of particles is generally less than that of a
single parte falling in an infinitely wide pond; the more crowded the
particles, the slower they fall. This is called “hindered setting
The £ theory (Section 3.4.3(a)) implicitly deals with the initial captur
isolated particles from virtually clear supernatant. For capture to be
completed, however, these particles need to be consolidated into a film on
the underside of the discs so that they can subsequently migrate towards
the bow! periphery. I, as a result of hindered setting, a “traffic-jam
prevents this consolidation, the disc-space wil fll up with solids and lead
to carry-over in the centrate.
‘As with the derivation of, the problem of hindered setting will first be
in terme thickener (cf Section 3.3.4). This will
Appendix A, reference [19].
The calculated ultima allows one to rapidly decide
‘whether a desired concentration can sen ieved by
centrifugation.
3.3Scale-Up When Hindered is the Limiting Process
33.1 Background
The terminal velocity of a swarm of particles is generally less than that of a
single parte falling in an infinitely wide pond; the more crowded the
particles, the slower they fall. This is called “hindered setting
The £ theory (Section 3.4.3(a)) implicitly deals with the initial captur
isolated particles from virtually clear supernatant. For capture to be
completed, however, these particles need to be consolidated into a film on
the underside of the discs so that they can subsequently migrate towards
the bow! periphery. I, as a result of hindered setting, a “traffic-jam
prevents this consolidation, the disc-space wil fll up with solids and lead
to carry-over in the centrate.
‘As with the derivation of, the problem of hindered setting will first be
in terme thickener (cf Section 3.3.4). This will
33.3 Batch Thickeners
Figure 14
dered sett
Figure 14
dered sett
Figure 15
There will often be a minimum in his cur urring at
concentration G* in the above plot. If, as is frequently Ihe case, the
feed concentration is lower than G* and the sludge concentration
higher, then G* represents the maximum solids flux that the
thickener can handle.
H a greater solids loading is imposed, then the thickener can n
longer cope. It wil fl up with material of concentration G* and
3s solids will be lost to the previously clear overt
33.4 Gravity Thickeners - Interpretation of Batch Settling Test
In a batch gravity test, the sediment Interface height above the ves
There will often be a minimum in his cur urring at
concentration G* in the above plot. If, as is frequently Ihe case, the
feed concentration is lower than G* and the sludge concentration
higher, then G* represents the maximum solids flux that the
thickener can handle.
H a greater solids loading is imposed, then the thickener can n
longer cope. It wil fl up with material of concentration G* and
3s solids will be lost to the previously clear overt
33.4 Gravity Thickeners - Interpretation of Batch Settling Test
In a batch gravity test, the sediment Interface height above the ves
Interface Height
time t
Figure 16
script o refers to the inital conditions)
hioka Method can now be used to determine the limiting flux rate.
This method is detailed in reference [10] and only the mechanistic step
are shown below
From the plotof Interface height against time
tling velocity u = dh/dt
Solids flux ”
And hence a plot © can
of hindered sertli constant and h
constant and this rises with concentration
A more time consuming, but potentially more accurate, method would be
to make up feeds of varying concentration and measure the initial setting
time t
Figure 16
script o refers to the inital conditions)
hioka Method can now be used to determine the limiting flux rate.
This method is detailed in reference [10] and only the mechanistic step
are shown below
From the plotof Interface height against time
tling velocity u = dh/dt
Solids flux ”
And hence a plot © can
of hindered sertli constant and h
constant and this rises with concentration
A more time consuming, but potentially more accurate, method would be
to make up feeds of varying concentration and measure the initial setting
Figure 17
The intercept yu o used to culculate the minimum settling area
ectsamry do achieve the desired discharge. calculat
where D is the thickened solids volumetric discharge ra
33.5 Application to Centrifuges
33.5.1 Batch Flux Curve
The batch setting test will typically be performed in a strot
centrifuge. The equivalent gravity test can be obtained by
expanding the time scale proportionately to the mean g-force
prevailing,
The preceding analysis will then give the minimum allowable flow
The intercept yu o used to culculate the minimum settling area
ectsamry do achieve the desired discharge. calculat
where D is the thickened solids volumetric discharge ra
33.5 Application to Centrifuges
33.5.1 Batch Flux Curve
The batch setting test will typically be performed in a strot
centrifuge. The equivalent gravity test can be obtained by
expanding the time scale proportionately to the mean g-force
prevailing,
The preceding analysis will then give the minimum allowable flow
(2) is not clear what concentration to choose for the underflow” In the
above graphical construction, It is probably ML the nozzle discharge
centration but this should give a ve result
(3) The identification of points a and b is generally far from straightforward
given the quality of data experienced in the real world. Point bis the most
important to determine as the anal
it may be most appropriately determined experimentally as the
centration to which particle te under gravity in the
re. An alternative approach tothe problem
However, the E ratio betw
relative capacities for hindered setting limitation. T a great
relief itis a simple result from a complex situation.
4 ASYSTEMATIC APPROACH TOCENTRIFUGE SCALE-UP
The following steps are recommended:
(1) Unless there is strong supporting evidence one way or the other, It is
dangerous to jump to a conclusion regarding the limiting mechanism In
trfuge operation.
above graphical construction, It is probably ML the nozzle discharge
centration but this should give a ve result
(3) The identification of points a and b is generally far from straightforward
given the quality of data experienced in the real world. Point bis the most
important to determine as the anal
it may be most appropriately determined experimentally as the
centration to which particle te under gravity in the
re. An alternative approach tothe problem
However, the E ratio betw
relative capacities for hindered setting limitation. T a great
relief itis a simple result from a complex situation.
4 ASYSTEMATIC APPROACH TOCENTRIFUGE SCALE-UP
The following steps are recommended:
(1) Unless there is strong supporting evidence one way or the other, It is
dangerous to jump to a conclusion regarding the limiting mechanism In
trfuge operation.
le-up methods to predict a f a given
type will be needed to just cope with the envisaged duty
(6) lt must be remembered tl
from their theoretical maximum to ensure centrate remains clear
jte minor process fluctuations. A 20% installed spare
(7) Further allowances for longer term deviations from ideal
nditions may be appropriate.
(8) CIP sequences (e.g. up to 1 hour per shift) need to be accommodated.
some spare capacity to cope with the inevitable mechanical
1ould be considered.
me worked examples.
Validity of Different Methods
Certain of the procedures - e.g. the Ab initio approach to calculation of
centrifuge performanc
type will be needed to just cope with the envisaged duty
(6) lt must be remembered tl
from their theoretical maximum to ensure centrate remains clear
jte minor process fluctuations. A 20% installed spare
(7) Further allowances for longer term deviations from ideal
nditions may be appropriate.
(8) CIP sequences (e.g. up to 1 hour per shift) need to be accommodated.
some spare capacity to cope with the inevitable mechanical
1ould be considered.
me worked examples.
Validity of Different Methods
Certain of the procedures - e.g. the Ab initio approach to calculation of
centrifuge performanc
In either case, the E or Ki
ion of the tials may
ss provide valuable insight. We suggest that all scale-up
s be tested and quantified as far as possible, but that
ntuition are stil important when working out the
implications. The need for a representative number of trials before
imiting large apital cannot be overstressed.
cedure is displayed diagrammatically in Figures
18 and 19 whilst details on how to perform individual scale-up metho:
are provided in Tables 1-7.
ion of the tials may
ss provide valuable insight. We suggest that all scale-up
s be tested and quantified as far as possible, but that
ntuition are stil important when working out the
implications. The need for a representative number of trials before
imiting large apital cannot be overstressed.
cedure is displayed diagrammatically in Figures
18 and 19 whilst details on how to perform individual scale-up metho:
are provided in Tables 1-7.
Calculate the free fall velocity under a single gravity (equation (8).
Caleulate the C theory for the centrifuge under consideration (equation
(©) using manufacture
Calculate the maximum centrate rate per machine (equation (10))
Calculate the maximum operational feed rate by simple mass balanc
given the di
Compare with other scale-up methods
Table 2
ical Scale-Up for Clarification Limited Systems
(see Section 3.4.3(a)
Calculate I values for the semi-tech and production centrifuges (equation
(6).
‘Scale centrate rates achievable on semi-tech centrifuge using equation
(11) to give maximum centrate rate possible on the production centrifuge.
Check semi-tech behavior as a function of number of discs or feed
concentration, le establishes whether clarification is the limiting factor.
Caleulate the C theory for the centrifuge under consideration (equation
(©) using manufacture
Calculate the maximum centrate rate per machine (equation (10))
Calculate the maximum operational feed rate by simple mass balanc
given the di
Compare with other scale-up methods
Table 2
ical Scale-Up for Clarification Limited Systems
(see Section 3.4.3(a)
Calculate I values for the semi-tech and production centrifuges (equation
(6).
‘Scale centrate rates achievable on semi-tech centrifuge using equation
(11) to give maximum centrate rate possible on the production centrifuge.
Check semi-tech behavior as a function of number of discs or feed
concentration, le establishes whether clarification is the limiting factor.
Calculate the internal pressure required to give the desired solids
nitration (reference [19],
Calculate the maximum internal pressure that can be generated by the
entrfuge being considered (equatio
uficiont for the required duty. If
ntrituge or modify the feed material. itis,
check the clarification and compaction kine!
Table 4
Steps in Semi-Empirical Scale-Up for Solids Limited Systems
(see Section 3.4.3(b)
Determine breakthrough solids content as a function of nozzle rate for a
small-scale centrifuge, of comparable design to the subject (possible
production) machine,
Estimate relative maximum allowable Inventories of thickened slurry for (a)
nal -scale machine; (b) subject centrifuge (see text for detal
Evaluate relative values of P for (a) small-scale machine; (b) subject
Centrifuge, from their known speed, dimensions and solids loadin
From Il and Ill estimate the scale-up factor between the centrifuges as
nitration (reference [19],
Calculate the maximum internal pressure that can be generated by the
entrfuge being considered (equatio
uficiont for the required duty. If
ntrituge or modify the feed material. itis,
check the clarification and compaction kine!
Table 4
Steps in Semi-Empirical Scale-Up for Solids Limited Systems
(see Section 3.4.3(b)
Determine breakthrough solids content as a function of nozzle rate for a
small-scale centrifuge, of comparable design to the subject (possible
production) machine,
Estimate relative maximum allowable Inventories of thickened slurry for (a)
nal -scale machine; (b) subject centrifuge (see text for detal
Evaluate relative values of P for (a) small-scale machine; (b) subject
Centrifuge, from their known speed, dimensions and solids loadin
From Il and Ill estimate the scale-up factor between the centrifuges as
tling curve(s) in strobe centrifuge
From | evaluate ¢, ie. the residence time, to give required solids content In
strobe centrifuge
Evaluate P (the consolidating pressure) in strobe centrifuge from its known
speed, dimensions, and solids loading (reference [19])
Evaluate P in subject centrifuge from its known speed, dimensions and
olids loading,
Using equation (15), evaluate required residence time for subject
centrifuge (assuming y = 1)
From centrifuge dimensions evaluate an approximation to maximum
achievable Inventory at thickened slurry in subject machine (see text for
details).
Estimate maximum allowable nozzle rate, to give required thickening in
‘equation (13)
ulate the maximum feed-rate from simple mass balance.
(Compare with other scale-up methods.
From | evaluate ¢, ie. the residence time, to give required solids content In
strobe centrifuge
Evaluate P (the consolidating pressure) in strobe centrifuge from its known
speed, dimensions, and solids loading (reference [19])
Evaluate P in subject centrifuge from its known speed, dimensions and
olids loading,
Using equation (15), evaluate required residence time for subject
centrifuge (assuming y = 1)
From centrifuge dimensions evaluate an approximation to maximum
achievable Inventory at thickened slurry in subject machine (see text for
details).
Estimate maximum allowable nozzle rate, to give required thickening in
‘equation (13)
ulate the maximum feed-rate from simple mass balance.
(Compare with other scale-up methods.
rate achievable on the semi-technical centrifuge (for desired
ration) to give maximum probable rate on production
hine by multiplying flux by & production | semi-tech.
Calculate the: onding maximum feed rates.
Compare results with those determined by other techniques.
#NB Ab Initio scale-up from strobe centrifuge data should be treated with caution
but such a calculation is needed as a guide it should follow the lines of Table 7.
Table 7
Steps in Ab Initio Guide Calculations for Systems Limited
by Hindered Settling Systems’
(see Section 3.4.3(c)
Determine batch setting flux for a known centrifugal field using strobe
centrifuge and the calculation procedure detailed in Section 3.4.3(¢).
Convert above results to their equivalent for a single gravity settler
Calculate the E value for the centrifuge under consideration (equation (6))
using manufacturers specifications.
ration) to give maximum probable rate on production
hine by multiplying flux by & production | semi-tech.
Calculate the: onding maximum feed rates.
Compare results with those determined by other techniques.
#NB Ab Initio scale-up from strobe centrifuge data should be treated with caution
but such a calculation is needed as a guide it should follow the lines of Table 7.
Table 7
Steps in Ab Initio Guide Calculations for Systems Limited
by Hindered Settling Systems’
(see Section 3.4.3(c)
Determine batch setting flux for a known centrifugal field using strobe
centrifuge and the calculation procedure detailed in Section 3.4.3(¢).
Convert above results to their equivalent for a single gravity settler
Calculate the E value for the centrifuge under consideration (equation (6))
using manufacturers specifications.
(1) À European Prut parates biomass from fermenter liquor (3
lating the cells, concentrating the
subsequently centrifuging the cells to 17 wt %.
s a diferent floc
Flocs which are more amenable to centrifugation, The flotation
ould be omitted.
many centrifuges would be required for
a full-scale plant using the n ss, and to quantity the possible
benefits of using a larger but untested centrifuge.
The existing plant has three Alfa-Lavat FEUX 320 disc-no irfugos
each capable of producing up to 2.9 te/h (dry basis> of 17 wt % material
(see Figure 20).
‘Semi-technical trials on a Westfalia NA7 machine (see Table 8) were
Carried out on both the existing and the proposed new flocculation routes,
called Pland P2 respectivel
A possible alternative to the FEUX 320 machines was the new Westfalia
HDA 300 disc-nozzle centrifuge, a more expensive but larger machine.
(8) Experience has shown that increasing the number of discs in the FEUX
320 machines had no effect on instantaneous capacity, i.e. the process ls
probably solids limited. (Indeed, narrow detrimental due
lating the cells, concentrating the
subsequently centrifuging the cells to 17 wt %.
s a diferent floc
Flocs which are more amenable to centrifugation, The flotation
ould be omitted.
many centrifuges would be required for
a full-scale plant using the n ss, and to quantity the possible
benefits of using a larger but untested centrifuge.
The existing plant has three Alfa-Lavat FEUX 320 disc-no irfugos
each capable of producing up to 2.9 te/h (dry basis> of 17 wt % material
(see Figure 20).
‘Semi-technical trials on a Westfalia NA7 machine (see Table 8) were
Carried out on both the existing and the proposed new flocculation routes,
called Pland P2 respectivel
A possible alternative to the FEUX 320 machines was the new Westfalia
HDA 300 disc-nozzle centrifuge, a more expensive but larger machine.
(8) Experience has shown that increasing the number of discs in the FEUX
320 machines had no effect on instantaneous capacity, i.e. the process ls
probably solids limited. (Indeed, narrow detrimental due
Figure
Estimated limiting solids fluxes (dev
basis) for Pl material thickened in a
Saleulated using only she "volume
factor and is hence expected to proxide
jer bound te possible. output, EM
Upper curve mas estimated using both
Van and #100" terme
Estimated limiting solids fluxes (dev
basis) for Pl material thickened in a
Saleulated using only she "volume
factor and is hence expected to proxide
jer bound te possible. output, EM
Upper curve mas estimated using both
Van and #100" terme
ant throughput required = 11.4 te/h (db). On the old P material, the
can produc lb) of 17% material
Jo of FEUX
machines req! d
(PD and Over (P2) are the maximum
ached for a given thicks concentration on th
hen running the 347.
9 machines operat
Habs
Generally, plants operate centrifuges away from breakthrough conditions,
and so in practice 4-5 machines will be required to be online at any one
time. Regular CIP sequences Cup to 1 hour per shift) and breakdowns
means that 6-7 machines should be installed,
Tr ith the estimate obtained from the a prior
approach (see later» and is In fair accord with the (supposed) slightly
more refined semi-empirical (1) technique:
522 Standard Semi-Empirical Approach
can produc lb) of 17% material
Jo of FEUX
machines req! d
(PD and Over (P2) are the maximum
ached for a given thicks concentration on th
hen running the 347.
9 machines operat
Habs
Generally, plants operate centrifuges away from breakthrough conditions,
and so in practice 4-5 machines will be required to be online at any one
time. Regular CIP sequences Cup to 1 hour per shift) and breakdowns
means that 6-7 machines should be installed,
Tr ith the estimate obtained from the a prior
approach (see later» and is In fair accord with the (supposed) slightly
more refined semi-empirical (1) technique:
522 Standard Semi-Empirical Approach
alia MP
Estimated Limiting
determined by
solida fluxes
he semi-sepirical
procedure, for P2 materiel thickened
In a PEUX 320, The key de as for
Figure 20. The crosses are expected
2 Solids Content 1X wi
determined by
solida fluxes
he semi-sepirical
procedure, for P2 materiel thickened
In a PEUX 320, The key de as for
Figure 20. The crosses are expected
2 Solids Content 1X wi
1000 rpm exp
oT
at 500 and 1000 rpm in
oT
at 500 and 1000 rpm in
Volume” term. However, no further elaboration of the basis of the
calcule justified at present. This require inthe
fundamental science underlying the problem.
rization data are available, the
rd with experien:
‘and the more
discussed, was also satisfactory
The approach has also been applied to prediction of the perform
a Westfalia HDA 300 centrifug eto that of an Alfa-Lavat
Pruleen” feed cream concentration. The data provided in Table
uggests a scale factor of circ
Results from actual operation of such a machine Indicate a scale-up of the
order of 1.4 is observed, provided a close-spaoed disc stack (* 1 mm) is
mployed. Unfortunately this rather Agreement is somewhat
ered by the further observation that with a 2.5 mm stack the relative
performance of the HDA 300 fel to only ~ 1.15 of that of the FEUX 320. At
this stage the reasons for this effect are not clear. However, itis perhaps.
most likely that due to particular design features comparatively more discs
are needed in the HDA 300 to achies tory particle acceleration
and a reasonable approach to the theoretical limit to performance given by
the calculation: Also, the machine ¡sr e | ai ve! y taller and so
maldistibution may be a problem: the extra discs m la
calcule justified at present. This require inthe
fundamental science underlying the problem.
rization data are available, the
rd with experien:
‘and the more
discussed, was also satisfactory
The approach has also been applied to prediction of the perform
a Westfalia HDA 300 centrifug eto that of an Alfa-Lavat
Pruleen” feed cream concentration. The data provided in Table
uggests a scale factor of circ
Results from actual operation of such a machine Indicate a scale-up of the
order of 1.4 is observed, provided a close-spaoed disc stack (* 1 mm) is
mployed. Unfortunately this rather Agreement is somewhat
ered by the further observation that with a 2.5 mm stack the relative
performance of the HDA 300 fel to only ~ 1.15 of that of the FEUX 320. At
this stage the reasons for this effect are not clear. However, itis perhaps.
most likely that due to particular design features comparatively more discs
are needed in the HDA 300 to achies tory particle acceleration
and a reasonable approach to the theoretical limit to performance given by
the calculation: Also, the machine ¡sr e | ai ve! y taller and so
maldistibution may be a problem: the extra discs m la
observed to tal ‘minutes
Using equation (15), Section 3.4,3(b), with y
required residence tim:
degrees of thickening :20 machine which can deliver
48 times the concentrating pressure of the stroboscopic instrument.
Taking the maximum allowable inventory of thickened slury in the
ipment to be ca 1/2 of machine internal volume predicts limiting
fluxes of 4.0 and 3.3 tonnes / hour of solids. This is in good
agreement with observed performance.
NB key approximation used, le required time is Inversely
proportional to applied pressure, is certainly not a good one for
‘consolidating sediments close to their limiting solids content but,
empirically, It has been found to hold reasonably well for a variety
of flocculated systems in the earlier stages of concentration
53 Calculation of Thickening Ultimate Limit
Example: "Structural" limit for Thickening in “Pruteen” Centrifugation
As previously noted, Itis useful to be able to estimate the ‘structural limit on
thickening to be expected under a particular centrifuge regime, even though
practical constraints over available residence times may prevent such solid
Contents being approached closely. In particular, such calculations are likely to
be valuable in Identifying when the kinetics of consolidation of cohesive sediment
Using equation (15), Section 3.4,3(b), with y
required residence tim:
degrees of thickening :20 machine which can deliver
48 times the concentrating pressure of the stroboscopic instrument.
Taking the maximum allowable inventory of thickened slury in the
ipment to be ca 1/2 of machine internal volume predicts limiting
fluxes of 4.0 and 3.3 tonnes / hour of solids. This is in good
agreement with observed performance.
NB key approximation used, le required time is Inversely
proportional to applied pressure, is certainly not a good one for
‘consolidating sediments close to their limiting solids content but,
empirically, It has been found to hold reasonably well for a variety
of flocculated systems in the earlier stages of concentration
53 Calculation of Thickening Ultimate Limit
Example: "Structural" limit for Thickening in “Pruteen” Centrifugation
As previously noted, Itis useful to be able to estimate the ‘structural limit on
thickening to be expected under a particular centrifuge regime, even though
practical constraints over available residence times may prevent such solid
Contents being approached closely. In particular, such calculations are likely to
be valuable in Identifying when the kinetics of consolidation of cohesive sediment
determined using the Triton VRC Frozen Image Centrifuge. Examples for the P2
material are shown in Figure 22. On taking the unknown cell density as ca 1.07
g/cmi agreement between prediction and observed limiting, strobe centrifuge
lids contents were satisfactory (Table 9). (Equation (4)) Appendix A o
reference [19] was used to calculate values for the pressure applied by the
centrifuge using known values for speed, rotor diameter and so on.)
lids content curves, and estimating Ihe pressure
320 centrifuges from various geomet
id never thicken the materials in
(Table 9). These proved to be w nd the values
ight) to which concentration was required and hence it at
engih of centrifugal field per se would not be a problem. It should be noted
that extrapolation of actual centrifuge data (eg Figure (Le. infinite
residence time in the field) gave ultimate limits of the order of those calculated
(Table
material are shown in Figure 22. On taking the unknown cell density as ca 1.07
g/cmi agreement between prediction and observed limiting, strobe centrifuge
lids contents were satisfactory (Table 9). (Equation (4)) Appendix A o
reference [19] was used to calculate values for the pressure applied by the
centrifuge using known values for speed, rotor diameter and so on.)
lids content curves, and estimating Ihe pressure
320 centrifuges from various geomet
id never thicken the materials in
(Table 9). These proved to be w nd the values
ight) to which concentration was required and hence it at
engih of centrifugal field per se would not be a problem. It should be noted
that extrapolation of actual centrifuge data (eg Figure (Le. infinite
residence time in the field) gave ultimate limits of the order of those calculated
(Table
t ream (PL route)
Pruteen* Works) in strobo
fentrifuge running at 1000 rpm
pH 3.5, 9
scopic centrifuge running
00 rpm
H 3.5, 90°C heat shock> in
stroboscopic centrifuge run
at 500 pe
Pruteen" Works centrifuge fa
cream in Westfalia BA7
Infinite residence time
As above but in Alfa-Laval FEUX
P2 material
«pk 3.5, 90
« above but in Alfa-Laval FEUX 320
Pruteen* Works) in strobo
fentrifuge running at 1000 rpm
pH 3.5, 9
scopic centrifuge running
00 rpm
H 3.5, 90°C heat shock> in
stroboscopic centrifuge run
at 500 pe
Pruteen" Works centrifuge fa
cream in Westfalia BA7
Infinite residence time
As above but in Alfa-Laval FEUX
P2 material
«pk 3.5, 90
« above but in Alfa-Laval FEUX 320
iously, the second Involving recovery of I nism,
ligenes Eutrophus:
5.4.1 Ab Initio Clarification Calculation for P2 Type Single Cell
Protein
The initial setting rate for the P2 material, for a particular "9" was given by
the tangent at the origin to the settling curve, Figure 23. This was equal to
0.0435 em/s at 180 9.
2.42 x 10
308 x 104 m
quatiom (10)
ntrate rate 2.42 x 10% x 6.306 x 10° 3"
0.153 m/s Pa
of 4.61% % thicks, 1 te feed gives
73 te of centrate
hus centrifuge feed rate
Hence maximum solids throughput x 755 te/h (dry solids
basis:
ligenes Eutrophus:
5.4.1 Ab Initio Clarification Calculation for P2 Type Single Cell
Protein
The initial setting rate for the P2 material, for a particular "9" was given by
the tangent at the origin to the settling curve, Figure 23. This was equal to
0.0435 em/s at 180 9.
2.42 x 10
308 x 104 m
quatiom (10)
ntrate rate 2.42 x 10% x 6.306 x 10° 3"
0.153 m/s Pa
of 4.61% % thicks, 1 te feed gives
73 te of centrate
hus centrifuge feed rate
Hence maximum solids throughput x 755 te/h (dry solids
basis:
us time for settling of P2 material
trobe centrifuge at 1000 rpm (= 180 8)
initial suspended solids content wi
al solids of approximately 4
trobe centrifuge at 1000 rpm (= 180 8)
initial suspended solids content wi
al solids of approximately 4
Table 10
Calculation of E and KO scale factors for Westfalia NA7 and NDA 300 centrifuges,
and for Alfa-Laval BPR 2075 and FEUX 320 machines
(NB = 2 wh cotO Nr? = ko = 2 coto
3 > 35
Speed o
Machine | (rpm - values | (cot © in No of discs (N ” xa
in rads/sec in | parenthese: (metres) | (mt) | (MKS units)
parentheses)
Westfalia 7500 62 0.036 3.316 x 10%
NAT (785.4) 0.5 mm spa 3
118 xl a 7.43 x 10?
Alfa-Laval | 6150 mm spacing)
BRPX 207s (644.0) 38 2.39 x 102
mm spacing)
Alfa-Laval 3600 a 60 . 4.254 x 103
FEUX 320 (377.0) (1.192) mm spacing)
Westfalia 3000 280 1.867 x 10%
HDA 300 (314.2) (1.881)
Calculation of E and KO scale factors for Westfalia NA7 and NDA 300 centrifuges,
and for Alfa-Laval BPR 2075 and FEUX 320 machines
(NB = 2 wh cotO Nr? = ko = 2 coto
3 > 35
Speed o
Machine | (rpm - values | (cot © in No of discs (N ” xa
in rads/sec in | parenthese: (metres) | (mt) | (MKS units)
parentheses)
Westfalia 7500 62 0.036 3.316 x 10%
NAT (785.4) 0.5 mm spa 3
118 xl a 7.43 x 10?
Alfa-Laval | 6150 mm spacing)
BRPX 207s (644.0) 38 2.39 x 102
mm spacing)
Alfa-Laval 3600 a 60 . 4.254 x 103
FEUX 320 (377.0) (1.192) mm spacing)
Westfalia 3000 280 1.867 x 10%
HDA 300 (314.2) (1.881)
al polymer poly-3-hydroxybutyrate (PHB) the
umulated by the bacteria as litle intracellular granules and the
fermentation is continued unti at k of the suspended dry matter:
polymer. Since the normal solid density
density of these bacteria at the end of the accumulation is probably
abnormally high ("1.10 g/cm31 for this type of system.
pic centrifuge measurements using the Triton WRC machine at
180 9 gave 0.3 om secimentation ofa clear upper nerace in 30
minutes#. This is equivalent to a gravitational free sedimentation
velocity of 3.3 x 10° mihr at 9.2 x 10” cm/set, which can be combined
the appropriate E: value from Table 10 (10.9 x 10° m‘ to give 360 lter/hr
for the maximum calculated centrate rate for total recovery on the BRPX
207 centrifuge. For this particular example, this would also be the
‘approximate limiting feed rate as feed concentrations would be very low
(see Figure 25).
Figure 24 shows the proportional recovery
function of feed rate on the Alfa Lavat BRPX 270s disc nozzle centrifuge.
The machine was set up with 118 discs at 0.5mm spacing and three 0.45
mm nozzles (probably the minimum number that can be safely used due
to risk of blockage etc). The recovery curve clearly indicates that feed
rates as low as - 500 Iter/hr are needed before -100% retention of solk
achieved. Clearly there is remarkably good agreement between
observation and the performance predicted from calculation, particula
when one considers the difficulty in obtaining a reliable value for the
sedimentation velocity.
umulated by the bacteria as litle intracellular granules and the
fermentation is continued unti at k of the suspended dry matter:
polymer. Since the normal solid density
density of these bacteria at the end of the accumulation is probably
abnormally high ("1.10 g/cm31 for this type of system.
pic centrifuge measurements using the Triton WRC machine at
180 9 gave 0.3 om secimentation ofa clear upper nerace in 30
minutes#. This is equivalent to a gravitational free sedimentation
velocity of 3.3 x 10° mihr at 9.2 x 10” cm/set, which can be combined
the appropriate E: value from Table 10 (10.9 x 10° m‘ to give 360 lter/hr
for the maximum calculated centrate rate for total recovery on the BRPX
207 centrifuge. For this particular example, this would also be the
‘approximate limiting feed rate as feed concentrations would be very low
(see Figure 25).
Figure 24 shows the proportional recovery
function of feed rate on the Alfa Lavat BRPX 270s disc nozzle centrifuge.
The machine was set up with 118 discs at 0.5mm spacing and three 0.45
mm nozzles (probably the minimum number that can be safely used due
to risk of blockage etc). The recovery curve clearly indicates that feed
rates as low as - 500 Iter/hr are needed before -100% retention of solk
achieved. Clearly there is remarkably good agreement between
observation and the performance predicted from calculation, particula
when one considers the difficulty in obtaining a reliable value for the
sedimentation velocity.
|
|
|
| Input Con
= Feed Rate (17h
|
|
| Input Con
= Feed Rate (17h
ram
2 FEUD
ce FEUX Capacity
BAT Capacity
Note that "capacity" here refers to centrate volume
Note also that use of KQ (rather than 2) is equally valid
comparing two centrifuges for clarfying duties and would
capacity ratio of 12.8.
Sample Calculation:
NA7 can handle 1.5 te/h of 10% feed when concentrating up to
13%.
What could the FEUX 320 handle?
Centrate rate from KAT
2 FEUD
ce FEUX Capacity
BAT Capacity
Note that "capacity" here refers to centrate volume
Note also that use of KQ (rather than 2) is equally valid
comparing two centrifuges for clarfying duties and would
capacity ratio of 12.8.
Sample Calculation:
NA7 can handle 1.5 te/h of 10% feed when concentrating up to
13%.
What could the FEUX 320 handle?
Centrate rate from KAT
Te. the cenirfuge
capacity of 2.9 te)
Then 4.3 te rate implie
handle 5.6 x 0.06
This demonstrates the great importance of feed conc
clarification limited duties, even though itis often irreleva
solids limitations
lt should be noted that th ulations under -predict actual
performance. This has litle significance as the calculation method
is invalid for solids limited applications - one cannot scale-up a
lids limited breakthrough curve using clarification limited theory.
We would re-emphasize that this semi-empirical approach does not
give a bound (either upper or lower) to performance.
55 Ab Initio Calculation for Hindered Settling Limitation
Finally it s ngle cell protein example to illustrate use
of the method for calculating hindered setting imitation:
Rates of setting versus solids content could be found by drawing tangents to the
curve in Figur batch settling curve vas party
capacity of 2.9 te)
Then 4.3 te rate implie
handle 5.6 x 0.06
This demonstrates the great importance of feed conc
clarification limited duties, even though itis often irreleva
solids limitations
lt should be noted that th ulations under -predict actual
performance. This has litle significance as the calculation method
is invalid for solids limited applications - one cannot scale-up a
lids limited breakthrough curve using clarification limited theory.
We would re-emphasize that this semi-empirical approach does not
give a bound (either upper or lower) to performance.
55 Ab Initio Calculation for Hindered Settling Limitation
Finally it s ngle cell protein example to illustrate use
of the method for calculating hindered setting imitation:
Rates of setting versus solids content could be found by drawing tangents to the
curve in Figur batch settling curve vas party
Batch
ng curve, from 1
strobe
»mtrifage data (Figure 23
for P2 material
5 Flux x 10?
ng curve, from 1
strobe
»mtrifage data (Figure 23
for P2 material
5 Flux x 10?
Fig 27 Definition of Terms for Settling Rate Calculation
However, at ow solids contents the rate of settling does not vary and is not, as
assumed in the hindered setting theory, a function of concentration. Thus below
the critical concentration where the falling rate zone commenced Yo was taken
as C x setting velocity, where the latter was, of course, a constant. #
Row for thickening to 17% by weight of total solids (= 15.5 wt. % of suspended
solids) the intercept Y m = 0.295 x 10° g/em'/sec, if we chose Co = 15.5%. Note
that this is for 180 g and thus, scaling already, Ynin (1 g) = 0.131 x 10° gicm?isec.
‘Thus Yen = 8.45 x 10 m/sec,
Yan
where D 18 the thickened solids volumetric discharge rate
However, at ow solids contents the rate of settling does not vary and is not, as
assumed in the hindered setting theory, a function of concentration. Thus below
the critical concentration where the falling rate zone commenced Yo was taken
as C x setting velocity, where the latter was, of course, a constant. #
Row for thickening to 17% by weight of total solids (= 15.5 wt. % of suspended
solids) the intercept Y m = 0.295 x 10° g/em'/sec, if we chose Co = 15.5%. Note
that this is for 180 g and thus, scaling already, Ynin (1 g) = 0.131 x 10° gicm?isec.
‘Thus Yen = 8.45 x 10 m/sec,
Yan
where D 18 the thickened solids volumetric discharge rate
settling area for the
lids throughput of the contrifuge:=
56 Comments
The above calculation apparently yields about the correct magnitude for the
centrifuge capacity (see previous examples). Although all data suggest that the
machine is solids-limited it must be remembered that the regions of hindered
settling and compression very much overlap.
Idealized separation into different regimes is generally only seen in textbook
Examination ofthe setting curves in the form of solids content versus time back
this up (Figure 22). We would remind the reader, too, that itis rarely practical to
operate a centrifuge deep In the compression zone where con:
lids throughput of the contrifuge:=
56 Comments
The above calculation apparently yields about the correct magnitude for the
centrifuge capacity (see previous examples). Although all data suggest that the
machine is solids-limited it must be remembered that the regions of hindered
settling and compression very much overlap.
Idealized separation into different regimes is generally only seen in textbook
Examination ofthe setting curves in the form of solids content versus time back
this up (Figure 22). We would remind the reader, too, that itis rarely practical to
operate a centrifuge deep In the compression zone where con:
F A Records, “Filtration Theory in Cyclic Centrifuges’, Chem
Process Eng, (November 1972)
CC Donaldson and R F Stewart, "The Thickening of Floceulated
“Pruteen'-Orgenién Suspensions by Centrifugation", Report IC
91985 (1984)
EF Stevart and D Sutton, in "Solid-Liquid Separation"
Ed J Gregory), ppi1i-126 (Ellis Harwood, Chichester), (198
Y Aster, CC Donaldson, J MeMahon and R F Stewart, "The
Control of Floc Structure and Strength XII An Investigation
of the Structure and Mechanical Properties of Flocculated
"Pruteen'-Organisn Suspensions". Report CL/8/01/1736/8.
P D A Mills, D Watkinson and G E Yates, "A Proliminary
Experimental Study of the Kinetics of Sedimentation of
Flocculated Fino Suspensions”, Report to be issued (1985)
D Macdonald, Chen Engineer, 15 March 1985)
A Rushton and M Spear. | Eiltzationnd Senaratione,ay/J)
1975», 254; A Rushton, Erocess Eng, (September 1900), 49.
DF Moir and H J Pearse, "The Use and Selection of
Flocculante", (Harwell = Warren Spring State of the Art Report
SARI6 - available as ¥D/21088), 19%
Dunlop, "Floceulation o£ ASL by the “Prvteen" Partial
Lysis) Route Part I", Agricultural Division Report, 421,10
Gost
P Chokrabeti, "Industrial Solid/Liquid Separation - A Review
and Proposals for Research", Report CL-R/28/1986/4 (1978)
A Harrison, "Seale-Up of Centrifuges for Biochemical Pr
Part A: Sealing up the "Pruteen' Process. Agricultural
Division, Planning & Coordination Department Report. 192
September, 1983
D J Tagg, "Optimisation of Disc Stack Spacing", Agricultural
Division Research Minute, 20th July 1981
x 5 Rogers. "Protein 1 Centrifuges, Conclusions and
Recommendation’, Projects and Engineering Department Niaut
20th March 1979.
B Fitch, “Current Theory and Thickener Design (1975). Parts
1,2 and 9", Eilteation and Separation, July-December 1975
Process Eng, (November 1972)
CC Donaldson and R F Stewart, "The Thickening of Floceulated
“Pruteen'-Orgenién Suspensions by Centrifugation", Report IC
91985 (1984)
EF Stevart and D Sutton, in "Solid-Liquid Separation"
Ed J Gregory), ppi1i-126 (Ellis Harwood, Chichester), (198
Y Aster, CC Donaldson, J MeMahon and R F Stewart, "The
Control of Floc Structure and Strength XII An Investigation
of the Structure and Mechanical Properties of Flocculated
"Pruteen'-Organisn Suspensions". Report CL/8/01/1736/8.
P D A Mills, D Watkinson and G E Yates, "A Proliminary
Experimental Study of the Kinetics of Sedimentation of
Flocculated Fino Suspensions”, Report to be issued (1985)
D Macdonald, Chen Engineer, 15 March 1985)
A Rushton and M Spear. | Eiltzationnd Senaratione,ay/J)
1975», 254; A Rushton, Erocess Eng, (September 1900), 49.
DF Moir and H J Pearse, "The Use and Selection of
Flocculante", (Harwell = Warren Spring State of the Art Report
SARI6 - available as ¥D/21088), 19%
Dunlop, "Floceulation o£ ASL by the “Prvteen" Partial
Lysis) Route Part I", Agricultural Division Report, 421,10
Gost
P Chokrabeti, "Industrial Solid/Liquid Separation - A Review
and Proposals for Research", Report CL-R/28/1986/4 (1978)
A Harrison, "Seale-Up of Centrifuges for Biochemical Pr
Part A: Sealing up the "Pruteen' Process. Agricultural
Division, Planning & Coordination Department Report. 192
September, 1983
D J Tagg, "Optimisation of Disc Stack Spacing", Agricultural
Division Research Minute, 20th July 1981
x 5 Rogers. "Protein 1 Centrifuges, Conclusions and
Recommendation’, Projects and Engineering Department Niaut
20th March 1979.
B Fitch, “Current Theory and Thickener Design (1975). Parts
1,2 and 9", Eilteation and Separation, July-December 1975
GBH ENTERPRISES, LTD
VULCAN Catalyst Process
Technology Consultancy
Sales and Service
Gerard B. Hawkins
Managing Director, C.E.0.
Stype: GBHEnterrises
Ce: +1-312-235-2610
+52 55 2108 3070
[email protected]
HawkineGerard@@amall.com
VULCAN Catalyst Process
Technology Consultancy
Sales and Service
Gerard B. Hawkins
Managing Director, C.E.0.
Stype: GBHEnterrises
Ce: +1-312-235-2610
+52 55 2108 3070
[email protected]
HawkineGerard@@amall.com
Download
Download Slideshow Get the original presentation fileQuick Actions
Statistics
- Views 9,096
- Slides 74
- Favorites 7
- Age 4400 days
Related Slideshows
11
8-top-ai-courses-for-customer-support-representatives-in-2025.pptx
JeroenErne2
48 views
10
7-essential-ai-courses-for-call-center-supervisors-in-2025.pptx
JeroenErne2
47 views
13
25-essential-ai-courses-for-user-support-specialists-in-2025.pptx
JeroenErne2
37 views
11
8-essential-ai-courses-for-insurance-customer-service-representatives-in-2025.pptx
JeroenErne2
34 views
21
Know for Certain
DaveSinNM
22 views
17
PPT OPD LES 3ertt4t4tqqqe23e3e3rq2qq232.pptx
novasedanayoga46
26 views