You are on page 1of 68

MSc & MEng

Batch mixing of fluids and particles

Solid-liquid mixing

Dr Gül Özcan-Taşkın
Professor Chris D Rielly
Lecture outline
❖Introduction
❖Definitions
❖Suspension of heavy solids
❖Liquid-solid mass transfer
❖Distribution of heavy solids
❖Drawdown of floating particles

2
Liquid-solid operations
Process objectives
❖ suspension of particles (for ex. catalyst particles), initially resting
on the bottom of the tank
❖ expose the maximum solid-liquid surface area for mass transfer
and chemical reaction
❖ formation and maintenance of a uniform distribution throughout the
liquid volume:
❖ product in the form of a slurry
❖ particular requirement of systems with draw-off of slurry
product, or catalyst recycling

3
Solid liquid mixing
❖Most common mixing application

❖ 90 % of “fluid only” plants;


❖ 60 % of solid handling plants achieve desired process result

❖ ~ 3 months start up time for a “fluid only”;


❖ 9- 18 months for a solids handling plant
Industrial Practice- Needs
▪ Avoid solid accumulation at the base of a tank

▪ Maximize contacting area between the solid and


liquid phases for mass transfer

▪ Achieved required homogeneity

▪ Select appropriate equipment


Suspension conditions
❖ Just suspension (complete suspension) conditions (NJS)
❖ "no particles remain on the bottom of the tank for more than 1-2 seconds”
(Zwietering, 1959)

❖ Just drawdown conditions (NJD)


❖ no particles remain floating at the liquid surface for a certain period of time

❖ Homogeneous suspension
❖ no solids concentration gradients are present within the vessel, i.e. the solids are
uniformly distributed
❖ in practice, requires very high specific power input
❖ often no advantage in terms of mass transfer rate
❖ may be required for certain products in the form of a powder dispersion or for product
removal

6
Solid-Liquid Mixing
Suspension Distribution Draw down
(s>  L ) ( s<  L)

Speed and power to achieve


no particles at the a given degree of no particles at the
vessel base homogeneity liquid surface

NJS NJD
OFF-BOTTOM SUSPENSION
OF HEAVY SOLIDS

8
Solid suspension
Factors affecting off-bottom suspension

➢ Impeller type, size, pumping mode, position


➢ Impeller speed\power input
➢ Tank and tank base shape
➢ Tank volume\scale of operation
➢ Tank internals
➢ Particle properties: size and density with respect to that of the
liquid
➢ Particle concentration and slurry rheology
➢ Solid shape (spherical, needle)
Mechanistic models for solid suspension

➢ Turbulence model
➢ Suspension due to energy transfer from turbulent eddies of a size
similar to that of dp .

➢ Fluid velocity model


➢ Suspension due to average velocity and hydrodynamic forces (lift and
drag) acting upon the particles.
Flow patterns

Interaction of the
flow patterns
generated by an
impeller and those
acting on particles
on the vessel base
Models for solid suspension (ct.)
• Neither of the two classes of models alone represent the exact
physical conditions.

• Both convection and turbulence act on settled particles.


Semi-empirical correlation by Zwietering

0.45
 g 
N JS = s   
0.1 0.13 0.2
 X dp D -0.85

 L 

No solids remain at the vessel base for more than 1-to-2 s


Just-suspension conditions
➢ Zwietering's (1959) empirical
correlation covers a wide range of
conditions
➢ accurate to about ±30% and N < NJS
dimensionless!!!

0.45
0.1 0.2  g 
NJS = s d p   D −0.85 X 0.13
   fillets of particles:
last to be removed

16
Limitations and applicability
❖ “s” values have to be known

❖ Not applicable outside the experimental conditions

❖ Applies to free and hindered settling conditions (i.e. low to moderate


X), but may not be valid in significantly non-Newtonian systems (i.e. at
high X)
Limitations and
Parameter Range covered
applicability (kg/m 3) 560 – 1800
d p (mm) 140 – 520
Operating conditions over which X(%) 0.5 – 20
Zwietering correlation was obtained T(m) 0.1 – 0.6
D/T 1/6 – 1/2
nb 4
C/T 1/20 – 1/2
Impel. type 4
Vessel base Flat, dish &
shape conical
mL(mPa s) 0.3 - 9
‘s’ values for Zwietering Correlation

Flat base tank


‘s’ values for
Zwietering Correlation (1)

Dish base tank


(the first value in the parenthesis is the
impeller diameter; the second blade width)

(1) Handbook of Industrial Mixing (Paul et al, 2003)


Recommendations
• Impeller Geometry

– Axial flow impellers (e.g. hydrofoils) pumping downwards are the most;
radial flow (e.g. 6DT) and sawtooth impellers are the least energy
efficient for off-bottom suspension.

– Low C generally more energy efficient, except at extremes (e.g.


D/T>0.5 C/T<1/6)
Recommendations
• Tank Geometry

– Flat base tanks not recommended: cone and fillet formation at the
bottom

– Multiple impellers for processes where H varies


Scale-Up of Suspension Processes
Geometrical similarity

D1 D 2 W1 W 2 H 1 H 2
= ; = ; = ; etc .
T1 T 2 D 1 D 2 T1 T2
Scale up (ct’ed)
• Zwietering’s Correlation

NJS  D-0.85

NJS decreases as scale increases


It is worth bearing in mind that there is some disagreement
on the value of the exponent on D reported by different
researchers.
Power Consumption
 Pjs   Po  av N D 
3 5
  =  
 V   V 

V = k1 T3, NJS = k2 D-x, D = k3T

where k1, k2, k3 are constants


Power Consumption (2)
• By substitution
 Pjs  ( 2− 3 x )
 T
 V 

• For x = 0.85  Pjs  − 0.55


   T
 V 
• Power per unit volume decreases on scale up.
Power Consumption (3)
• For x = 0.67
 P js 
   T 0

 V 

• Conservative approach: keep P/V ct on scale up (reported


values of x: 0.67 - 1.0)
• can lead to under design when scaled down!
Scale-up
Rules
LIQUID-SOLID MASS
TRANSFER

29
Liquid-solid mass transfer
C As liquid-solid mass
CA transfer coefficient
C As 6 s 6m
as = =
CA dp  pd p
particle
J A = ks as (C As − C A )

molar transfer rate


❖ CAs is the concentration of A at the surface of the particle of species A per
❖ CA concentration of A in the liquid bulk unit volume
❖ as is the specific interfacial area of the particles (m2/m3)

30
Liquid-solid mass transfer coeff
log

Mass transfer coefficient,


fully (Harnby et al. 1985)
increasing movement and
suspended
suspension
kJS

surface
rolling aeration
ks

NJS NSA Stirrer speed, N


log
❖ No significant gain for N > NJS in terms of ks
❖ despite significant increase in power: P  N3
❖ surface aeration reduces liquid-solid interfacial area

31
Mass transfer correlations
• Often these are based on the Froessling equation

Sh = 2 + 0.72 Re1p/ 2 Sc1/ 3


ks d p mL
Sh =  L d pus Sc =
DL Re p =  LDL
mL
• however, usually the particle slip velocity is unknown and so we write  d p4
Re p = 3
• where  = P /  LV L
1/ 4
 d 
1/ 3
 mL 
4
ks d p
• Sh = = 2 + 0.4   
e.g. p

DL  
  L DL 
3
 L  Sano et al. (1974)

32
Liquid-solid mass transfer
• Usually solid particles are finely divided
– liquid-solid mass transfer coefficients are typically large
compared to gas-liquid coefficients
– liquid-solid mass transfer is not the rate limiting step

• It is costly and with no major mass transfer benefits to


operate under homogeneously suspended conditions
– except where the tank is a holding vessel for solid-liquid mixtures
– solid product must be drawn off continuously

33
DISTRIBUTION OF SOLIDS IN
A LIQUID

34
Solid Distribution
– Off-bottom suspension  mass transfer

– Certain degree of homogeneity required for other


processes\products

– The mechanisms considered for distribution: convection and


turbulence. Convection distributes particles by drag force, the
turbulence eddies act to distribute at a smaller scale by fluctuating
drag forces acting in all directions.
Measurement and Assessment
• Different experimental techniques:
conductivity, sampling, visual observations,…

• Mixture quality assessed in terms of:


– Slurry height,
– Relative Standard Deviation,
– Concentration profile
Experimental Techniques- Slurry Height
N1 N2 N3
Experimental techniques- Conductivity
Conductivity probe for local solids concentration

• Reliable but intrusive


technique epoxy
body
• Probe positions need
careful selection to get
representative values

Pt electrodes

38
Probe Installation
Probe Positions in Vessel
• Measurement positions in experimental studies by Taylor, 2002

T=0.61,1.83 m

h/H=0.836

h/H=0.672

h/H=0.508
H=T

h/H=0.344 R 2R/3
1 =45E
h/H=0.180
C=T/4
D

45E 4 blade pitched blade turbine


Relative Standard Deviation (RSD)
1
1  1 n 2
( )
2
RSD =  
Cm  n − 1 1
C ij
− C m 

Cm: Calculated mean bulk solid concentration


Cij : Measured local solid concentration
n : Number of samples
RSD = 0 for perfect homogeneity
Solid Distribution Curve
Measurement and assessment
– RSD: a single value assigned to the whole mixture but does not provide
information on how solids are distributed

– Concentration profiles are useful for a detailed assessment and to correctly


position the inlet and outlet in a CSTR
Homogeneous suspension
Axial solids concentration profiles
h/H
Increasing impeller speed
1
N increasing

“belly” plot

44
RDT: D=T/3 (Baldi et al., 1981)
1.0
d p = 385 µm

position, h / H
d p = 240 µm
0.8
d p = 89 µm

Axial position, z / T 0.6


N = N js
0.4
Axial

0.2
Impeller disc plane

0.0
0.1 1.0 10.0
Normalised
(Local solids
concentration) concentration
/ (Mean concentration)

45
Solids distribution
T/6 from the wall
(ct’ed) 1
0.8
0.6

h/H
0.4
– “Belly plot” on the axial
traverse 0.2
– Position and value of max.
depends on N, C/T, D/T,
0
particle properties,... 0 0.5 1 1.5 2
C/Cmean
Major points related to solids distribution
2T/3 from the top
2

1.5

C/Cmean
1

0.5

0
0 0.2 0.4 0.6 0.8 1
r/R
Practically uniform radial concentration profiles
Table 3. Scale up criteria proposed by different researchers (Taylor, 2000)
Scale-up Criteria
Author Scaling criteria

Impeller speed Power/Volume


-0.67 0
Mak and Ruszkowski (1992) NαT P/VαT

Rieger et al. (1988)

Buurman et al. (1985) NαT -0.78 P/VαT -0.34

Magelli et al. (1990) NαT -0.9 P/VαT -0.70

Barresi and Baldi (1987) NαT -1 P/VαT -1

For comparison: Zwietering (1957) NjsαT -0.85 P/VαT -0.55


-1 -1
For comparison: Tip speed NαT P/VαT

For comparison: Specific power NαT -0.67 P/VαT 0


10
Major points on Rieger et al. (1988)
Mak and Ruszkowski (1992)
solids distribution Constant P/V

(ct’ed) 1

0.1

0.01

1 10 100
Scale factor
Comparison of Vtip and P/V for scale up

D. Taylor, PhD thesis, Cranfield University


Dual Impeller Configurations
• Commonly used in ‘tall’ tanks

• No advantage for suspension: Pjs higher than in


a single impeller configuration.

• Dual impellers improve distribution quality


SOLID SUSPENSION USING A
LIQUID JET

53
Solid Suspension Using a Jet
– Alternative to mechanical agitation

– Good off bottom suspension but poor top-to-bottom


distribution

– Limited access above the vessel, there may not be enough


headroom to install a gear box and motor

– Portable/flexible
Jet Mixer for
Solid
Suspension
Advantages of using jets
Mechanical constraints on agitator size in very large
vessel

Energy efficiency for solid suspension in comparison


to agitation tanks depends on the specific design

Easier to install in wide, shallow vessels: agitator


system requires extra supports
Advantages of using jets
•No moving parts in vessel : jet motion induced by
an external pump (reduced maintenance)

•Low maintenance cost: jet mixing system is very


simple and pump is remote from the tank

•Low capital cost: pump may already be present


for tank drainage.
DRAWDOWN OF FLOATING
PARTICLES

58
Example industrial processes

– Polymerisation reactions,
– Food industry,
– Manufacture of consumer care products,
– Waste water treatment,
– Minerals processing,
– Pharmaceuticals.
Draw down/ Introduction
Particles may float due to (Etchells, 2001):

– Density difference
– Poor “wettability”
– “Low bulk density”
Issues to Consider
– Speed and power required to draw down (and disperse) solids
(agglomerates);
– Variations in liquid level;
– Preventing gas/vapour entrainment from head space;
– Mech. vibrations (in the presence of a vortex);
– Slurry homogeneity;
– Rate of draw down;
– Change in rheology;
– Preventing fouling on vessel internals
Incorporation of floating particles
Depending on the particle type the process may involve several
steps (Freudig et al, 1999):
– Wetting
– Drawing down
– Dispersing
– Dissolving
Definitions
Different researchers used different criteria:
NJD1 No particles remain stationary at the
liquid surface for more than 1-2 seconds
NJD1-2 No particles remain floating at the liquid
surface for more than 1-2 seconds
NJD2-4 No particles remain floating at the liquid
surface for more than 2-4 seconds
Findings and conclusions
Only empirical correlation

4-blade 45º PBT, water, single surface baffle


0.27 < T< 1.8 m 0.29<D/T< 0.60 0.11 <C/T<0.33
0.1< /L<0.76 2<dp<13mm
−3.65 0.42
D   
FrJD =k   
  
T   L 
Drawdown Mechanisms
Draw down of Draw down of Draw down of
solids solids and air solids

Downward pumping impellers Upward pumping impellers

a) Through the recirculation loop b) Through surface aeration c) Through the main circulation loop
1 or 2 baffles and at high
Fully baffled vessels off bottom clearances
Özcan-Taşkın, 2001
Effect of D/T
Effect of impeller diameter on drawdown power
with a downward pumping PBT4-45 (X=1%; four baffles; H=T)

600

D=T/2 D=T/3
500 SA

400
P/V (W/m3)
300

200

100

0
0.2 0.4 0.6 0.8
Cs/T

Small D/T requires high Njd but is more energy efficient


Effect of pumping mode Figure7b. Effect of pumping mode on drawdown power at varying
liquid levels using an LE20 (4 baffles ;T=0.61 ; D=T/3 )
500

LE20 400

(4 baffles; T=0.61; D=T/3)

P/V(W/m³)
300

200

100

0
0 0.2 0.4 0.6 0.8 1
Upward pumping mode more Cs/T
energy efficient H=T PU H=5T/6 PU H=2T/3 PU H=T/2 PU

H=T PD H=5T/6 PD H=2T/3 PD H=T/2 PD


Video clips
Video clips
Video clips

You might also like