Dynamic Modeling of a Batch

Bioreactor for Trans-esterification of

Waste Vegetable Oil
Nabeel Adeyemi
Department of Mechanical Engineering,
Faculty of Engineering,
International Islamic University Malaysia
Malaysia
Supervisor: Prof A.K.M. Mohiuddin (Mechanical Eng Dept, IIUM),
Co-Supervisor: Assoc.Prof Dr Tariq Jameel (Biotechnology
Engineering Dept, IIUM)
Objectives
This primary aim is this work is to model the
transesterification of waste cooking oil (WVO)
using Computational Fluid dynamics (CFD)
techniques where reaction and flow in a reactor
are simultaneously considered
Introduction
Biodiesel Production based on Food Oils.
Fuel vs Food debate
Alternative raw material
WCO-thermally-degraded Food Oils
(>17hrs domestic/ industrial use at 60-90C or more).
Introduction (Contd)
Factor affecting biodiesel production
temperature, ratio of alcohol to oil,
catalyst type and amount, FFA and
water content and mixing intensity
Optimization of biodiesel production
carried out in lab flask accounting
ONLY for the kinetics related effect.

Problem Statement
Yield in WCO transesterification is about
80-85% in batch reactors due to mass
transfer and kinetics-related limitations.
The reason alluded to this is with little
reference to physical condition of
reactor, where the mixing of the
reactant can be significantly improved
during the reaction
RESEARCH PHILOSOPHY
If the hydrodynamics in bioreactor are
considered and accounted for, along
the reaction kinetics, reaction
parameters can be easily optimized for
varying process condition in reactor
design using CFD (Multiphysics
approach)
Coupled multiphysics
phenomena
Momentum
(Navier-Stokes Equations)
Energy
(Convection and Conduction, Heat transfer)
Mass
(Convection and Diffusion, Reaction)
Velocity, pressure
Temperature
Density, viscosity
Thermal conductivity
Heat capacity
Reaction rate
reaction rate
Concentration
1
2
3
Previous Milestone
Concluded WVO transesterification using
Taguchi L9 orthogonal array design
Formulated a kinetic model for the reaction
Simulated the flow in 2 D axi-symmetrical mode

METHODOLOGY
Kinetic model
Nonlinear regression (Least square fitting method)
CFD simulation-Multiple Reference frame (MRF)
model 3D
(RANS k-e, RSM and LES) turbulent models
Exploitation of P.I.V. measurements:
data analysis (RANS k-e, RSM and LES) turbulent
models

Results
Kinetic Model for Transesterification
TRI + OH<->DG + ES (1)
r1f=k1f*[TRI]*[OH]; r1r=k1r*[DI]*[ES]
d[TRI]/dt= k1f*[TRI]*[OH]-k1r*[DI]*[ES]
DG + OH <->MG + ES (2)
r2f=k1f*[DI]*[OH] r2r=-k1r*[MG]*[ES]
d[DI]/dt = =k1f*[TRI]*[OH]-k1r*[DI]*[ES]-k2f[DG][OH]+k2r*[MG]*[ES]

MG +OH <->tr + ES (3)
r3f=k3f*[MG]*[OH], r3r=k3r*[tr]*[ES]
d[MG]/dt = =k3f*[MG]*[OH]-k3r*[tr]*[ES]+k2f[DG]*[OH]-k2r*[MG]*[ES]
d[ES]/dt = k1f*[TRI]*[OH]-k1r*[DI]*[ES]]-k2f[DG][OH]+k2r*[MG]*[ES]+k3f*[MG]*[OH]-
k3r*[tr]*[ES]

With initial value of TG=0.49746, DG=0.82212, MG=0.08876, GLY=0.01378
Equation 1-3 solved
Select Initial value for preliminary fitting
specie concentration(TG, DG, MG)
k values
Compare between the experimental data and the
kinetic model
estimated using non linear regression by
minimizing the standard
Effect of impeller type, position and speed on
yield: Rushton Impeller







Figure 2: FAME wt (%) at (a) 60 (b) 65 (c) 70 C for N = 600, 650, 700 rpm and
IBC = 20, 25, 30 mm for Rushton impeller in baffled reactor













Figure 3: FAME wt (%) at (a) 60 (b) 65 (c) 70 C for N = 600, 650, 700 rpm and IBC = 20, 25, 30 mm
for Rushton impeller in unbaffled reactor
Effect of impeller type, position and speed on
yield: Elephant ear Impeller















Figure 4: FAME wt (%) at (a) 60 (b) 65 (c) 70 C for N = 600, 650, 700 rpm and IBC = 20, 25, 30 mm for Elephant
Ear impeller in baffled reactor














Figure 5: FAME wt (%) at (a) 60 (b) 65 (c) 70 C for N = 600, 650, 700 rpm and IBC = 20, 25, 30 mm
for Elephant Ear impeller in unbaffled reactor
Table 2: S/N ratio of yield
Temp(
C)
Speed (rpm)
IBC
(mm)
Rushton
unbaffled
S/N ratio
Elephant Ear
(Baffled)
S/N ratio
60 600 20 38.86 39.25
60 650 25 39.20 39.37
60 700 30 39.09 39.22
65 600 25 39.05 39.23
65 650 30 39.10 39.23
65 700 20 38.86 39.00
70 600 30 39.25 39.11
70 650 20 39.05 39.18
70 700 25 39.12 39.31
Linear relationship of temperature, speed and
Impeller bottom distance
60 65 600 650 20 25
Yield using Rushton Impeller
39.0624 0.0146t 0.0604t 0.0111S 0.0509S 0.1387I 0.0578I = + +
60 65 600 650 20 25
Yield using Elephant ear Impeller
39.06 0.13t 0.12t 0.29S 0.07S 0.27I 0.21I = + +
(1)
(2)
Table 3: Statistical correlation of temperature; Speed and IBC to yield
and peak yield time
Rushton (Peak Yield time) Elephant ear (Peak Yield time)
SE Coeff T p SE Coeff T p
Constant 2.07 9.13 0.01 0.42 218.57 0
T 3.83 2.05 0.18 0.77 -0.57 0.63
S 3.83 -1.74 0.22 0.77 -0.72 0.55
IBC 3.83 -2.74 0.11 0.77 1.34 0.31
T*S 5.74 -2.49 0.13 1.16 1.4 0.30
T*ID 5.74 0.87 0.48 1.16 -0.82 0.50
S*IBC 5.74 1.87 0.20 1.16 -0.03 0.98
p-value < 0.05 for all parameters
Non
parametric
model

















Figure 5: surface plot of FAME weight yield as function
of peak yield time and temperature
60
65
70
5
10
15
400
410
420
430
440
450

temp (C)
peak yield time (min)

F
A
M
E

w
t

(
g
)
415
420
425
430
435
440
445
3
0
2
5
2
0
1
5
1
0
5
peak yield tim
e (m
in)
60
61
62
63
64
65
66
67
68
69
t
e
m
p

(
C
)
350
350
360
360
370
370
380
380
390
390
400
400
410
410
F
A
M
E

w
t

(
g
)
F
A
M
E

w
t

(
g
)







Figure 6: surface plot of FAME peak yield time as a
function of IBC distance and speed
20
25
30
600
620
640
660
680
700
0
5
10
15
20
25
30
35

speed (rpm)
distance (mm)

p
e
a
k

y
i
e
l
d

t
i
m
e

(
m
i
n
)
5
10
15
20
25
30
600
650
700
20
22
24
26
28
30
4
6
8
10
12
14
16
speed (rpm)

distance (mm)

p
e
a
k

y
i
e
l
d

t
i
m
e

(
m
i
n
)
5
6
7
8
9
10
11
12
13
14
15








Figure 7: surface plot of FAME weight yield as a function of
temperature and speed
600
620
640
660
680
700
60
65
70
350
360
370
380
390
400
410
temperature (C)

speed (rpm)

F
A
M
E

w
t

(
g
)
355
360
365
370
375
380
385
390
395
400
405
60
62
64
66
68
70
600
650
700
400
410
420
430
440
450

speed (rpm)
temp (C)

F
A
M
E

w
t

(
g
)
410
415
420
425
430
435
440
445
PIV
Reactor Property Dimension
impeller bottom clearance, C (mm) 0.11T-0.27T
Height, H (mm) 150
Tank diameter, T (mm) 130
Total Liquid Height, L (mm) 0.34T
Impeller Diameter, D (mm) 0.23T
Table 1: Physical dimension of biodiesel Reactor
RESULTS (Contd)
Figure 1. Simulated mean tangential, radial and axial velocity at impeller
bottom clearance, C= 0.11T, 0.15T, 0.19T, 0.23T and 0.27T for unbaffled
Figure 2: Time average velocity field (m/s) for full tank using Rushton
Impeller (unbaffled)
Figure 3: Time average velocity field (m/s) for full tank using Rushton
Impeller (unbaffled)

0
0.2
0.4
0.6
0.8
1
1.2
1.4
1.6
1.8
2
0.1 0.0 0.8 0.9 1.0
r/D
U
/
U
m
a
x
(
m
/
s
)
K-E
LES
RSM
PIV
Figure 4: Comparison of mean Tangential velocity for
RAN (k-e), LES, RSM models at 30 mm from shaft centre
Physical Model
Bioreactor with Rushton Impeller

0 100 200 300 400 500 600 700 800 900 1000 1100 1200 pix
0
100
200
300
400
500
600
700
800
900
1000
pix
Statistics vector map: Vector Statistics, 3931 vectors (1209)
Size: 12801024 (0,0)
Figure 5: Velocity (vectors) of stirred flow with
20m PSP using Rushton impeller at 600 rpm

0 100 200 300 400 500 600 700 800 900 1000 1100 1200 pix
0
100
200
300
400
500
600
700
800
900
1000
pix
Statistics vector map: Vector Statistics, 3931 vectors (1209)
Size: 12801024 (0,0)
Figure 6: Time average velocity field (m/s) for full tank using
Elephant Ear (unbaffled) (a) CFD (b) PIV

Figure 8: Time average velocity field (m/s) for full
tank (a) CFD (b) PIV
Conclusion
Goals achieve
Effect impeller on FAME yield
Non-parametric model using temperature, speed and
bottom distance to model FAME yield and peak time
PIV evaluation of the velocity structure by Rushton and
Elephant ear
Local rate of energy dissipation
Future Thrust
PIV resolution to determine kolmorogov scale
for spatial variability of parameter during mixing
(macro,meso, or micro) mixing
Time average shearing field (s-1) computed by
Local rate of energy dissipation
Coupling of reaction and flow model
Thank you
.u 0 V =
T
u
(u. )u . I+ ( u ( u) ) F p
t
q
c
( + V = V V + V +

c
Governing Equations
For the fluid flow, the governing equations (continuity and momentum)
for incompressible flow were used;
F is source term dependent on the velocity component and u is
the absolute velocity, which can be calculated by
The variables velocity and stress variables are decomposed into time average
and fluctuating component which are substituted into the governing equations
for incompressible flow to give the RANS
to the momentum equation called the Reynolds stress tensor
Decomposing the variables in Navier-Stokes equation yields an additional term,
i j
u u
' '
( )
u r v e = +
r ur r r
TURBULENT MODEL
This gives a system of equations of 13 unknown variables and 9 equations and is
therefore not closed
2
3
j
i
i j t ij
j i
u
u
u u
x x
o k
| |
c
c
' ' = + |
|
c c
\ .
The closure of the Reynolds stress term results in the development of turbulence
models such as the standard k-e, RNG k-e, realizable k-e and Reynolds stress models
Relationship between the Reynolds stresses to the mean flow velocity gradients
and can be expressed as
2
t
k
C

c
=
( )
2
1
. U. = U U.
2
T
T
T
t
q
k
q k k q c
o
k
c
V + V + V V + V
c
(
| |
| |
(
| |
\ .
( \ .

( )
( )
2 2
1
. U. = U U.
1 1
2
T
T
C C
T
t
c
q
c c c
q c c q
c c
o k k
c
V + V + V V + V
c
(
| |
(
|
( \ .

k the turbulent kinetic energy, defined as
1
2
i j
u u k
' '
=
1 2
1.3, 1, 1.44, 1.92, 0.09
k
C C C
c c c
o o = = = = =
Cylindrical Coordinates
( )
r
r r
r
r
z
r r
r
r
g
V
r z
V V
r
rV
r r r
r
P
z
V
V
r
V V
r
V
r
V
V
t
V

u u

u
u u
+
(

c
c

c
c
+
c
c
+
|
.
|

\
|
c
c
c
c
+
c
c
=
|
|
.
|

\
|
c
c
+
c
c
+
c
c
+
c
c
2 2
2
2
2
2
2
2 1 1
z
z z z z
z
z z
r
z
g
z
V V
r r
V
r
r r z
P
z
V
V
V
V
r
V
V
t
V

u
+
(

c
c
+
c
c
+
|
.
|

\
|
c
c
c
c
+
c
c
=
|
.
|

\
|
c
c
+
c
c
+
c
c
+
c
c
2
2
2
2
2
1 1
( )
u
u u
u
u u u u u u

u u

u u

g
V
r z
V V
r
rV
r r r
P
r z
V
V
r
V V V
r
V
r
V
V
t
V
r
z
r
r
+
(

c
c
+
c
c
+
c
c
+
|
.
|

\
|
c
c
c
c
+
c
c
=
|
.
|

\
|
c
c
+ +
c
c
+
c
c
+
c
c
2 2
2
2
2
2
2 1 1
1
Centrifugal force
Coriolis force