You are on page 1of 11

CHAPTER THREE

DEVELOPMENT OF MATHEMATICAL MODEL



3.0: Modeling of RBC Process
The rotating biological contactor (RBC) process offers the specific advantages of a
biofilm system in treatment of wastewater for removal of soluble organic
substances and stabilisation of nitrogen compounds. Being a unique adaptation of
the moving-medium biofilm system, it facilitates easy and effective oxygen
transfer. The RBC process can be optimized and adapted under different
conditions using modeling tools. In this work, it has been attempted to frame a
mathematical model for a four-stage RBC process in simple and realistic ways.
The model is based on the well-known principles of one-dimensional mass
transfer and transport of substances. Biofilm models are more complex than
models for conventional activated sludge systems. This is because of mass transfer
limitation of substrates in the different reaction processes. The biochemical
conversion process is represented by a mixed-culture biomass environment.





The processes considered to take place inside the biofilms are as enumerated
below:
1. Molecular diffusion of substrates into and of the end products out of the
biofilm.
2. Metabolic reactions within the biofilm
3. Microbial growth and endogenous respiration, which govern the population
dynamics inside the biofilm.
4. Solid displacement by advection and diffusion.
5. Attachment and detachment of biomass at the biofilm surface.
The basic principle for modeling quantitatively is conservation of mass, which
expressed as:

system the in
component
the of
n consumptio
of Rate
system the in
component
the of
production
of Rate
system the
of out
component
the of
flow Mass
system
the o
component
the of
flow Mass
system the in
component
of mass of
on accumulati
of rate Net
int
3.1
Developing a mathematical model for a multi-species biofilm system that will suit
a predictive tool for fast performance and fewer complication requires making
certain assumptions and such assumptions must not ignore any important event of
the system.
The following assumptions were made in developing the model:
1. The model assumes a one-dimensional multi-substrate homogenous biofilm
system. The concentration gradients are only in the direction perpendicular to the
support surface.
2. The active fraction of biofilm remains constant. Surface detachment is
proportional to biofilm growth rate. Monod kinetics was assumed to model
the growth rate of the microorganisms.
3. Complete mixed flow conditions exist due to aeration and mixing as the
disc rotates. The liquid film entrained on the disk surface is stripped off and
completely mixed with the bulk liquid as soon as the disk re-enters the
liquid.
4. The system is isothermal and the transport of substrates in the biofilm is
governed by molecular diffusion and obeys Ficks law.
5. The same mass of liquid adheres to the same bio-film throughout the disc
rotation.
6. Diffusivity coefficients of substrate and oxygen are based on water as
medium.
7. Model is valid only for steady state condition, since the bio-film thickness
is constant.
8. The disc is 45% submerged in the wastewater.
Figure 3.2 is a schematic representation of a four-stage rotational biological
contactor RBC employed for the treatment of oil refinery wastewater.

Figure 3.2: Schematic Representation of the four-stage RBC Process
(Castillo et al, 2007)

3.1: Mass Balance in RBC
The biofilm is treated as a continuum [Wanner, 1996] consisting of a liquid phase
and solid phase. The biofilm liquid phase consist of dissolved components, while
the biofilm solid phase consist of particulate components. Mass transfer takes
place across the liquid-biofilm interface as diffusion and biochemical reactions
within the biofilm. Owing to the dynamic nature of oxygen transfer, which is the
typical limiting substrate in crude oil refining wastewaters, the boundary layer is
assumed to be completely mixed and concentrations averaged over the entire
volume.
3.1.1: Mass Balance of Substrate
Considering the convection, diffusion and reaction in the biofilm, a mass balance
for the substrate in the RBC shown in Figure 1 is given as:

=
ism microorgan by
n consumptio Substrate
convection and
diffusion by out Flow
convection and
diffusion by in Flow
on Accumulati
Substrate
(3.2)
Equation 1 is mathematically expressed as:
S
V dZ
e
S
e
F
Z
S
A
S
D
Z
S F
Z
S
A
S
D S F
Z
S
A
S
D
t
S
V 9 +
c
c

c
c
+ +
c
c
+
c
c
=
c
c

|
.
|

\
|
0 0 0 0
(3.3)
Where, F
0
and F
e
= Volumetric flow in and out (Lh
-1
) respectively, V = Volume of
biofilm submerged in bulk liquid fluid (L), A = Submerged area in the plane
perpendicular to the direction of diffusion (m
2
), S = Substrate concentration (mgL
-
1
), D
S
= Diffusivity of substrate (m
2
/h), t = time (h),
S
9
= Rate of substrate
utilization.
Multiplying through Equation (3.3) reduces it to:
S
V dZ
Z
e
S
e
F AdZ
Z
S
S
D
t
S
V 9
c
c

c
c
=
c
c
2
2
(3.4)
Putting
dV AdZ=
into Equation (3.4) and dividing through by
AdZ
gives:
S
Z
e
S
A
e
F
Z
S
S
D
t
S
9
c
c

c
c
=
c
c
2
2
(3.5)
Assuming uniform concentration and flowrates, Equation (3.5) becomes:
S
Z
S
A
F
Z
S
S
D
t
S
9
c
c

c
c
=
c
c
2
2
(3.6)
Assuming that at any given time step, the convection through the biofilm is
negligibly small compared to the diffusion process [Famularo et al. 1978, Wanner
et al. 1995], Equation (3.6) reduces to
S
Z
S
S
D
t
S
9
c
c
=
c
c
2
2
(3.7)
In RBC units, the substrate and oxygen are simultaneously growth limiting. The
rate of substrate utilization by microorganisms,
S
9
is related to specific growth
rate, biomass concentration and oxygen utilization by the expression:


U
U
K
U
S
K S
S
Y
X
m
S
+ +
= 9
u
(3.8)

Where,
m
u
= Maximum specific growth rate (h
-1
), Y = Yield coefficient (mg/mg),
X = Biomass concentration (mg/L), K
S
= Half saturation constant (mg/L), U =
Oxygen concentration (mg/L), K
U
= Monods saturation concentration (or half-
saturation concentration) of oxygen (mg/L).
Substituting Equation (3.8) into (3.6) will give:

U
U
K
U
S
K S
S
Y
X
m
Z
S
r
Z
S
S
D
t
S
+ +

c
c

c
c
=
c
c
u
e
2
2
(3.9)
At steady state,
0 =
c
c
t
S
, therefore Equation (3.9) reduces to:

U
U
K
U
S
K S
S
Y
X
m
dZ
dS
r
dZ
S d
S
D
+ +
=
u
e
2
2
(3.10)

3.1.2: Mass Balance for Oxygen
The mass balance for oxygen generation, transfer and utilization in the biofilm is
given as:

Uc Ug
Z
U
r
Z
U
U
D
t
U
9 9 +
c
c

c
c
=
c
c
e
2
2
(3.11)

Where D
U
= Diffusivity of oxygen (m
2
/h),
Ug
9
= rate of oxygen generation and
transfer,
Uc
9
= rate of oxygen consumption.
The rate of oxygen generation and transfer, as modified in Kim and Molof model
(Vijay et al, 2004) is given as:

( ) U
S
U a
L
K
Ug
= 9
(3.12)
Where K
L
= Oxygen transfer coefficient (m/h), a = Specific surface area of
exchange (m
-1
), U
S
= Saturation concentration of oxygen in air (mg/L).
The rate of oxygen consumption for biomass growth and maintenance is given as

U
U
K
U
S
S
K
S
m
U X
Y
U
UC
+ +
= 9 u
m
/
(3.13)
Where
U
m
= Specific respiration rate coefficient, that is mass of oxygen consumed
per mass of living cell (mg/mg),
U X
Y
/
= Yield coefficient (mg/mg).
Substituting Equations (3.12) and (3.13) into (3.11) will give:
( )
U
U
K
U
S
S
K
S
m
U X
Y
U
U
S
U a
L
K
Z
U
r
Z
U
U
D
t
U
+ +
+
c
c

c
c
=
c
c
u
m
e
/
2
2
(3.14)
At steady state
0 =
c
c
t
U
, therefore Equation (3.14) reduces to:
( )
U
U
K
U
S
S
K
S
m
U X
Y
U
U
S
U a
L
K
Z
U
r
dZ
U d
U
D
+ +
+
c
c
= u
m
e
/
2
2
(3.15)

3.1.3: Mass Balance for Biomass
Similarly, the mass balance for the biomass in the RBC unit, with the cognizance
that feed stream is assumed sterile, is derived as:

|
|
.
|

\
|
|
|
.
|

\
|

+ +
=
d
K
U
U
K
U
S
S
K
S
X
r
V
out
X
out
F
dt
dX
r
V
max
u
(3.17)
Where X = Biomass concentration (mg/L),
d
K
= Endogenous decay rate (h
-1
)
Dividing Equation (3.17) by Vr and noting the assumption that biomass
concentration is uniform yields:

|
|
.
|

\
|
|
|
.
|

\
|

+ +
+ =
d
K
U
U
K
U
S
S
K
S
X X
r
V
out
F
dt
dX
max
u
(3.18)
Substituting
r
V
out
F
= u
into Equation (3.18) yields:

|
|
.
|

\
|
|
|
.
|

\
|

+ +
+ =
d
K
U
U
K
U
S
S
K
S
m
X
X
dt
dX
u
u
(3.19)
Where
u
= Hydraulic retention time (h)
At steady state
0 =
dt
dX
, and Equation (3.19) by rearrangement becomes:
|
|
.
|

\
|
|
|
.
|

\
|
|
|
.
|

\
| +
=
+
+
u
u
u
d
K
U
U
K
U
S
S
K
S
m
1
(3.20)
Substituting Equation (3.20) into (3.10) and (3.15) will respectively yield:
|
|
.
|

\
| +
=
u
u
e
d
K
Y
X
dZ
dS
r
dZ
S d
S
D
1
2
2
(3.21)
( )
|
|
.
|

\
| +
+
c
c
=
u
u m
e
d
K
U X
Y
U
U
S
U a
L
K
Z
U
r
dZ
U d
U
D
1
/
2
2
(3.22)
Where a
L
K , (volumetric oxygen transfer coefficient) is expressed as (Kim and
Molof, 1982):

732 . 0
) ( 0011 . 0 Nv a
L
K = (3.23)
And
d
D
Nv
5 . 0 5 . 1
=
= (3.24)
Where Nv = ., = = rotational speed (S
-1
), D = Disk diameter (m), d = disk
spacing (m).
Equations (3.21) and (3.22) are the model equations for substrate and oxygen
utilization in the biofilm of the Rotational Biological Contactor (RBC) system.
The models, which comprises experimental measurements, the reaction and
transport characteristic terms are subject to the following conditions:
With initial condition, at
0
, 0 ,
0
, 0 , 0 S
f
S
b
S S U U Z t = = = = < < = o
And boundary condition,
( ) ( )
f
S S and
f
U U air to osure on
b
S S and
b
U U bulkliquid the in
dZ
dS
dZ
dU
Z and t at
= =
= =
= = = >
: exp
, :
; 0 , 0 o

Where
o
= thickness of the liquid film boundary layer (m), S
0
= initial substrate
concentration (mg/L), S
b
= substrate concentration in the bulk, S
f
= substrate
concentration in liquid film (mg/L), U
0
= initial concentration of oxygen in the
liquid film (mg/L), U
b
= oxygen concentration in bulk, U
f
= oxygen concentration
in liquid film (mg/L).

3.3: Solution to Model Equations
The model equations, which are parabolic ordinary differential equations, are not
amenable to analytical solution technique. The substrate and oxygen concentration
profiles within the biofilm were approximated as parabola and Liebnizs rule was
applied to an expansion of Equations (3.21) and (3.22). The boundary values were
converted into an initial value problem and solved using fourth order Runge-Kutta
method. A computer program using Visual Basic 6.0 allowed the interactive loop
calculations for solving the models.

3.3: Parameter Selection and Determination
For the model solution, coefficients and calibration are necessary. Sensitivity of
different Kinetics is observed by using a range of values for each kinetic
parameter. The range of values considered for various kinetic parameters are taken
from literatures as indicated in Table 3.1 below. While Table 2 indicates RBC
design/operation parameters of plant-scale RBC (Warri Refining Company).
Table 1: Kinetic Parameters
Parameters Values
Parameters Values
Yield coefficient 2.0 5.0
Endogenous Death rate 0.1 hr
-1

Diffusivity Coefficient for
oxygen
5.0 * 10
-5
cm
3
/sec (Chan and Stenstron,
1979)
Diffusivity Coefficient for
Substrate
0.64*10
-5
cm
3
/sec (Chan and Stenstron,
1979)
Table 2: Process Parameters of the RBC Unit
Parameters Values
Reactor Length 5.5 m
Bio Reactor Radius 0.5 m
Number of disc 12
Diameter of each disc 3.6 m
Width of the trough 5.2 m
Depth of trough 1.2 m
Rotational Speed 1.6 rpm
Hydraulic Retention Time 10 hrs

You might also like