Mathematical Modelling for Chemical Engineers

Mathematical Modelling
in Chemical Engineering

Lecture Notes
(with some gaps which should be completed during the lecture hours)

y qy+Δy

qx
qz qz+Δ z

qx+Δ x qy
z

∂ 2T ∂ 2T ∂ 2T ∂T
x kz 2
+ ky 2
+ kx 2
= ρC p
∂z ∂y ∂x ∂t

Prepared by
Dr Serpil Serbetcioglu Awdry
The Middle East Technical University

Mathematical Modelling for Chemical Engineers

CONTENT

1. Systematic Approach to Model Building
1.1 The Idea of a ‘Model’ and Definition of a Mathematical Model
1.2 How to Use Mathematical Models Chemical Engineering
1.3 Using Model as a Tool
1.4 Brief Review on Mathematical Modelling Elements
1.5 Ingredients of Process Models
1.6 Classification of the Model Equations
1.7 Principles of Model Building
1.8 Techniques of Building a Mathematical Model
Mathematical Modelling Problem Sheet 1

2. Formulation of Heat Transfer Phenomena
2.1. Heat Transfer by Conduction on a Plane Wall
2.2. Possible Boundary Conditions on a Plane Wall
2.3 Energy Balance on the Surface of a Medium
Mathematical Modelling Problem Sheet 2

2.4 Heat Transfer in a Flowing Fluid System
Mathematical Modelling Problem Sheet 3

3. Formulation of Mass Transfer with a Diffusional Flow
3.1 Diffusion versus Bulk Convection
Mathematical Modelling Problem Sheet 4

3.3 Mass Transfer and Catalytic Reactions
3.3.1 Mass Transfer Steps in a Catalytic Reaction System

3.3.2 External Mass Transfer
3.3.3 Diffusional Mass Transfer and Effectiveness Factor
3.3.4 Choosing the Rate Controlling Step
Mathematical Modelling Problem Sheet 5

4. Formulation of Heat Transfer from Extended Surfaces
Mathematical Modelling Problem Sheet 6

5. Modelling of Simultaneous Heat &Mass Transfers in Heterogeneous Systems

Mathematical Modelling Problem Sheet 7

6. References and Sources

2
Mathematical Modelling for Chemical Engineers

1 Systematic Approach to Model Building

1.1 The Idea of a Model & Definition of a Mathematical Model

Any model is just about imitation of reality. There are many type of models used
everywhere in human activities: physical or mathematical, i.e economic models,
climate change, cosmology, theology or engineering, you name it.

So, why is there such an interest in the activity of model building and use of models in
human activities? Well, It is clearly a means of gaining insight into the behaviour of
systems, probing them, controlling them, optimizing them. One thing is certain: this is
not a new activity:

“So, just about every noun is a model of some event, process or idea”
(Plato, 400BC)

“A theory is just a model of the universe, or a restricted part of it, and a set of rules that
relate quantities in the model to observations that we make.
A theory is a good theory if it satisfies two requirements: it must accurately describe a
large class of observations …, and it must make definite predictions about the results of
future observations.”
(Stephane Hawkins, 1989 A Brief History of Time, from the Big Bang to Black Holes)

A mathematical model is a formulation that represents a physical process in
mathematical terms based on empirical observations or experiments. We should
never forget this and get so distracted by the model that we forget the real
application, which is driving the modelling. In the process of model building we are
translating our real world problem into an equivalent mathematical problem, which
we solve and then attempt to interpret. We do this to gain insight into the original real
world situation or to use the model for control, optimization or possibly safety studies.

In an elementary picture of the scientific method, there is a ‘real –which is external

world’ and a ‘conceptual world- which is the world of the mind. The conceptual world
can be viewed as having three stages: observation, modelling, and prediction as shown
in the Figure below:

In the observation part of the scientific method we measure what is happening in the
real world and gather empirical evidence.

The modelling part is concerned with analysing the observations. The models
developed based on these observations should explain why that behaviour and results

3
Mathematical Modelling for Chemical Engineers

occurred as they did; and should allow us to predict future behaviours that are as yet
unmeasured or unseen.


Real World Conceptual World
(Physical phenomena) (Modelling)


Phenomena i.e. OBSERVATION
Measuring and gathering information
Gravity,
Global warming,
Absorption,
Reaction,
CREATING
Combustion,
A MODEL of a Phenomena
Conduction,
(Analysing)
Vaporisation,

Diffusion,
Population grows, PREDICTIONS from the model
Central heating, …. COMPARISON (Simulations)
with the real phenomena
(Validation)

(Adopted from Dym and Ivey, 1980)

1.2 How to use Mathematical Models in Chemical Engineering

There are many applications in the Chemical Engineering discipline that mathematical
models can be used, as summarised below:


Mathematical
Models in Chem.
improve
Engineering can be used in R&D
understanding the
training & education.
behavior of a
complex system.

assist in experimental
help to predict the be used in simulation,
design, & trouble
design, control & sensitivity analysis, &
shooting.
operation variables. optimisation.




Modelling & Simulation are valuable tools, since it is much safer and cheaper to
perform tests on the model using computer simulations rather than carrying repetitive
experimentations and observations on the real system, provided that the model used
is VALIDATED.

4
Mathematical Modelling for Chemical Engineers

Model Application Areas in Chemical Engineering

Application Area Model use and aim

Process design Feasibility analysis of novel designs

Technical, economic, environmental
assessment
Effects of process parameter changes on
performance
Optimization using structural and
parametric changes
Analysing process interactions
Waste minimization in design

Process control Examining regulatory control strategies

Analysing dynamics for setpoint changes
or disturbances
Optimal control strategies for batch
operations
Optimal control for multi-product
operations
Optimal startup and shutdown policies

Trouble-shooting Identifying likely causes for quality
problems
Identifying likely causes for process
deviations

Process safety Detection of hazardous operating regimes
Estimation of accidental release events
Estimation of effects from release
scenarios (fire etc.)

Operator training Startup and shutdown for normal
operations
Emergency response training
Routine operations training

Environmental impact Quantifying emission rates for a specific
design Dispersion predictions for air and
water releases Characterizing social and
economic impact Estimating acute
accident effects (fire, explosion)

5
Mathematical Modelling for Chemical Engineers

1.3 Using Model as a ‘Tool’
Mathematically formulated model can be as a tool to predict the behavior of a
particular system for unknown new conditions. However, in order for a model to be
used as a simulation tool it should represent the reality as close as possible. Therefore,
validating of the model using data from the real phenomenon/experiment is an
absolute necessity to gain confidence in the model before using it in a simulation
processes.

Schematic model building steps can be summarised in the following diagram:


PHYSICAL

PROCESS SIMULATION TOOL
DEFINITION OPTIMISATION

PARAMETER ESTIMATION
IDENTFY CONTROLLING

PHENOMENA
Good match ☺
&MODELLING GOAL Bad match !
Use as a
Tool
VALIDATION &
DERIVE MATHEMATICAL
COMPARISON
EQUATIONS, ESTABLISH
THE RESULTS.
BOUNDARY/INITIAL
CONDITIONS

APPLY NUMERICAL COMPUTER

ALGORITHM SOLUTION




1.4 Brief Review on Elements of Mathematical Modelling
Building a mathematical model for a chemical engineering system requires to have the
knowledge of the physical and chemical laws such as rates of mass, heat, and
momentum transport, rates of adsorption-desorption, absorption, diffusion, rates of
reactions governing the processes all of which take place within the boundaries of the
system concern.
The body of knowledge and disciplines involved in building a mathematical model for
chemical engineering processes can be summarized schematically below:

6
Mathematical Modelling for Chemical Engineers

The System and its Boundaries (VERY IMPORTANT)
For constructing a mathematical model for a process, first of all we need to
understand the system to be modelled and its boundaries:
When a system is chosen for modelling, it should be described by its elements, the
interaction between the elements, and its relation with the environment. A system
should have a boundary that distinguishes it from the environment.

A system is influenced by its surroundings or environment via its inputs and generates
influences on its surroundings by its outputs which occur through its boundary.


Balance region (or Control volume)

When we develop models, it is necessary to define regions in the system where we
apply conservation principles and basic physical and chemical laws in order to derive
the mathematical description. These are the balance regions also called 'control
volume or differential element. In model building, the choice of the right 'balance
region' is very important step since this region will be used to represent the whole
system to be modelled. The balance region can vary depending on the system, what
degree of accuracy and what specific results are required from the model. It can be a
total reactor as in Batch or a section of a reactor as in PFR shown below:

7
Mathematical Modelling for Chemical Engineers

BATCH PLUG FLOW REACTOR

Differential volume

Balance region: total reactor Balance region: differential volume

C= f(time) C = f(time, position)
Lumped system (Macroscopic) Distributed system (Microscopic)

Mathematical Representation of a Physical Phenomenon:
In developing a model of a real system, the governing phenomena or mechanism
should be represented in mathematical terms as differential or algebraic equations:

• Differential equations: the rate of change of some system variables with
respect to other variables (i.e. time or system geometry) within the system of
concern, e.g. the symbol of dv/dt may state 'the rate of change of velocity with
respect to time'. A vast number of real world applied problems are governed
by differential equations since many physical and economic laws relate rates of
change.
• Algebraic equations: equations in the form of a polynomial having a finite
number of terms i.e. y = ax2+bx+c
• Non-linear equations: model equations can be linear or non-linear depending
on their unknowns, if they appear as variables of a polynomial of degree higher
than one (as non-linear) or power of one (as linear).

• Independent variables: quantities that are specified by either the geometry of
the system i.e. axial length, radius or state of the system i.e. time

• Dependent variables: quantities which are functions of the independent
variables and evaluated within the model equations, e.g. in a system if
concentration is a function of space & time then it can be shown as C=f (z, t)

• Lumped parameters: quantities that have no spatial variation in the system,
e.g. concentration in Batch Reactor.


• Distributed parameters: quantities that have spatial variation e.g.
concentration in
plug flow reactors.

8
Mathematical Modelling for Chemical Engineers

1.5 Ingredients for Building a Model for a Chemical Process

Model building is not simply formulating a set of equations. The variables which
characterize the system being studied should be clearly defined during the building a
model. Also for the clarity and consistency, model building process should incorporate
a lot more information on all key ingredients i.e.:

• Assumptions: in order to simplify the real phenomena we state some
assumptions which should be reasonable and justifiable.

• Input variables: quantities that determine subsequent evaluations within the
model,
e.g. initial flow rate, initial velocity.

• Output variables: quantities that are consequences of given value of input
variables and parameters, and cannot be given arbitrary values.

• Boundary conditions (B.C): a complete definition of differential equations must
include numerical values for the boundary conditions which are the specific
values of the dependent variables at particular values of the independent
variables.

• Initial conditions (I.C): if the boundary values of dependent variables specified
for the initial values of independent variable, they are called initial conditions.

9
Mathematical Modelling for Chemical Engineers

• Initial and Boundary Value Problems: If the boundary conditions are specified
at one point is called ‘initial value problem, at more than one point the
problem is termed as a ‘boundary-value problem’.

• Sufficiency & Redundancy: In order to obtain a solution to a set of equations, it
is necessary to specify as many independent variables as there are dependent
variables in other words number of unknowns should be equal to number of
equations.


Example 1: Use the equation below representing temperature changes due to
conduction of heat in a rectangular bar:
∂ 2T ∂ 2T ∂ 2T ∂T
k 2
+k 2
+k 2
= ρCv
∂t
∂x ∂y ∂z

Answer the following questions:
What is the dependent variable in this equation?
What is the independent variables in this equation?
Is the equation linear or non-linear?
Is the equation ODE or PDE?


Example 2: A rod of length ‘L’ has its two ends at temperatures T1 and T2 , it loses heat
by convection at lateral surface as shown below:

Model equation of this system gives the following differential equation, which shows
temperature changes within the rod as a function of z:
d 2T
k = hA(T − Tair )
dz 2

a) What are the dependent/independent variables of this system?
b) Decide if this is initial or boundary value problem
c) How many initial or boundary conditions do we need to solve this equation, and
what would they be?

10
Mathematical Modelling for Chemical Engineers

Example 3: Calculate the values of x, y z in the following system of algebraic equations:

1.6 Classifications of the Model Equations
There are several ways of classifying model equations. Each leads to a variety of model
characteristics which have an impact on the solution techniques as well as the
potential application areas where they can be used. Some model types are
inappropriate in certain circumstances, i.e. a steady-state model for batch reactor
start-up analysis.

The section below explains some key aspects of the model classifications depending
on the various factors such as geometry, state, variables etc.:

Based on the State of the System

Model equations can be steady state or transient depending on the time variation
within the system.

Steady-State Model: Quantities of the variables are not changing with respect to time,

⎛ ∂ 2T ∂ 2T ∂ 2T ⎞ Steady-state system
k ⎜⎜ 2 + 2 + 2 ⎟⎟ = 0
⎝ ∂x ∂y ∂z ⎠

Unsteady-state (Transient or Dynamic) Model: Quantities of the variables are

changing with respect to time, i.e.

⎛ ∂ 2T ∂ 2 T ∂ 2T ⎞ ∂T
k ⎜⎜ 2 + 2 + 2 ⎟⎟ = ρCv
⎝ ∂x ∂y ∂z ⎠ ∂t

11
Mathematical Modelling for Chemical Engineers

Based on the Geometry of the System

Model equations can be one, two or three-dimensional differential equations depending

on the nature of the physical phenomena concern within the geometry of the system
modelled:
One-dimensional model: e.g. heat conduction in a long and thin slab as only
conduction in axial direction (z) and convection at the axial edges of the slab are
considered to be dominant physical phenomena hence, the system can be described
by the following model equation, considering only one spatial position (axial direction,
z)

d 2T
k = hA( T − Ta )
dz 2 z



Two-dimensional model: e.g. heat conduction in a cylindrical bar if the variation in
heat transport is considered in two spatial positions: radial and axial directions.

δ 2T 1 δ ⎛⎜ δ T ⎞⎟ r
k + rk =0
δ z2 r δ r ⎜⎝ δ r ⎟⎠
z

Three-dimensional model: e.g. the variations in heat conduction in three spatial

positions:

⎛ δ 2T δ 2T δ 2T ⎞ y
k⎜ + + ⎟=0
⎜ δ x2 δ y2 δ z2 ⎟
⎝ ⎠
x
z

12
Mathematical Modelling for Chemical Engineers

Based on the Physical Phases Involved within the System:

Model equations can be constructed on one phase or multi-phase systems:
Homogeneous model: if there is only gas or liquid or solid phase in the system.
Heterogeneous model: if there is more than one phase in the system, i.e. gas-
liquid-solid catalyst systems.
Pseudo homogeneous model: solving multi-phase, heterogeneous system as if
there were only one phase in the system.

Based on the Number of Independent Variables:

Ordinary Differential Equations (ODE): Derivatives arise with respect to one

independent variables only.

Changes in ‘T’ depends on ‘z’ only,

d 2T T = f (z)
k = h( T − Ta )
dz 2

Partial Differential Equations (PDE): Derivatives arise with respect to more than one
independent variables.

⎛ ∂ 2T ∂ 2T ⎞ ∂T Changes in ‘T’ depends on

k ⎜⎜ 2 + 2 ⎟⎟ = x & y directions and time, t,
⎝ ∂x ∂y ⎠ ∂t T = f (x, y, t )

Based on the Order of the Differential Equations:

Order of the differential equation is the number of times a dependent variable has
been differentiated.

dx First order (with respect to t, time)

=v
dt

d 2T Second order (with respect to z, length)

k 2 =q
dz
Please note that order of the differential equations indicates how many boundary
conditions required to solve the system, i.e. a second order differential equation
requires two boundary conditions to solve the equation.

1.7 Principles of Model Building

The development of a mathematical model that incorporates the basic phenomena
occurring in the process requires a lot of skill, ingenuity and practice. Since, in reality,
most physical phenomena are too complicated to create their exact model, an

13
Mathematical Modelling for Chemical Engineers

engineering compromise between a rigorous description and getting an answer that is

good enough is always required which may involve making as many simplifying
assumptions as are reasonable and make sense. Therefore, the most vital role that
modellers play is in exercising their judgement as to what assumptions can be validly
made. Basically, good modellers should not only have a deep understanding of the
physical phenomena, mathematical knowledge, but also an intuition, imagination and
curiosity.

Mathematical modelling is a principled activity that has both physics behind it and
methods should be successfully applied. The methodological modelling principles can
be captured in the following list below which are key for building a good mathematical
model:

• Why Why do we the modelling? (Identify what we are looking for)
• What What do we want to know? (List the data we are seeking)
• How How should we look at the model? (Identify the governing
physical principles, and the equations)
• Have What do we know? (Identify the relevant data available)
• Assume What can we assume? (Identify the circumstances that apply)
• Predict What will our model predict?
• Verified Are the predictions good? (Identify tests that can be made
to verify the model)
• Validate Are the predictions valid? (Identify tests that can be made
to validate the model)

• Improve Can we improve the model?

(Identify parameters/values that are not adequately known,
variables that should have been included and/or
assumptions/restrictions that could be lifted)

• Usage How will we apply the model? What will we do with it?

Please note that ‘validation and verification of the model’ against empirical
observations is very important part of modelling process in order to ensure that the
model is reliable and consistent.

Also remember that the computational effort required for the solution should be
reasonable as the model to be developed

• should not be more detailed than absolutely necessary!

Everything should be made as simple as possible but not simpler.

Albert Einstein

14
Mathematical Modelling for Chemical Engineers

• should contain as few parameters as possible!

How many things I can do without.

Socrates

• should have reliable correlations for the parameters!

There are three kind of lies: lies, damned lies and statistics.
Benjamin Disraeli



1.8 Techniques of Building a Mathematical Model
Like other engineering tasks, good practice requires mathematical models to be
constructed following some modelling techniques which basically help answering the
philosophical questions posed above. The techniques and mathematical principles
include: dimensional homogeneity, abstraction, scaling, conservation and balance
principles which are briefly explained below:

• Scaling
The notion of scaling includes several ideas, the effects of geometry on scale
(macro or micro), the relationship of function to scale, and the role of size in
determining limits—all of which are needed to choose the right scale for a model
in relation to the reality we want to capture.

• Conservation and Balance Principles
In developing mathematical models, balance or conservation principles should be
applied to assess the effect of maintaining or conserving levels of important
physical properties. This is based on the fundamental principle of physics that
mass, energy and momentum are neither destroyed nor created but simply
change their form.

• Abstraction
An important decision in modelling is choosing an appropriate level of detail for
the problem considered. This process is called abstraction and it typically requires
a thoughtful approach to identifying those phenomena on which we want to
focus, that is, to answering the fundamental question about why a model is being
sought or developed.

• Dimensional Homogeneity and Consistency
Every equation derived in the modelling of a process, should be dimensionally
homogeneous or dimensionally consistent.

15
Mathematical Modelling for Chemical Engineers

The development of a model is inherently iterative in its nature hence using a

hierarchic step by step procedure can be necessary. Firstly, the definition of the
system to be modelled should be stated very clearly, and the modelling goals should
be fully identified correctly, as well as the geometry, boundary and the system
variables. The following 8-step procedure for developing a model is good guideline to
create a ‘sound’ model specially with dealing complicated processes.

(1) Define the problem: start with drawing a sketch of the system to be
modelled, and labelling the various geometric, physical and chemical
quantities on the sketch would help to identify the parameters clearly.
(2) Identify the key controlling mechanisms, i.e., mass (i.e diffusion, convection)
heat (i.e. conduction, radiation) or momentum transport, with a chemical
reaction etc.
(3) Identify the characterising variables, i.e. dependent and independent
variables.
(4) Establish balance region: choosing a ‘control volume’ for a lumped or
differential element of the system to be modelled, (draw it on the sketch
and indicate all inflow-outflow paths).
(5) Define the ‘conservation principles’ within the control volume: formulate
the model equations, based on the principles of mass, energy or momentum
transport phenomena, and introduce specific forms of the terms (i.e. flux,
reaction rate).
(6) Introduce necessary justifiable and simplifying assumptions.
(7) Identify all the possibilities for the boundary values of the dependent
variables.
(8) Select an appropriate solution method and implement it to solve the
equations.

Test Yourself: Please Complete Modelling Problem Sheet 1 in your own time before the
next session

16
Mathematical Modelling for Chemical Engineers

Mathematical Modelling Problem Sheet 1:

Understanding Mathematical Modelling Terms

1. How many initial values are required to solve the following simultaneous set of
equations?
dy
= x 2 − y 2 + 3z (1)
dt
dz
= y − 2z + x (2)
dt

x = 5z 2 − y + 6 (3)

2. Consider all the equations given below and answer the following questions:
(Please note that x, y, z, r are the coordinates, t is time)

(a) Are the equations ODE or PDE?

(b) What are the orders of the equations?
(c) Are they transient or steady state?
(d) Are they linear or non-linear?
(e) What are the dependent and independent variables?
(f) What are the dimensions of the equations (i.e. one-di, two-di)?
(g) How many initial and/or boundary conditions are required to solve each
equation?
(h) If each represents a model equation of a system, what would you say about
the coordinates of the system modelled?

du k (1)
=− u u+B
dt m

d 2T dT (2)
a −b = 5T 2
dx 2 dx

∂ 2C D ∂ ⎛ ∂C ⎞ (3)
D + ⎜r ⎟=0
∂z 2 r ∂r ⎝ ∂r ⎠

∂T ⎛ ∂ 2T ∂ 2T ∂ 2T ⎞ (4)
ρ Cp = k⎜ 2 + 2 + 2 ⎟
⎜ ∂x
∂t ⎝ ∂y ∂z ⎟⎠

17
Mathematical Modelling for Chemical Engineers

2 Formulation of Heat Transport Phenomena

Intended learning outcomes (ILO’s) of this session:

• Setting up a general heat transfer model equation
• Understanding the difference between heat flow and heat flux
• Performing Heat Balance on the plane wall
• Estimating temperature profile on a plane wall
• Understanding various boundary conditions for the heat transfer form a plane wall.
• Performing Heat Balance on flowing heated fluid systems

Can ‘Heat’ be defined as ‘Energy’?

The first law of thermodynamics states that during an interaction between a system
and its surroundings, the amount of energy gained by the system must be exactly
equal to the amount of energy lost by the surroundings: energy is neither created nor
destroyed, but energy can be transferred from one system to another.

Heat is energy in transit due to a temperature difference. When a temperature

gradient exists in a stationary medium (solid or fluid) the term ‘conduction’ is used to
refer to the heat transfer that will occur across the medium. Where else, the term
‘convection’ refers to heat transfer that will occur between a surface and a moving
fluid when they are at different temperatures.

Energy versus Heat
Energy is a property, but heat is not.
Energy is associated with a state; heat is associated with a process.
Heat is defined as form of energy that is transferred between two systems by virtue of
a temperature difference:
‘Heat’ is an ‘Energy in Transition’ (also called Thermal Energy) which is recognised only
as it crosses the boundary of a system.
Hence ‘a human body would contain energy but not heat’.

Application of Heat Transfer Phenomena in Chemical Engineering;
In chemical engineering processes i.e. heat exchangers, distillation columns, reactors,
chemical engineers are always interested in how to predict rates of heat transfer.
Therefore, in all the processes where there is a temperature change, to find out
information such as temperature distribution, the amount of heat transferred, etc. we
must formulate an energy balance within the system studied.

18
Mathematical Modelling for Chemical Engineers

[Nomenclature used in the formulation of Heat Transport Phenomena - SI unit-]

Revision -List of Some Energy Balance Terms-

(Connect the left hand side definitions with the right hand side formula)

Fourier’s conduction law ρvC P
dT
dt

Accumulation rate of thermal energy hAs (T1 − T2 )

Bulk convection dT
−k
dz
Interphase heat transfer
( 4
εσ TSystem 4
− TSurr )

Heat of reaction
ρvC p ΔT

Heat transfer by radiation (− ri )(ΔΗ R )

Setting Up a General Energy (Heat) Balance Equation

In this unit, since we are only interested in Energy Flow in terms of ‘ Heat’, therefore,
we will be using the following equation for a heat balance within a system studied:

⎧rate of flow ⎫ ⎧⎪ rate of flow ⎫⎪ ⎧⎪ generation or consumption ⎪⎫ ⎧⎪ rate of accumulati on ⎫⎪
⎪ ⎪
⎨ ⎬ −⎨ ⎬ ±⎨ ⎬=⎨ ⎬
⎪⎩of heat ⎪⎭ in ⎪⎩of heat ⎪⎭ out ⎪⎩ of heat within system ⎪⎭ ⎪⎩of heat within
19 system ⎪⎭
 Mathematical Modelling for Chemical Engineers

∂T
(
Qin − Qout ± Qgen / cons = V(ρC P ) ) ∂t

W [=] J s -1 ( )( )(
m 3 kg m-3 J kg -1 K -1 K s −1 )

Q: heat flow
Q = m! C p ( ΔT ) J s-1 =W

m! [=] kg s -1
C P [=] J kg -1 K -1


q, heat flux
q = ρvC p ( ΔT )

v [=] m s -1
ρ [=] kg m-3
ρv [=] kg m- 2 s -1
[J ] =[ kg m2s-2]
J s-1m-2=W m-2



Heat Flow versus Heat Flux

General Heat Balance Equation in terms of Heat Flux Q = q(Ac )

⎧ generation or consumption⎫ ⎧ rate of accumulation ⎫
(h eat flux)in Ac − (heat flux)out Ac ± ⎪⎨ of heat within system ⎪⎬ = ⎪⎨of heat within system ⎪⎬
⎪⎩ ⎪⎭ ⎪⎩ ⎪⎭

20
Mathematical Modelling for Chemical Engineers

(qin Ac ) − (qout Ac ) ± ( Qgen / cons ) = V (ρCP ) ∂T
∂t

J s -1 m −2 m 2 ( )( )(
m 3 kg m-3 J kg -1K -1 K s −1 )

Ac is the cross sectional area, normal to the flow.

2.1 Heat Transfer by Conduction on a Plane Wall (as one-di, ODE systems)

Conduction is an electronic/atomic mechanism of transferring energy from one place
to another in solids, and a molecular mechanism of heat transfer in liquids and gases.
The section below summarises the formulation of heat transfer on a plane wall in one
direction and explains the related possible boundary conditions.

In-class Exercise 1: Consider that the edges of a wall shown below are insulated with
the temperatures of To and TL in which To>TL then the heat will flow normal to the
surface, that is in z direction. (Assuming that w and H is much smaller than L) . The
area normal to the heat flow is AC.

We want to know the steady state temperature profile on the axial direction of the
wall (from 0 to L):
For that, we need to formulate a model equation within the axial length of the wall
which describes the heat transfer on the wall.

We take a segment on the wall (Δz) and perform heat balance on the segment Δz.
Control volume of this increment is (Ac Δz).

21
Mathematical Modelling for Chemical Engineers

{Heat flow in } z − {Heat flow out }z + Δz = 0

Qz − Qz + Δz = 0

q z ( Ac ) − q z +Δz ( Ac ) = 0

Dividing the equation by the control volume, AC Δz , gives:



and taking the limit as Δz approaches zero, gives:



qz (heat flux on the solid wall) is described by Conduction –Fourier’s Law- as:

dT
q z = −k
dz

replacing the conduction heat flux on the model equation above, gives:





If conduction coefficient is not function of z, then it can be written as:




This is a second order ODE which requires two B.Cs. to solve, as:


B.C 1: at z=0

B.C 2: at z=L

22
Mathematical Modelling for Chemical Engineers

What is the analytical solution to this problem?

d ⎛ dT ⎞
⎜ ⎟=0
⎜ ⎟
dz ⎝ dz ⎠






From B.C.1:
→ To = C 2


From B.C.2: TL = C1L + C2



TL = C1L + T0


T −T
C1 = L 0
L
Hence the analytical solution gives the temperature profile on a plane wall in axial
direction is which can be expressed as follows:





This tells us that the temperature profile in the slab must be a straight line as shown in
the figure below:

23
Mathematical Modelling for Chemical Engineers

2.2 Possible Boundary Conditions on a Plane Wall

24
Mathematical Modelling for Chemical Engineers

2.3 Energy Balance on the Surface of a Medium (Lumped Model)

If the conservation of energy requirement needs to be applied at the surface of a

medium, under steady state conditions, the control surface do not include mass , or
volume, hence the steady state surface energy balance can be written as follows:

qcond − qconv − qradn = 0

qcond = qconv + qradn

25
Mathematical Modelling for Chemical Engineers

k
(T1 − T2 ) = h( T2 − T∞ ) + εσ( T2 4 − T∞4 )
L

In-class Exercise 2: The hot combustion gases of a furnace are separated from the
ambient air by a brick wall 0.15m thick with a thermal conductivity of 1.2 W/m/K and a
surface emissivity of 0.8. Under steady state conditions the outer surface temperature
of 100oC is measured. Free convection heat transfer to the air adjoining the surface is
characterised by a convection coefficient of h = 20 W m-2 K-1. Calculate the brick inner
temperature.
σ = 5..67 x10 −8 Wm −2 K −4

A surface energy equation, in a steady state condition gives: qcond − qconv − qradn = 0
k
(T1 − T2 ) = h( T2 − T∞ ) + εσ( T2 4 − T∞4 )
L

1.2

0.15
( )
(T1 − 373) = 20.( 373 − 298 ) + 0.8 5.67 x10−8 ( 3734 − 2984 )

Test Yourself Please Complete Modelling Problem Sheet 2 in your own time, before
the next lecture.

26
Mathematical Modelling for Chemical Engineers

Mathematical Modelling Problem Sheet 2:

Steady State Conduction

1) Consider a room where three sides perfectly insulated, and one side is
covered with a double-glazed window. The equivalent thickness of the
window is given as 20 mm, (see below a schematic cross-sectional
drawing). Heat transfer coefficient of air is 18 W m-2 K-1, and equivalent
conductivity of the glass is 0.93 Wm-1 K-1 .

If the temperature inside the room is -somehow- kept constant at 25oC, and
outside air temperature is 5oC, what would be the temperature of the
window facing outside? (Assume that the heat flux is at steady state
condition.)

Answer: T = 292K = 19.4oC

2) Considering each of the following cases given below, solve the model
equations analytically, using the boundary conditions provided to obtain the
temperature profile equation for each case:

a) A plain wall where there is a heat generation (qgen, J/s/m3 ) within the wall,
and symmetrical boundary conditions can be applied:

27
Mathematical Modelling for Chemical Engineers

Ts1 Ts ,1 = Ts ,2 = Ts
Ts ,2

⎛ d 2T ( z ) ⎞
The appropriate heat equation: ⎜k ⎟ + q gen = 0
⎜ dz 2 ⎟
⎝ ⎠

Answer:

q gen L2 ⎛ z 2 ⎞
T( z ) = ⎜1 − ⎟ + Ts
2k ⎜⎝ L2 ⎟⎠

b) A cylindrical system where there is a conductive heat transfers in radial

direction:

Ts ,2

1 d ⎛ dT ( r ) ⎞
The appropriate heat equation: ⎜r ⎟=0
r dr ⎝ dr ⎠

Answer:
Ts ,1 − Ts ,2 ⎛r⎞
T( r ) = ln⎜⎜ ⎟⎟ + Ts ,2
⎛r ⎞ r
ln⎜⎜ 1 ⎟⎟ ⎝ 2 ⎠
⎝ r2 ⎠

28
Mathematical Modelling for Chemical Engineers

2.4 Heat Transfer in a Flowing Fluid System

In a system where there is a heated fluid flowing e.g. in a pipe or in a tubular reactor,
the heat transfer occurs through bulk convection due to the heat capacity of fluid and
through conduction due to conductivity characteristic of the fluid itself.

qconv=ρvCpΔTfluid


In-class exercise 3: Heated fluid ‘A’ enters in an insulated cylindrical reactor, and
reacts to form ‘B’. Assume that there is a plug flow velocity distribution and a perfect
mixing in the radial direction, derive a steady state heat balance equation which gives
temperature distribution of fluid ‘A’ along the reactor as shown in the sketch below:

Solution:
Heat balance on increment ‘Δz’

(qin Ac ) − (qout Ac ) − (− rA )(ΔH )R V = V (ρCP ) ∂T
∂t


(q) z AC − (q)z+Δz AC − (− rA )(ΔH R )(AC Δz ) = 0

Dividing by the equation with the ‘control volume’ , ‘AC Δz’ , gives:

29
Mathematical Modelling for Chemical Engineers

Heat flux within the fluid, qz can be both through bulk convection and conduction,
Therefore, replacing these terms (given above) into the equation gives:

d ⎛ dT f ⎞
− ⎜− k f + ρvC p dT T ⎟f − (− rA )(ΔΗ R ) = 0
f ⎟
⎜
dz ⎝ dz ⎠

d 2T f dT f
kf 2
− ρvC p − (− rA )(ΔΗ R ) = 0
dz dz

Where kf is the conduction of the fluid.
This is a second order ODE, hence two boundary conditions are required to solve it:

Since for most fluids bulk convection term is more significant than conduction term:

⎛
⎜ d 2T f ⎞⎟ ⎛ dT f ⎞
⎜⎜ k f << ⎜ ρvC p ⎟ ⇒
2 ⎟⎟ ⎜ dz ⎟
⎝ dz ⎠ ⎝ ⎠


Therefore, we can ignore the conduction term, which then gives:

dT f
− ρvC p − (− rA )(ΔΗ R ) = 0
dz
To solve this equation, we only need one boundary condition, i.e. at z=0, T=T0.

30
Mathematical Modelling for Chemical Engineers

Test Yourself : Please Complete Modelling Problem Sheet 3 in your own time, before the
next lecture.

Mathematical Modelling Problem Sheet 3:

Heated Fluid

1. A fluid with a temperature of 25oC enters in a cylindrical tube where the

wall of the tube is externally heated at a constant temperature of 95oC. The
length of the tube is 100 cm and the diameter is 5 cm. The velocity, density
and viscosity of the fluid are 0.75 m s-1, 1.05 kg m-3 and 1.30x10-5 Pa s ,
respectively. Derive a differential model equation, using that you can
estimate the temperature profile of the fluid within the tube and the time
period which takes for the system to reach to steady state.

a) Sketch the system; show the control volume on the sketch.

b) Formulate an appropriate differential model equation within the control
volume. State and discuss your assumptions.
c) Express appropriate initial and/or boundary condition(s).
d) If the velocity of the fluid reduced to 0.3 m s-1, how this would affect the
solution method? Formulate the model equation and discuss your
boundary conditions.

2. Boiling water in a 2400 watts rated electric kettle:

Consider that your system is the boiling water inside the electric kettle
a) Identify the key transport processes which occur in your system,
b) Choose the most dominant controlling phenomena,
c) Identify all the relevant dependent & independent variables of the
dominant phenomena if you represent this phenomena with a differential
equation.
d) And finally, calculate how long it would take 1 litre of water to boil
assuming that the kettle was well insulated and the initial water
temperature was 15oC.

31
Mathematical Modelling for Chemical Engineers

3. Formulation of Mass Transfer Phenomena with a Diffusional Flow

(two-di, 2nd order )

Intended Learning Outcomes of this section:
Understanding the difference between
• Molar Flow, FA , versus Molar Flux, NA
• Diffusional versus Convectional Mass Transfer
• Performing Material Balance Equation and applying it in a chemical
engineering processes

Nomenclature used in Material Balance Formulations in SI units

32
Mathematical Modelling for Chemical Engineers

Mathematical representation of some of the terms used in ‘Material Transfer’:

(Connect the left hand side definitions with the right hand side formula)

33
Mathematical Modelling for Chemical Engineers

In most cases, mass transfer occurs due to both: convective bulk fluid and molecular
diffusion motion. Therefore, the description of mass -or molar- flux, includes both
bulk convection and diffusion terms as given below :
Molar flux of species ‘ i ’ in z direction

dC A Where;
N A z = −D + vC A
dz D is the bulk diffusion,
mol
2
m s C is the concentration,
Convective
v is the velocity,
Diffusional Flow
Bulk Flow
(Fick’s Law) z is the axial direction,
i represents the species in the
fluid

34
Mathematical Modelling for Chemical Engineers

Discussion on Molar Flow versus Molar Flux on a tubular reactor:
So far, in any plug flow reactor example, we used molar flow term for the inlet and
outlet streams as:

F
Aout


However, a model equation in terms of FA only represents, mass transfer through bulk
convection, and would not consider the molecular diffusion. This is acceptable for a
plug flow reactor model, since the key assumptions are; fast flow system, perfect
mixing and no agitation, hence we can ignore the diffusion term. However, if the fluid
flows in a laminar region the effect of diffusion will be significant and these
assumptions no longer will be valid. Therefore, during the construction of a mass
transfer balance equation, using ‘molar- flux ’ term, N, instead of ‘molar flow’, term,
F , would be more appropriate.

35
Mathematical Modelling for Chemical Engineers

3.1 Diffusion versus Bulk Convection
Although mass is transferred both by diffusional flux and convection, in most cases
one phenomenon would be more significant than other. There are some guide lines to
help you to decide which one would be more dominant to add into your model:

Let’s consider liquid ‘B’ , with a boiling point of 70oC, kept in a large reservoir. A
capillary tube is connected on the reservoir where the liquid might evaporate and
move through the capillary into the surrounding air at various temperatures.

Below we discuss how the movement of the liquid through capillary might change at
different air temperatures i.e. from 10 to 80oC.

In-class exercise 4: Liquid benzene (B.P=80.1oC), stocked in a reservoir. Create a
model to solve transient concentration distribution of benzene which evaporates and
moves out
through the capillary (as sketched below) into the surrounding room at the following
three different temperatures:
(a) T=6oC

36
Mathematical Modelling for Chemical Engineers

(b) T=60°C
(c) T=82°C
Solution:

⎧Molar flow of ⎫ ⎧Molar flow of ⎫ ⎧Mole of species B ⎫ ⎧Mole of species B⎫
⎨ ⎬−⎨ ⎬±⎨ ⎬=⎨ ⎬
⎩species B at z ⎭ ⎩species B at z + Δ z ⎭ ⎩produced/ consumed ⎭ ⎩accumulating ⎭
FB = vz AcCB
FB = Ac N B

VΔCB
Ac N B z − Ac N B =
z + Δz
Δt
ΔCB V = Ac Δz
Ac N B z − Ac N B = Ac Δz
z + Δz
Δt

NB z − NB ΔCB
z + Δz
=
Δz Δt
Taking the limit when both Δt and Δz approach to zero, gives:

∂N B ∂C
− = B
∂z ∂t


(a) for T=6°C

at this temperature mass transfer would be mostly through diffusion of ‘Benzene’.

dC
N B z = −D B
dz

Replace NBz into the model equation, gives:

37
Mathematical Modelling for Chemical Engineers

∂ 2C B ∂C B
D =
2
∂z ∂t


(b) for T=60°C


at this temperature both diffusional and convectional mass transfer would be

significant!

dC B
N B z = −D + vz CB
dz
Replace NBz into the model equation, gives:


∂ 2C B ∂C ∂C
D 2
− vz = B
∂z ∂z ∂t

(c ) for T=82°C
at this temperature mass transfer through bulk convection would be more significant!

N B z = vz CB


Replace NBz into the model equation, gives:

∂C B ∂C
− vz = B
∂z ∂t






Test Yourself : Please Complete Modelling Problem Sheet 4 in your own time, before the
next lecture.

38
 Mathematical Modelling for Chemical Engineers

Mathematical Modelling Problem Sheet 4:

Mass Transfer and Diffusion

1) A sphere shape sparingly soluble material ‘B’ (assume that the sphere’s
size does not change much during the dissolution process), quickly
dissolves in the surrounding solvent, so that the solute’s concentration at the
sphere’s surface is saturated. Because the sphere is immersed in a large
fluid volume, the concentration far from the sphere can be taken as zero.
We would like to estimate both the dissolution rate and the concentration
profile around the sphere.
Starting from a material balance of species ‘B’, derive a model equation for
dissolution of a spherical particle as described below:

Ro =radius of sphere
r
B Solute flux away

Cross sectional area:

Acr = 4πr 2
Control Volume:

Acr Δr = 4πr 2 Δr

2) Species ‘A’ diffuses into a rectangular sponge as shown in the sketch below.
Set up a mass transport model in rectangular coordinates that will produce a
concentration distribution of species ‘A’ within the slab under unsteady-state
conditions.

y
NAz
x
V z

NAx

NAy

39
Mathematical Modelling for Chemical Engineers

3.3 Mass Transfer in Heterogeneous Systems with Catalytic Reactions

Intended learning outcomes of this section:
• Understanding Internal and External Mass Transfer Resistances
• Learning Transfer Steps in a Catalytic Reaction System
• Understanding the Concept of ‘Effectiveness Factor’


Packed-bed reactors, with porous catalyst particles are widely used in the chemical
industry. The purpose of a packed bed is typically to improve contact between two
phases in a process. A common design is a cylindrical column filled with catalyst
coated packing materials. The packing can be ‘randomly filled’ or ‘structured packing’
(see the figure below)

40
Mathematical Modelling for Chemical Engineers

In some systems, the catalyst pellets can be fluidized by the action of a gas or liquid
flowing through a reactor.
The fluid enters at the bottom of a bed of catalyst particles, and the particles are
fluidized or made to “float” by the shear force that the fluid exerts on their surface. At
the same time, species dissolved in the fluid react on the surface and inside the porous
catalyst particles.
Schematic Representation of External Mass Transfer Resistances and Diffusion

3.3.1 Mass Transfer Steps in a Catalytic Reaction System

The sketch below illustrates the mass transfer steps for a system where a bulk gas
flowing on a catalytic surface, species ‘A’ in the gas diffuses inside a pore on the
catalyst, where it reacts chemically to produce species ‘B’ which travels back from the
pore to the bulk gas.

41
Mathematical Modelling for Chemical Engineers

Step 1: External mass transfer from bulk to the catalyst surface
Step 2: Diffusion into the pores from surface
Step 3: Catalytic reaction on the pore walls of the pellet
Step 4: Diffusion back out of the pores to the surface
Step 5: External mass transfer back to bulk from the catalyst
surface

42
Mathematical Modelling for Chemical Engineers

3.3.2 External Mass Transfer (Step 1 and 5)

From the result of concentration differences of the species ‘mass transfer’ occurs
across an interface and the rate of this external mass transfer of a given species
depends on some of the physical properties of the system and on the degree of
turbulence of the phases involved.
However, these parameters are not easily quantified, therefore mass transfer
coefficients have been defined that lump them all together within the hypothetical
uniformly thick fictitious layer around the catalyst as formulated below:

2
Molecular diffusivity, m /s
2
Average eddy diffusivity, m /s

Diff + ε M
NA = − (C A,b − C A,s )
Δz

43
Mathematical Modelling for Chemical Engineers

Thickness of the fictitious layer

or we can write it as:
N A z = −k m (C A,b − C A,s )


Concentration Concentration of ‘A’
of ‘A’ in the bulk flow on the catalyst surface


Where km (m/s) is described as the mass transfer coefficient within the fictitious layer
around the catalyst.

Often it is represented as

(
N Az = km C As − C Ab

)
mol m mol
2
[=]
s m3
m s



3.3.3 Diffusional Mass Transfer (Steps 2 and 4)

The diffusional mass flux can be described as:


dC A
N A = −DK
dς

Where
DK is the Knudsen Diffusion (diffusion into the pores) and
ζ is the pore length or catalyst radius

Effectiveness Factor:

Most industrial reactions are carried by contacting fluids with pellets of porous solid
catalysts, the reaction rates of these catalysts are expected to be proportional to the
total surface area available for the reaction, e.g. if we double the surface area, we
expect that the reaction rate will be doubled. However, on the contrary, in some
cases, the rate of reaction is observed to decrease as the size or amount of catalyst

44
Mathematical Modelling for Chemical Engineers

particle increases, despite of having larger external surface area on the catalyst fluid
interphase.

The reason of the decrease on the rate of reaction is the result of the slow diffusion
into the pores of the catalyst (interphase diffusion). However to calculate the flux of
this diffusional flow is rather difficult, requires big effort and time. Hence, for
simplification, we use a term called ‘Effectiveness Factor, η, ’ which represents how
much the reaction rate is decreased with the slow diffusion rate in the pores:

(actual reaction rate including diffusion )
η= 1 ≥η > 0
(
reaction rate as if diffusi on were ins tan tan eous )

If η =1 means that no effect of the diffusional resistance on the reaction rate.

If η is close to zero means that the diffusion into pores is very slow which reduces the
reaction rate on the catalyst significantly.

Modelling Challenge
When creating a computational model of a packed bed reactor that takes into account
the geometrical shape, we are faced with a major challenge:
A realistic reactor can contain hundreds of thousands of pellets of various sizes, and
each pellet contains countless microscale pores where those mass transfer steps and
reactions occur.
Modelling with a CFD software –COMSOL-

45
Mathematical Modelling for Chemical Engineers

3.3.4 Choosing the Rate Controlling Step

When we model such a complex phenomenon, we tend to simplify the system by

determining which step would be more dominant and which step can be ignored.

In the summary chart given below, some -rule of thumb- guidelines are introduced
which would help us for determining the rate controlling (or limiting) step, and
therefore simplifying the transfer mechanism in a complicated catalytically reacting
system:


Rough guidelines to determine the rate controlling steps

phenomenon rate controlling step

If the reaction is irreversible, Mass transfer from bulk
and flow dependent to the catalyst surface

If the reaction is irreversible,
Diffusion into the
and flow independent, but
catalyst pores
pellet size dependent

If the reaction rate varies sharply
with temperature, and is Catalytic reaction
independent of pellet size on the pore walls
of the catalyst

If the reaction is reversible,
Diffusion back out
flow independent, but
of the pores
pellet size dependent
46
Mathematical Modelling for Chemical Engineers

If the reaction is reversible, Mass transfer back to bulk
and flow dependent from the catalyst surface

In-class exercise 6:


For the case given below, decide which steps can become rate controlling steps:
Reactant A forms product B, in an irreversible, exothermic reaction on the surface of
the very small and non-porous pellets.

A# B −r = k C ks: Intrinsic rate constant
A s AS
Since it is irreversible reaction and the particles are nonporous and very small then at
steady state conditions the rate of mass transfer of A to the surface must be equal to
the rate of disappearance of A by reaction, therefore the mass transfer mechanism can
be written as:

External mass transfer and catalytic reaction: (steps 1 and 3)

k s C A,s = k m a m(C A,b − C A,s )

am: the external surface area of the
C = k m am C catalyst available for mass transfer
A,s A,b
1 + k m am per unit volume (1/m)

(− rA ) = 1
C A,b = k a C A,b ka: apparent rate constant (1/s)
1 1
+
k s k m am

In-class exercise 7: Calculation of Mass/Mole Transfer Rate

A large volume of pure water at 25oC is following parallel to a flat plate of solid benzoic
acid which dissolves into water. The solubility of benzoic acid in water (at the
interphase) is 0.02948 mol m-3, mass transfer coefficient, km, is 5.85x10-6 m s-1, and the
length of the plate is 0.25 m in the direction of flow. Calculate the molar flux of
benzoic acid into the pure water.

Solution

47
Mathematical Modelling for Chemical Engineers

C B ,s = 0.02948 molm −3

km = 5.85 x10 −6 ms −1 N Benz = km (CB,s − CB,water )


N Benz x10 −6 (0.02948 − 0)
= 5.85

N Benz = 1.726x10 −7 mol m−2 s −1


Various representation of mass transfer coefficients

For Gases

( ) (
N A = km C A1 − C A2 = kG PA1 − PA2 = k y y A1 − y A2

) ( )

For Liquids

N =k C
A m A1
( −C A2) = k (C L A1 −C ) = k (x
A2 x A1 −x )
A2

Test Yourself: Please Complete Modelling Problem Sheet 5 in your own time, before the
next lecture.

48
Mathematical Modelling for Chemical Engineers

Mathematical Modelling Problem Sheet 5:

Mass Transfer in Catalytic Systems

1. Consider the following first order reaction on catalytic spherical pellets:

A#B (− rA ) =k sC A

The intrinsic rate constant is given as 1.2 m3 kg-1 s-1 in terms of per kg of
catalyst. The mass transfer coefficient is 0.96 m s-1 between the bulk gas and the
4 mm diameter spherical pellets with smooth non-porous surface. The catalyst
density is given as 2000 kg m-3.
(a) Decide which mass transfer steps would be more dominant.
(b) Compute the apparent reaction rate constant (s-1)
(c) Compute the intrinsic rate constant in terms of volume of catalyst (m3 s-1).

Answers: 900 1/s , 8.0x10-5 m3/s

2. Initially pure water is passed through a tube constructed of solid benzoic acid.
Since benzoic acid is slightly soluble in water (denote solubility as C* moles
acid/cm3 solution), the inner walls of the tube will dissolve very slowly. By
weighing the dried tube before and after exposure, it is possible to calculate the
rate of mass transfer. Assuming that water flows in the turbulent flow region,
derive a steady sate material balance equation for the benzoic acid in the tube,
and choose all the boundary conditions required to solve the model equation.

Answer:
dCB 4
−v + km ⎛⎜ C* − C ⎞⎟ = 0
dz D ⎝ B B ⎠

49
Mathematical Modelling for Chemical Engineers

4. Formulation of Heat Transfer from Extended Surfaces

(Convective Heat Loss from Hot Solid Surfaces) (Boundary Value
Problems)


Intended learning outcomes (ILO’s) of this section:
• Setting up a heat transfer model on an extended surface
• Understanding the significance of simplification a complex system
• Understanding the solution of a Boundary Value Problem
• Application of Finite Difference Method

The term extended surface is commonly used in reference to a hot solid where an
energy transfer by conduction occur within its boundaries, while heat is transferred by
convection (and or radiation) between its boundaries and surroundings.
Examples of such a system sketched below:
Tair qconvection = hAs (Ts − Tair )

Ts

In-class exercise 7: Heat conduction in a cooling fin: (Bird et.al., 1960 )
Set up a model equation to estimate the temperature profile in the axial direction of
the cooling fin shown in the figure below, and write all the boundary conditions
required to solve the equation.

o
T =15 C
air

AC = 2BW

B is much smaller than L

50
Mathematical Modelling for Chemical Engineers

When a system to be modelled is too complicated, a reasonably good description of

the system may be obtained by approximating the true physical situation by a
simplified model (or a set of assumptions): i.e.

Real phenomena Simplified model

⎧Heat flow in ⎫ ⎧Heat flow out ⎫ ⎧⎪Heat loss ⎫⎪
⎨ ⎬−⎨ ⎬−⎨ ⎬=0
⎩by conduction ⎭ ⎩by conduction ⎭ ⎪⎩from fin to air ⎪⎭

Q z − Qz+Δz − Qloss = 0
qz Ac − qz +Δz Ac − hAs ( T − Tair ) = 0

qz ( 2BW ) − qz +Δz ( 2BW ) − h( 2WΔz )( T − Tair ) = 0

Dividing this equation by the control volume (2BW∆z) and taking the limit as ∆z
approaches to zero gives:

q z ( 2 BW ) − q z + Δz ( 2 BW ) h( 2WΔz )( T − Tair )
− =0
2 BWΔz 2 BWΔz
dq z h
− − (T − Tair ) = 0
dz B
Replacing qz with Fourier’s Law

gives:
d 2T h
2
− (T − Tair ) = 0
dz kB
where k is the thermal conductivity of the fin and h is heat transfer coefficient
between solid wall and air.

51
Mathematical Modelling for Chemical Engineers

This equation can be solved with the following ‘Boundary Conditions’:

Please note that in the next computing tutorial, you will solve a similar system using
Matlab, applying the numerical solution techniques (i.e. finite difference method) that
you have learned in Numerical Analysis part of the course.























Test Yourself: Please Complete Modelling Problem Sheet 6 in your own time, before
the next lecture.

52
Mathematical Modelling for Chemical Engineers

Mathematical Modelling Problem Sheet 6:

Heat Transfer on an Extended Rod

1. A 30 cm long cylindrical rod, 5 cm in diameter has one end attached to a wall

which is maintained at 125oC and the other is insulated as shown in figure below.
The lateral surface of the rod is exposed to ambient air at 15oC with a heat
transfer coefficient of 110 W m-2 K-1.
(a) Set up a model equation which describes the heat transfer on the rod,
(b) Express all the assumptions to simplify the system,
(c) Choose related boundary conditions required for solution.
(d) Solve the model equation to estimate the temperature profile in axial
direction within the cooling rod for the following three materials:
i. Copper ( k=401 W/m/K) ,
ii. Aluminum (k= 205 W/m/K),
iii. Stainless Steel (k=16 W/m/K).

o
Tw=125 C

L=30 cm

o
Tair=15 C
y

53
Mathematical Modelling for Chemical Engineers

5 Modelling of Heat & Mass Transfers in Heterogeneous

Systems
Most mathematical models in chemical processes which involve reactions require
consideration of both the energy and material transports within the system. The
model equations describing material and energy changes should be solved
simultaneously since they share the reaction term in which temperature and
concentration of the species are variables.

Formulating of a Mathematical Model for a Catalytic Converter

Typical example of such systems is catalytic monolith reactors which are widely used
as pollution reduction devices. The most popular application of the monolith reactor
currently used is in the automobiles as a catalytic converter fitted at the end of
exhaust pipes.

Catalytic converters use a catalyst (in the form of platinum and palladium) to convert
three harmful compounds; unburned hydrocarbon (fuel) , carbon monoxide and
nitrogen oxides, into less harmful compound (i.e CO2, H2O, N2 and O2) .

The sketch below shows schematic drawing of a typical square channel monolith
reactor (A), with the zoom in images of the monolith cell (B) and the catalyst which is
coated on the wall of the cell (C):

54
Mathematical Modelling for Chemical Engineers

55
Mathematical Modelling for Chemical Engineers

In order to increase the external surface area of the catalyst, the washcoat is applied
on the surface of the wall of the monolith channels which creates a porous surface of
the wall and in some cases –depending on the size of the pores- causes an intraphase
and interphase diffusional mass transfer steps as shown on the schematic drawing
below:

56
Mathematical Modelling for Chemical Engineers

In-class exercise 8: Building a Model for a Catalytic Converter:

A monolithic catalytic converter is used at the end of the exhaust pipe in order to
eliminate the emission of CO from the car exhausts, and oxidizes CO to form CO2 with
the following reaction:
CO + 1/2O2 → CO2 ∆H = -283 kJ/mol

Reaction rate on the solid surface can be in the following form:

mol
( −rCO )s = k s CCO s CO 2 s 2
mcat s

Figure below shows a schematic drawing of such a catalytic converter

Formulate the material and heat balance model equations to estimate

(a) the concentration profile of CO, and

(b) the temperature profile along the axial length of the converter, both on the gas
and solid phases.

Solution: First of all we need to simplify the system with some assumptions:

Assumptions for Simplifications Implications on the model

The converter works under No accumulation terms!

steady state condition, dT ⎞
⎛ dC ⎞ ⎛
⎜V ⎟ ⎜ mCp ⎟
⎝ dt ⎠ ⎝ dt ⎠

Flow distribution is assumed Single channel model!
to be uniform,

57
Mathematical Modelling for Chemical Engineers

There is a complete mixing in the ‘r’ direction,
hence the flux in ‘r’ direction can be neglected,
⎛ dT ⎞ dCCO
⎜k ⎟ N CO r = − D + vr CCO
⎝ dr ⎠ dr

CCO
If the axial bulk gas velocity is sufficiently high N CO z = − D + v z CCO
dz
the diffusional mass flux, and axial conduction
⎛ dT ⎞
⎜ kz ⎟
can be neglected, ⎝ dz ⎠

Diffusional effect into the catalyst is represented by ‘effectiveness factor’, so that, the
interphase diffusional flow term will not be required!

Formulation of the Model Equations:
There are more than one phase (gas and solid) and reaction happens on the solid
surface, therefore the concentration on the gas phase would be different than on the
solid phase. Hence, there should be two material balance equations: one on the gas
phase and one on the solid phase.

(a) Material balance on the increment Δz of the monolith bed

• in the gas phase

58
Mathematical Modelling for Chemical Engineers

N CO z

(Ac N CO ) z − (Ac N CO ) z + Δz − ΔAS k m (C co,g )
− C co,s = 0

Dividing by ∆V (= AC∆z) , and taking limit when ∆z approaches to zero, gives

dN CO ,g ΔAS
−
dz
−
AC Δz
( )
k m C CO ,g − C CO ,s = 0

Where Ac is the cross sectional area, and AS is the surface area within the monolith cell
which is As = W.L

‘W’ is the perimeter in the monolith cell : W = π DH

DH is the hydraulic diameter

The increment of As would be:

ΔAS WΔz πDH 4

= = =

Ac Δz Ac Δz πDH2 / 4 DH ( )

dN CO,g 4
−
dz
−
DH
(
k m CCO,g − CCO,s = 0 )

N CO z = vCCO

dC
−v
CO,g
dz
−
4
DH
km C
CO,g
(
−C
CO,s
=0 )
59
Mathematical Modelling for Chemical Engineers

Required Boundary Condition: at z = 0 CCO = CCOin

• in the solid phase

ΔAs k m C ( CO,g
−C
CO,s
)− η(− r )
CO ,s s a p ΔV =0

η : effectiveness factor # represents the effect of the diffusional resistance

ap : external catalytic surface area per unit volume of the monolith bed

Dividing by ∆V (=ACΔz) , gives

ΔAs k m C( CO,g
−C
CO,s
) − η(− r ) = 0
CO ,s s a p
Ac Δz

ΔAS WΔz πDH 4
= = =
2
Ac Δz Ac Δz πDH / 4 DH ( )

4
DH
km C
CO,g
(−C
CO,s
)
− η(− rCO,s )s a p = 0

1 m ⎛ mol ⎞ mol 2
mcat
⎜ ⎟ 2
v mcat s 3
m s ⎝ m3 ⎠ mbed

4
DH
km C (
CO,g
−C
CO,s
) (
− η k s CCO,s CO2 ,s s a p = 0 )

60
Mathematical Modelling for Chemical Engineers

(b) Heat balance on the increment Δz of the monolith bed

• in the gas phase

(q z AC ) z − (q z AC )z + Δz − hΔAS (Tgas − Tsolid ) = 0

AS = WΔz
Dividing by Ac∆z and take limit when ∆z approaches to zero gives

dq z hW
− − ( Tgas − Tsolid ) = 0
dz Ac

Substitute convective bulk flow (we assumed that heat flux by conduction in the gas
phase would not be significant comparing the heat transfer by the bulk flow , hence

kJ
dq z = ρ vC p dTgas [ =]
m2 s

⎛ dTgas ⎞ hW
− ρvC p ⎜⎜ ⎟ A ( Tgas − Tsolid ) = 0
⎟−
⎝ dz ⎠ c

W π DH 4
= =
Ac (2
π DH / 4 DH )

⎛ dTgas ⎞ 4
− ρvC p ⎜⎜ (
⎟ − (− ΔΗ) kCCO g = h
⎟ )( Tgas − Tsolid )
⎝ dz ⎠ Dh

Boundary Condition:
at z = 0 Tg = Tg _ in

61
Mathematical Modelling for Chemical Engineers

• in the solid phase

(qz AC ) z − (qz AC )z + Δz − hAS ( TS ) + (− ΔΗ R )S η (− rCO ,s )s a p (Ac Δz ) = 0

− Tgas

Dividing by Ac∆z and take limit when ∆z approaches to zero gives

dqz hAS
−
dz
−
AC Δz
( )
( TS − Tgas ) + η − rCOS (− ΔΗ R )S a p = 0
dTS
q z = −k z
dz
d 2TS 4
kz 2
− h( TS − Tgas ) + η (− rCO )(− ΔH R )S a p = 0
dz Dh
Boundary Conditions:

If the ends are perfectly insulated

dTS
=0 at z = 0 and z = L
dz

If there is convective heat loss at the ends of the wall of the converter

⎛ dT ⎞
⎜⎜ k z S ⎟⎟ = h( TS - T∞ ) at z = 0 and z = L
⎝ dz ⎠ z = L

Test Yourself: Please Complete Modelling Problem Sheet 7 in your own time before the
next lecture.

62
Mathematical Modelling for Chemical Engineers

Mathematical Modelling Problem Sheet 7

Gas Turbines with Catalytic Combustion

1. You are working as a design engineer in a company developing catalysts for

various industrial applications. One area that the company is keen to invest is to
develop a high temperature catalyst for the catalytic oxidation of methane mainly
for the application of catalytic gas turbines (as seen in the Figure 1 below).

Figure 1: Sketch of the Conventional and Catalytic Gas Turbines

Catalytic oxidation which is flameless combustion occurs at lower

temperature than conventional combustion, therefore it results with lower
emission of harmful nitrogen oxide (NOx) gasses.
The one step catalytic combustion of methane reaction can be written as
follows:
CH 4 + 2O2 → CO2 + 2H 2O ΔHor = -803 kJ/mol

63
Mathematical Modelling for Chemical Engineers

The reaction rate expression of the catalytic oxidation of methane is given as

follows:

(− rCH )s = ksCCH CCO

4 4
2
2
−2 −1
mol mcat s

As a design engineer your task is to investigate the efficiency of one of the

company proprietary ‘palladium oxide, PDO’ catalysts for various catalytic
bed options. Your project team will decide, which of those configurations
shown below would be more efficient to use for the catalytic gas turbines
applications:

- Packed bed with spherical catalyst (Dpellet= 3 mm, ε = 0.45)

- Metallic foam with gauze structure coated with the catalyst (ε = 0.25)

- Square channel monolith coated with the catalyst (DH =1 mm, ε =

0.65)

Where ε is the void fraction of the catalyst bed and V is the volume:
Vvoid
ε=
Vvoid + Vsolid

Solution Hint:
To compare the efficiency of the configurations suggested, first of all, you need to
derive three different models for each of the system, then solve these models
separately to compare the results obtained, namely, their concentration and
temperature distributions within the catalyst bed.
However for this tutorial, I am not expecting you to solve the model equations,
your task is only to derive the equations correctly on the gas and solid phases for
each of the configurations.

64
Mathematical Modelling for Chemical Engineers

6. References and sources

The following references are used in preparation of these notes. Please note that
all the books used are available at the library of the University of Bath.

Bird B. R., Stewart W.E.,Ligthfoot E.N. (1960) Transport Phenomena, John Wiley & Sons, Inc.USA

Cussler E.L. (2009) Diffusion, Mass transfer in Fluid Systems, Cambridge University Press, UK

Comsol (2013) Introduction to COMSOL Multiphysics in Chemical Engineering Applications, USA

Dym L. Ivey E.S., (1980) Principles of Mathematical Modeling, 1st Edn, Academic Press, New
York

Elnashaie S.S.E.H and Elshishini S.S. (1993) Modelling, Simulation and Optimisation of Industrial
Fixed Bed Catalytic Reactors, Gordon and Breach Science Publishers, USA

Finlayson B. A. (2006) Introduction to Chemical Engineering Computing, Wiley, USA

Fogler H.S. (1986) Elements of Chemical Reaction Engineering, Prentice Hall International editions,
USA

Franks G.E.R. (1972) Modelling and Simulation in Chemical Engineering, Wiley-Interscience, USA

Geankoplis C.J. (1993) Transport Processes and Unit Operations, Prentice Hall

Jenson V.G., Jeffreys G.V. (1977) Mathematical Methods in Chemical Engineering, Acad. Press, USA

Hayes R.E., Kolaczkowski S.T.(1997) Introduction to Catalytic Combustion, Gordon &Breach

Science Publishers, Singapore

Incropera F.P., DeWitt D.P. (1996) Fundamentals of Heat and Mass Transfer, Wiley, USA

Law V.J. (2013) Numerical Methods for Chemical Engineers, CRC Press, USA

Ramirez W. F.(1997) Computational Methods for Process Simulation, Butterworth Heinemann, USA

Rice R.G., Do D.D. (1995) Applied Mathematics and Modelling for Chemical Engineers, John Wiley
and Sons, USA

Smith C.L., Pike R.W., Murrill P.W. (1970) Formulation and Optimization of Mathematical Models,
International Textbook Co, USA

65