CHME 403

Lecture 2: Models

Lecturer: Dr. Dhawal Shah

Chemical and Materials Engineering Department,

Nazarbayev University
Spring 2025
Modeling of chemical processes
 Learning objectives
 Formulate dynamic models based on fundamental
balances
 Understand why we need dynamic models

2 Dhawal Shah, 2024 22/1/2025

Control of continuous processes
 Tubular heat exchanger. A process fluid on the tube side is cooled by cooling water
on the shell side. Typically, the exit temperature of the process fluid is controlled by
manipulating the cooling water flow rate.

 Continuous stirred-tank reactor (CSTR). If the reaction is highly exothermic, it is

necessary to control the reactor temperature by manipulating the flow rate of coolant
in a jacket or cooling coil.

 Thermal cracking furnace. Crude oil is broken down (“cracked”) into a number of
lighter petroleum fractions by the heat transferred from a burning fuel/air mixture. The
furnace temperature and amount of excess air in the flue gas can be controlled by
manipulating the fuel flow rate and the fuel/air ratio.

 Kidney dialysis unit. Used to remove waste products from the blood of human
patients whose own kidneys are failing or have failed. The blood flow rate is
maintained by a pump, and “ambient conditions”

(3) Dhawal Shah 22/1/2025

Control of batch/semibatch processes
 Jacketed batch reactor. In a batch reactor, an initial charge (e.g., reactants and
catalyst) is placed in the reactor, agitated, and brought to the desired starting
conditions. For exothermic reactions, cooling jackets are used to keep the reactor
temperature at or near the desired set point.

 Semibatch bioreactor. For a semibatch reactor, one of the two alternative operations
is used: (i) a reactant is gradually added as the batch proceeds or (ii) a product stream
is withdrawn during the reaction. In order to better regulate the growth of the desired
microorganisms, a nutrient is slowly added in a predetermined manner.

 Semibatch digester in a pulp mill. Both continuous and semibatch digesters are
used in paper manufacturing to break down wood chips in order to extract the
cellulosic fibers. The endpoint of the chemical reaction is indicated by the kappa
number, a measure of lignin content. It is controlled to a desired value by adjusting the
digester temperature, pressure, and/or cycle time.

(4) Dhawal Shah 22/1/2025

Models
A model should incorporate all of the important
dynamics, while being no more complex than is
necessary. In other words, less important
phenomena may be omitted.

 “All models are wrong, but some are useful”

 Why model a process?

 improves process understanding
 provides predictive (extrapolative) capacity
 facilitates controller synthesis
 helps to optimize process design/ operating conditions
 helps to train operating personnel

(5) Dhawal Shah, 2024 22/1/2025

Modeling approaches
 Fundamental modeling
 Global
 Difficult to develop; need detailed process knowledge (material/energy
balances; heat, mass, momentum transfer chemical kinetics,
thermodynamics, physical property relationships)
 Usually complex (a large set of ODEs or PDEs)
 Fundamentally correct – useful in extrapolation
 Used for simulation (operator training), control, optimization

 Empirical Modeling
 Explain the observed response (cause-effect)
 Easier to develop; need experimental data
 Usually kept simple (lower order ODEs)
 Lacks fundamental correctness (dangerous to extrapolate)
 Should be used with caution
 Used for controller design

(6) Dhawal Shah, 2024 22/1/2025

Hierarchy of process control activities

(7) Dhawal Shah 22/1/2025

Development of dynamic models: Example
 A blending process
Stream 1 Stream 3

Stream 2
 Mixing tank with heating

Stream 4

(8) Dhawal Shah, 2024 22/1/2025

Example
 Ca1 = 1 g/L; v1 = 10 L/min
 Ca2 = 4 g/L; v2 = 20 L/min
 Ca3 = ? g/L; v3 = ? L/min
 V = 150 L

The process is running at steady state, so what is

output concentration?

At some time the operator misreads and changes the

flow rates. He puts stream 1 at 20 L/min and stream
2 at 10 L/min. What will the output concentration as
a function of time?

(9) Dhawal Shah, 2024 22/1/2025

Model
 Assumptions:
 Concentration does not change in the input system
 Perfect mixing
 Constant temperature

Model development

First find the steady state exit

concentration?

Apply mass balance for

process dynamics

(10) Dhawal Shah, 2024 22/1/2025

Example 2: Energy balance
 Ca1 = 1 g/L; v1 = 10 L/min; T1 = 25 oC
 Ca2 = 4 g/L; v2 = 20 L/min ; T2 = 55 oC
 Ca3 = ? g/L; v3 = ? L/min; ; T3 = ? oC
 V = 150 L

 Exit stream

v = ? C = ? T= 80 oC

 Energy Balance but before that: How much heat is

given by the heater?

(11) Dhawal Shah, 2024 22/1/2025

Develop a dynamic model
 Assumptions:
 Perfect mixing
 Constant V (inlet and outlet flow rates are equal)
 The density ρ and heat capacity C of the liquid are
assumed constant.
 Negligible heat loss

 From energy balance:

(12) Dhawal Shah 22/1/2025

Example
 A stirred-tank blending process with a constant liquid holdup of 2 m3
is used to blend two streams whose densities are both approximately
900 kg/m3. The density does not change during mixing.

 Assume that the process has been operating for a long period of time with
flow rates of w1 = 500 kg/min and w2 = 200 kg/min, and feed
compositions (mass fractions) of x1 = 0.4 and x2 = 0.75. What is the
steady state value of x?
 Suppose that w1 changes suddenly from 500 to 400 kg/min and remains
at the new value. Determine an expression for x(t) and plot it.
 Repeat part (b) for the case where w2 (instead of w1) changes suddenly
from 200 to 100 kg/min and remains there.
 Repeat part (c) for the case where x1 suddenly changes from 0.4 to 0.6
(in addition to the change in w2).
 For parts (b) through (d), plot the normalized response xN(t),

where x(0) is the initial steady-state value of x(t) and x(∞)

represents the final steady-state value, which is different for each
part.

(13) Dhawal Shah 22/1/2025

The blending process: Revisited
 The material balance equation gives us

 Combining all the terms we get

 In the state-space format:

(14) Dhawal Shah 22/1/2025

Electrically heated stirred tank
 Remember a hidden assumption: Heat is
transferred instantaneously!!
 Are you sure? What if heat is not transferred
immediately:
 Heat transfer coefficients of the element, he
 Temperature of element, Te

 Combining the two equations, gives us

(15) Dhawal Shah 22/1/2025

Liquid Storage tanks
A typical material balance across the tank will give
us:

Assuming constant density

 Three possible scenarios can develop

 q is constant: Using a pump (So inlet will be same as
outlet)
 q is based on certain resistance of the flow: valve
 q is function of pressure: realistic

(16) Dhawal Shah 22/1/2025

Liquid storage tank
 Case 2: Tank exit line functions simply as a
resistance to flow, Rv, from the tank. Assume flow to
be linearly related to level (analogy to Ohm’s law, E
= IR). Then,

 Case 3: To get a more realistic expression for q, fix a

valve on the exit line and assume turbulent flow.
Then pressure drop across the valve is

(Chapter 4, seborg: Non-linearization)

(17) Dhawal Shah 22/1/2025

What about a reaction
 CSTR are widely used for reaction.

 Considera simple liquid-phase, irreversible chemical

reaction where chemical species A reacts to form
species B. The reaction can be written as A → B. We
assume r = kcA, wherein k is koexp(-E/RT).
Assumption:

1.The CSTR is perfectly mixed.

2. The mass densities of the feed and product
streams are equal and constant. They are
denoted by ρ.
3. The liquid volume V in the reactor is kept
constant by an overflow line.

(18) Dhawal Shah 22/1/2025

 Material Balance
 Material balance: over all, if temperature is constant, density of
constant.

 Material balance over component A

ASSUMPTION Cont.. For energy balance

4. The thermal capacitances of the coolant and the cooling coil wall are negligible compared to
the thermal capacitance of the liquid in the tank.
5. All of the coolant is at a uniform temperature, Tc. (That is, the increase in coolant
temperature as the coolant passes through the coil is neglected.)
6. The rate of heat transfer from the reactor contents to the coolant is given by Q = UA(Tc −
T), where U is the overall heat transfer coefficient and A is the heat transfer area. Both of these
model parameters are assumed to be constant.
7. The enthalpy change associated with the mixing of the feed and the liquid in the tank is
negligible compared with the enthalpy change for the chemical reaction. In other words, the
heat of mixing is negligible compared to the heat of reaction.
8. Shaft work and heat losses to the ambient can be neglected.
(19) Dhawal Shah 22/1/2025
Example
 A classical CSTR; some nominal operating conditions

 The governing equation:

 The coolant temperature comes down to 290 K or

increases to 305 K?

(20) Dhawal Shah 22/1/2025

Answer
 As the coolant temperature goes down  more heat can be
removed from the system  reaction rate goes down  less
conversion  less heat produced  new equilibrium
temperature.

 As the coolant temperature goes up  less heat removed

from the system  high reactor temperature  more
conversion  more heat produced  less A available  less
reaction rate  less heat produced Fluctuations.

(21) Dhawal Shah 22/1/2025

Summary
 Here we reveal the approach to modeling through
derivation of simple dynamic models from first
principles (conservation equations)

 We also recognize that much of Chemical

Engineering concerns continuous processes
operating at steady state. Well and good, but there’s
more to it: continuous . processes may be disturbed
in a variety of ways and thus temperature, pressure,
flow and composition may rise or fall. Process
Dynamics is about evaluating the effects of
disturbances on the process and Process Control is
about managing disturbances for product quality and
safety.

(22) Dhawal Shah 22/1/2025