Instrumentation
and
Process Control
Lecture – 8
Steps in control system design – Introduction to Process modelling
Dr. Hemanth Kumar Tanneru
Department of Chemical Engineering
IIPE Visakhapatnam
In this lecture
• Block diagram for feedback control configurations
• Elements and steps in control system design
• Mathematical models of chemical processes
• Why we need them for control system design?
• Ways to build mathematical model?
• First principles models
• Empirical methods
• Other uses of mathematical models
• Methodologies for modelling
Lecture 8 : Instrumentation and Process control
From process system to block diagram
• Block diagram
• Effective way of showing the
blocks (elements) of control
system and how are the signals
getting transferred between them
• Consider
• Stirred tank heater
• Temperature control
• This can be represented as
block diagram
Lecture 8 : Instrumentation and Process control
Block diagram
Lecture 8 : Instrumentation and Process control
Design elements of a control system
1 2 3 4 5
Define control Select Select Select control Design the
objectives measurements manipulated configuration controller
variables
Lecture 8 : Instrumentation and Process control
Steps in
control system
design
Lecture 8 : Instrumentation and Process control
Mathematical model
The mathematical
The activity of building
representation of physical Mathematical modelling of
this mathematical models
and chemical phenomenon a chemical process
of process
occurring in the process
Requires knowledge from all
the areas of chemical
engineering
Model Mathematical modelling
• Thermodynamics
• Reaction engineering
• Transport phenomena
Lecture 8 : Instrumentation and Process control
Mathematical modelling approaches
• General process
• To investigate the process
behaviour
• 𝑜𝑢𝑡𝑝𝑢𝑡𝑠 = 𝑓 𝑖𝑛𝑝𝑢𝑡𝑠
• Two approaches
• Empirical approach
• Theoretical approach
Lecture 8 : Instrumentation and Process control
Process
• is physically available with you
To understand the behaviour
Empirical • You can deliberately change the inputs and
measurable disturbances
approach • Measure how the outputs are changing
w.r.t time
Demerits
• Time and effort consuming
• Large number of experiments –costly
Lecture 8 : Instrumentation and Process control
Theoretical approach
Often
• A control engineer needs to design the control system even before the original physical
plant is built
• In those cases
• Experimental approach cannot be relied upon
• Some other representation of a process is needed
• The other representation
• Usually set of differential equations, algebraic equations
• Solution of these ODEs will yield dynamic behaviour of the process
Focus
• Theoretical approach
Lecture 8 : Instrumentation and Process control
Need of mathematical models for control
Goal Why mathematical models for control?
• Keep the process variables at their • Understanding of the process is very
set points (Load control) crucial for control system design
• Ever ending disturbances • We might not have physical process
• Track the setpoint changes, when available with us
needed (Servo control) • Even if available the procedure of
understanding the complete dynamic
behaviour by conducting experiments
• Time and effort consuming
• Examples
Lecture 8 : Instrumentation and Process control
Example 1: Stirred tank heater
• Feedback control of temperature
• Design an Proportional integral
controller to control the
temperature in the tank to mitigate
the disturbances coming in the inlet
temperature 𝑇𝑖
Lecture 8 : Instrumentation and Process control
Example 1: Temperature control
• 𝑇𝑖 is the disturbance
• Assuming the flowrates of
inlet and outlet are constant
the level is maintained at the
desired value
• Goal:
• Maintain the temperature of
the tank (𝑇) at desired
temperature (𝑇𝑠 )
• Feedback
Lecture 8 : Instrumentation and Process control
Stirred tank heater-Temperature control
• Controlled variable
• 𝑇
• Manipulated variable
• 𝐹𝑠𝑡 (𝑄)
• 𝑄 = 𝐹𝑠𝑡 𝜌𝑠𝑡 𝐶𝑝 𝑠𝑡 (𝑇𝑠𝑡 − 𝑇)
• How should 𝑄 change to
keep 𝑇 at desired value?
• Model gives this answer
• How?
• We will look into it in detail
in the upcoming classes.
Lecture 8 : Instrumentation and Process control
Example 2: Feed forward
• Feed forward control configuration
• Disturbance is measured
• It is anticipated that it will effect the process
• To maintain the CV at its setpoint
• We need to change the MV such that it
cancels the effect of disturbance of the CV
Lecture 8 : Instrumentation and Process control
Example 2: Feed forward
• Question
• How much should I change the
manipulated variable?
• To answer this we need the following
relations
• 𝑂𝑢𝑡𝑝𝑢𝑡 = 𝑓1 (𝑑𝑖𝑠𝑡𝑢𝑟𝑏𝑎𝑛𝑐𝑒)
• 𝑂𝑢𝑡𝑝𝑢𝑡 =
𝑓2 (𝑚𝑎𝑛𝑖𝑝𝑢𝑙𝑎𝑡𝑒𝑑 𝑣𝑎𝑟𝑖𝑎𝑏𝑙𝑒)
• Model of the process provide this
information
• If the output remains same, the mv should
take a value that satisfy this relation
• 𝑓1 𝑑𝑖𝑠𝑡𝑢𝑟𝑏𝑎𝑛𝑐𝑒 − 𝑓2 𝑚𝑣 = 0
Lecture 8 : Instrumentation and Process control
Example 3: Inferential control
• Controlled variables
• Unmeasured outputs
• Unmeasured outputs = f (measured outputs)
• Obtaining such information
• Without mathematical models is not possible
• Model provides such information
Lecture 8 : Instrumentation and Process control
Need for mathematical models (control)
Feedback • 𝐶𝑉 = 𝑓(𝑀𝑉)
• 𝐶𝑉 = 𝑓(𝑀𝑉)
Feed forward
• 𝐶𝑉 = 𝑓(𝐷)
Inferential • 𝐶𝑉 = 𝑓(unmeasured outputs)
control • 𝐶𝑉 = 𝑓 (𝑀𝑉)
Lecture 8 : Instrumentation and Process control
Other uses of models
• To understand the process
• To train personnel working in plants
• Develop a control strategy for a new process
• Optimize process operating conditions
Lecture 8 : Instrumentation and Process control
Methodology : State variables and State
equations
Lecture 8 : Instrumentation and Process control
