CH4044: Process Dynamics and

Control
D y n a mic B e h avio r

Dr. Dilhara Sethunga; PhD(NTU, Singapore), BSc. Eng (UoM, Sri Lanka)
E mail: [email protected]|Contact: +94778589964

28 August 2023
 Dynamic Response
Single Input-Single Output System
The dynamic behavior of a chemical process is described by an nth
order liner differential equation (or linearized non-linear equation).

INPUT f(t) OUTPUT y(t)

PROCESS

𝑑𝑛 𝑦 𝑑 𝑛−1 𝑦 𝑑𝑦
𝑎𝑛 𝑛 +𝑎𝑛−1 𝑛−1 … … … … … + 𝑎1 + 𝑎0 𝑦 = 𝑏𝑓(𝑡)
𝑑𝑡 𝑑𝑡 𝑑𝑡

01
Dynamic Qualitative Characteristics of the System
Behavior Recap: Poles and zeros
𝑄(𝑠) 𝑄(𝑠)
𝐺 𝑠 = =
𝑃(𝑠) (𝑠 − 𝑃1 )(𝑠 − 𝑃2 ) 𝑠 − 𝑃3 𝑚 (𝑠 − 𝑃4 )(𝑠 − 𝑃4∗ )(𝑠 − 𝑃5 )

02
Dynamic Qualitative Characteristics of the System
Behavior Recap: Poles and zeros

03
Dynamic Laplace Transformation
Behavior Recap: Final and Initial Value Theorem

Final Value Theorem

The asymptotic value of y(t) for large values of time(i.e. y(∞)) can be
found from, providing that lim [sY(s)] exists for all Re(s) ≥ 0
𝑠→0

lim [y(t)] = lim [sY(s)]

𝑡→∞ 𝑠→0

Initial Value Theorem

lim [y(t)] = lim [sY(s)]
𝑡→0 𝑠→∞

05
Dynamic Laplace Transformation Various Functions
Behavior Recap

06
Dynamic Laplace Transformation Various Functions
Behavior Recap

07
Dynamic Laplace Transformation Various Functions
Behavior Recap

08
Dynamic Laplace Transformation Various Functions
Behavior Recap

09
Dynamic Behavior of 1st Order Systems
1st Order Lag/ Linear Lag/Exponential Transfer Lag Systems
𝑑𝑦
𝑎1 + 𝑎0 𝑦 = 𝑏𝑓(𝑡)
𝑑𝑡

10
1st Order 1st Order Lag Systems
Systems Remarks: Time Constant (𝝉𝑷 )
➢ This is a “self-Regulating” Process
➢ Slope of the response at t=0 is 1
➢ If the initial rate of change of y(t) were to be maintained, the
response would reach its final value after 1 time constant elapsed
➢ Smaller the time constant, steeper the initial response of the
system (fast response)
➢ Time constant of a process is a measure of the time necessary for
the process to adjust to a change in its input

Time elapsed 𝝉𝑷 𝟐𝝉𝑷 𝟑𝝉𝑷 𝟒𝝉𝑷

% y(t) from its ultimate value 63.2% 86.6 95 98

11
1st Order 1st Order Lag Systems
Systems Remarks: Steady State Gain (𝐾𝑷 )
➢ The ultimate value of the response is equal to KP for a unit step
change in the input or AKP for a step change with a magnitude of
A
➢ As t→∞, y →KP or AKP
➢ ∆(output)= KP ∆(input) or ∆(output)= AKP ∆(input)
➢ This slows to take the decision of how much of the change
should be applied to the input to obtain a desired change in the
output
➢ A small change in the input, if the system has a large KP (very
sensitive system)
➢ A large change in the input, if the system has a small KP

12
1st Order 1st Order Pure Capacitive Processes
Systems

➢ As t→∞, y →∞
➢ Also known as “Pure integrator” as it behaves as if there were an
integrator between the input and output
➢ This will cause serious control problems, because it cannot
balance itself.
➢ Therefore, termed as “ Non-self regulation processes”
➢ Commonly encountered in inventory systems for raw material and
products

13
1st Order Exercise 01
Systems First order system with a capacity for Mass Storage
Develop a mathematical model for this system assuming,
ℎ
𝐹0 =
𝑅
Compare the dynamic behavior of two systems
1. having different areas (A1>A2) with constant R
2. Having a relationship of A1/A2=R2/R1

Steps:
1. Apply mass balance to the system.
2. Obtain the deviation format of the model
3. Obtain time constant and steady-state gain
4. Draw the dynamic response and compare the results

14
1st Order Exercise 02
Systems Pure Capacitive System
Develop a mathematical model for this system assuming,
F0 is independent of h
Explain the dynamic behavior of the system

Steps:
1. Apply mass balance to the system.
2. Obtain the deviation format of the model
3. Draw the dynamic response and compare the results

15
Dynamic Behavior of 2nd Order Systems
2nd Order Systems
𝑑2𝑦 𝑑𝑦
𝑎2 2 +𝑎1 + 𝑎0 𝑦 = 𝑏𝑓(𝑡)
𝑑𝑡 𝑑𝑡

➢ Multi-Capacity processes: Processes that consists with two or

more capacities (first order systems)
➢ Inherently 2nd order systems
➢ A processing system with its controller

16
2nd Order Dynamic Behavior
Systems Case A: Over Damped Response (ζ>1)

➢ Sluggish behavior as compared to 1st order systems

➢ More sluggish as ζ increases
➢ Can be expected in multi-capacitive systems (1st order systems in series)

17
2nd Order Dynamic Behavior
Systems Case B: Critically Damped Response (ζ=1)

➢ Approaches its ultimate value faster than over damped systems

18
2nd Order Dynamic Behavior
Systems Case C: Under Damped Response (0<ζ<1)

➢ Underdamped response is initially

faster than the critically or
overdamped responses
➢ Although the underdamped
response is initially faster and
reaches its ultimate value quickly, it
does not stay they, but starts
oscillating with progressively
decreasing amplitude.
➢ Oscillatory behavior is more
pronounced with smaller values of
the damping factor
➢ Almost all the underdamped behavior are due to the interaction of
controller with the process units they control

19
 2nd Order Dynamic Behavior
Systems Characteristics of Under Damped Response

➢ Periods of Oscillations
1 − 𝜉2
𝜔 = radian frequenzy =
𝜏
1 2𝜋𝜏
=𝑇=
𝑓 1 − 𝜉2

➢ Overshoot: ratio of A/B; B=ultimate value of the response, A=maximum amount by which
the response exceeds its ultimate value.
−𝜋𝜉
𝑂𝑣𝑒𝑟𝑠ℎ𝑜𝑜𝑡 = exp
1 − 𝜉2
➢ Decay Ratio: C/A (the ratio of the amounts above the ultimate value of two successive peaks)
−2𝜋𝜉
𝐷𝑒𝑐𝑎𝑦 𝑅𝑎𝑡𝑖𝑜 = exp = 𝑜𝑣𝑒𝑟𝑠ℎ𝑜𝑜𝑡 2
1 − 𝜉2

20
 2nd Order Dynamic Behavior
Systems Characteristics of Under Damped Response

➢ Natural Periods of Oscillations (ζ=0)

1
𝜔𝑛 =
𝜏
1
= 𝑇𝑛 = 2𝜋𝜏
𝑓𝑛
➢ Response time: The response of an underdamped system will reach its ultimate value in a
oscillatory manner as t→∞. Practically, it is considered that the response reached its final
value when it came within ±5% of its final value and stays there. Time needed to reach this
situation is called the response time
➢ Rise Time: used to characterize the speed of the response of an underdamped system. It is
defined as the time required for the response to reach its final value for the first time. The
shorter the ζ, the smaller the rise time, but at the same time the larger the value of the
overshoot
➢ In designing controllers, the objective is proper selection of corresponding ζ and 𝜏 values, so
that resulting a smaller overshoot, shorter rise time, smaller decay ratio and smaller response
time.

21
Dead Time Dynamic Behavior
Dead- Time Processes

22
Dead Time Dynamic Behavior
Dead-Time (Time Delay) Processes

23
Dead Time Dead-Time (Time Delay) Processes
Exercise
The trickle-bed catalytic reactor utilizes product recycle to obtain satisfactory operating
conditions for temperature and conversion. Use of a high recycle rate eliminates the need for
mechanical agitation. The concentration of the reactant is measured at a point in the recycle
line where the product stream is removed. A liquid phase first-order reaction is assumed.
Under normal operating conditions, the following assumptions may be made:
I. The reactor operates isothermally with a reaction rate given by r = kc, where −r denotes
the rate of disappearance of reactant per unit volume, c is the concentration of reactant,
and k is the rate constant.
II. All flow rates and the liquid volume V are constant.
III. No reaction occurs in the piping. The dynamics of the exit and recycle lines can be
approximated as constant time delays, θ1 and θ2, as indicated in the figure. Let c1 denote
the reactant concentration at the measurement point.
IV. Because of the high recycle flow rate, the reactor is completely mixed.

a) Derive an expression for the transfer function C′1(s)∕Ci′(s).

b) Using the following information, calculate c′1(t) for a step change in c′i(t) of 2000 kg∕m3.

24
Dead Time Dead-Time (Time Delay) Processes
Exercise

Parameter Values
V= 5 m3
α = 12
q = 0.05 m3∕min
θ1 = 0.9 min
k = 0.04 min−1
θ2 = 1.1 min

25
Dead Time Approximations for Dead Time
Tayler Series, Pàde, Skogestad’s Half Rule

➢ Tayler Series Approximations (1st Order)

𝑒 −𝜃𝑠 = 1 − 𝜃𝑠
𝜃𝑠 −𝜃𝑠
1
𝑒 = 1 + 𝜃𝑠 ⇒ 𝑒 =
1 + 𝜃𝑠
➢ Pàde Approximation (1st and 2nd order)
𝜃𝑠
1− 2
𝑒 −𝜃𝑠 =
𝜃𝑠
1+ 2

𝜃 2 𝑠 2 − 6𝜃𝑠 + 12
𝑒 −𝜃𝑠 = 2 2
𝜃 𝑠 + 6𝜃𝑠 + 12
➢ Skogestad’s Half Rule
• For higher models that contains multiple time constants
• find largest neglected time constant (2nd largest time constant) in the
denominator and add one half of its value to the existing time delay (if any) and add
another half to the retained time constant. All other time constants should be added to
the time delay in the numerator.

26
Dead Time Approximations for Dead Time
Example

Consider the transfer function,

𝐾(−0.1𝑠 + 1)
𝐺(𝑠) =
5𝑠 + 1 3𝑠 + 1 (0.5𝑠 + 1)
Derive an approximate first order plus time delay model using
1. Tayler series expansions and
2. Skogestad’s half Rule
And compare the results

27
 Inverse Dynamic Behavior
Response Inverse Response

𝑏𝑚 𝑠 𝑚 + 𝑏𝑚−1 𝑠 𝑚−1 + ⋯ … … … … … + 𝑏0
𝐺0 𝑠 =
𝑎𝑛 𝑠 𝑛 + 𝑎𝑛−1 𝑠 𝑛−1 + ⋯ … … … … … + 𝑎0

➢ Where, one of the roots of the numerator has a positive real part
➢ Systems with inverse response are particularly difficult to handle
and need special attention

28
 Inverse Dynamic Behavior
Response Inverse Response

29
 Inverse Dynamic Behavior
Response Inverse Response-Example

For the case of a single zero in an overdamped second-order transfer

function,

𝐾1 (𝜏𝑎 𝑠 + 1)
𝐺𝑠 𝑠 =
𝜏1 𝑠 + 1 𝜏2 𝑠 + 1

calculate the response to a step input of magnitude M and plot the

results for τ1 = 4, τ2 = 1 and several values of τa.

29
 Dynamic Behavior of Higher Order
2ndOrder Systems Systems
Multi-capacitive 1st order processes in series
• The response has the characteristics of an overdamped system
(s-shaped and sluggish)
• Increasing the number of capacities in series increases the
sluggishness of the response
𝐾1 𝐾2 ………𝐾𝑁
𝐺0 𝑠 = 𝐺1 𝑠 𝐺2 𝑠 …………….. 𝐺𝑁 𝑠 = 𝜏1 𝑠+1 𝜏2 𝑠+1 ………..(𝜏𝑁 𝑠+1)

30
 Interacting and Non-interacting
Processes
Noninteracting process: Interacting process: downstream
changes in a downstream units affect upstream units, and vice
unit have no effect on versa
upstream units.

31