1.

Introduction
11 August 2021 12:48

Block Diagram of a control system

The external influences, which have the potential to affect the dynamics of the process, are termed as process inputs whereas the entities, which
themselves are affected due to the changes in process behaviour, are called process outputs. Flow rate of the feed stream is an example of process
input that can influence the process dynamics, whereas the purity of the product stream can be called a process output (as the purity is a direct
consequence of changes in process behaviour).

In an ideal control configuration, process inputs are manipulated in a calculated manner in order to maintain the process output at a predefined
desired set point. Input of a process can be of two types viz. Manipulated input and disturbance. Input, Which can be manipulated as per process
requirement, are named manipulated input. And the inputs which are outside the scope of any manipulation but whose changes are difficult to avoid,
are termed as disturbance. Feed flow rate and feed concentration are examples of manipulated input and disturbance respectively.

Output of a process can be of two types viz, measured and unmeasured. Product flow rate and product concentration are examples of measured &
unmeasured outputs respectively.

Aspects of good control strategy

The development of a good control strategy would ideally consist of identifying the following aspects:
• Control Objectives: The first and foremost aspect of control system design is to identify the control objectives. In other words, the designer should
find out the answer for “What needs to be ensured that would lead to best operation of the process?” It is rather common to divide a large process
plant into several smaller segments and identify the control objectives of each segment separately. However, it is not unusual to find some conflicting
objectives that would promise supposedly best operation. Hence, identifying control objective is not a trivial task.
• Operating Characteristics: One needs to find out whether the process runs on batch mode, semi-batch mode or continuous mode. The operability
issues at the time of start-up and shutdown may also be scrutinized.

Page 1
• Operational constraints: The operational constraints include various types of limitation that the associated equipment may have. For example, a
valve cannot open more than 100% even if the best operability demands so; a pump has to maintain the NPSH; flooding of plates in a distillation
column should be avoided; etc.

• Safety: In a plant operation, particularly in hazardous condition, safety has the top most priority. Human lives are most important and a control
engineer has to ensure maximum security for the people working inside the plant while designing its control system. For an example, if a furnace has
been designed to operate with maximum allowable temperature of 1000°C, one needs to ensure that the controlled temperature is maintained much
below this value. Alarm switches of different warning levels are integral parts of a control system.

• Environmental regulations: Effluent management is a critical issue for any plant operation. Government regulations are set in order to limit the
release of toxic gas (in atmosphere) and wastewater (into river). A control system is required to ensure that such regulations are strictly followed.

• Economics: Target of any meaningful business is to maximize its profit. In order to achieve that, one needs to minimize the production cost, losses,
wastage, energy requirement and human labour. A good control system should ensure an optimum level of performance that maximizes the profit
conforming to all safety & environmental regulations, product specifications and operational constraints.

Classification of Process Variables

Let us take the example of a continuous stirred tank reactor (CSTR) as shown below. Reactant A is entering into the CSTR at concentration CAi ,
flow rate Fi and temperature Ti. Product is coming out of the CSTR at concentration CA, flow rate F and temperature T. The tank volume is V. The
reaction is exothermic and hence coolant is passed through the cooling jacket at a flow rate Fc. Coolant is entering the jacket at temperature Tci and
exiting at Tco .

Page 2
 A process of CSTR

The process variables are as follows:

The coolant flow rate (Fc) can be manipulated to control the temperature (T) . Inlet temperatures of reactant and coolant (Ti, Tci) are disturbances
which can be measured, however concentration of reactant (CAi) is an unmeasured disturbance. Flow rates and temperatures can be measured and
can be regarded as primary output whereas volume of the liquid inside the tank can be estimated by measuring its level. Hence it is a secondary
output. Outlet concentration of product is an unmeasured output unless there is an on-line concentration analyzer in place.

The position of valve in the reactant/product line decides the identity of Fi and F. If valve is placed at inlet, Fi is the manipulated input and F is the
measured primary output. If the valve is placed at the outlet line, F is the manipulated input and Fi is the source of measured disturbance.

State variables and equations

Page 3
State variables and equations

These variables are a set of fundamental dependent quantities whose values describe the natural state of a given process. The state equations are
set of differential equations which describe the progression of the states with time. The state variables are primarily the fundamental quantities of a
process viz., mass, energy and momentum, whereas the state equations are generated out of conservation principle involving these fundamental
quantities.

Let S be the fundamental quantity. The state equation can be written as:

If there are N no. of streams going inside a process and M no. of streams coming out of a process, the mass balance equations for the process can
be written as:

where, ρi, j , Fi,j and Vi, j represent the density, volumetric flowrate and volume of the material inside and outside the system respectively and In case
of a chemical reaction, let CA and CA i,j be the concentrations of component A of material inside the process and the streams respectively and r be
the rate of the reaction. The component balance equations can be written as:

Let h and hi,j be the enthalpies of material inside the process and incoming/outgoing streams respectively, Q be the heat supplied to the system (or
heat removed from the system ), W is the work done on (or by) the system. Then the energy (E) balance equation for such system is written as:

In addition to the above state equations, a few algebraic equations are also useful for modelling of a chemical process. Examples of such equations
are:

Page 4
Page 5
2. Process Modelling
11 August 2021 20:28

Physical Examples

1. Mercury Thermometer

Consider the thermometer to be located in a flowing stream of fluid for which the temperature x varies with time. Assuming

• All the resistance to heat transfer resides in the film surrounding the bulb (i.e., the resistance offered by the glass and mercury is neglected).

• All the thermal capacity is in the mercury. Furthermore, at any instant the mercury assumes a uniform temperature throughout.

• The glass wall containing the mercury does not expand or contract during the transient response.

By applying the unsteady-state energy balance,

Rate of Input - Rate of Output = Rate of accumulation

We get,

where,
A = surface area of bulb for heat transfer, ft 2
C = heat capacity of mercury, Btu/(lb m · °F)
m = mass of mercury in bulb, lbm
t = time, h
h = film coefficient of heat transfer, Btu/(ft 2 · h · °F)

At steady state,

Page 6
At steady state,

or,

If we define the deviation variables to be the differences between the variables and their steady-state values

X = x - xS
Y = y - yS
we get,

If we let mC/hA = ࣎ , and take Lapalce of the above equation we get,

or,

2. Liquid level

Consider the system shown below , which consists of a tank of uniform cross-sectional area A to which is attached a flow resistance R such as a
valve, a pipe, or a weir. Assume that qo , the volumetric flow rate (volume/time) through the resistance, is related to the head h by the linear
relationship:

Page 7
From mass balance,

Rate of Input - Rate of Output = Rate of accumulation

or,

or,

At steady state,

If we define
Q = q - qS
and
H = h - hS
then,

and from Laplace Transform,

or,

Page 8
where, ࣎ = AR

or, if we put R = H(s)/Q0(s), we get

3. Liquid level with constant outlet flowrate

For this system also,

Introducing the deviation variable in the above equation as:

Taking the Laplace we get,

This can be treated as the special case of the liquid level system with variable flowrate at outlet where, R → ∞ (infinitely large resistance).

4. Mixing Process

Consider the mixing process shown below in which a stream of solution containing dissolved salt flows at a constant volumetric flow rate q into a
tank of constant holdup volume V. The concentration of the salt in the entering stream x (mass of salt/volume) varies with time. It is desired to
determine the transfer function relating the outlet concentration y to the inlet concentration x.

Page 9
If we assume the density of the solution to be constant, the flow rate in must equal the flow rate out, since the holdup volume is fixed. We may
analyze this system by writing a transient mass balance for the salt. Thus

from material balance,

Rate of Input - Rate of Output = Rate of accumulation
or,

At steady state,

Subtracting this equation from the previous and introducing the devation variables as:

X = x - xS
Y = y - yS
We get,

Taking the Lapalce and arranging it to get,

where, ࣎ = V/q

5. Heating Process

Consider the heating process shown below. A stream at temperature Ti is fed to the tank. Heat is added to the tank by means of an electric heater.
The tank is well mixed, and the temperature of the exiting stream is T. The flow rate to the tank is constant at w lb/h.

Page 10
A transient energy balance on the tank yields

Rate of Energy Input - Rate of Energy Output + Rate of Energy generated from heater = Rate of accumulation of energy in tank

where Tref is the reference temperature and C is the heat capacity of the fluid. At steady state, dT /dt is zero,

Subtracting this equation from above equation, we will get

If we assume that Ti is constant (and so Ti = Tis) and introduce the deviation variables as

T - Ts = T '
q - qs = Q
we get,

Page 11
Taking Lapalce of the above equation and rearranging it to obtain,

where, ࣎ = ρV/w

Standard form of First Order System

The general form of a first order system is,

where y is the output variable and x(t) is the input forcing function. The initial conditions are

Introducing deviation variables gives

X = x - xS
Y = y - yS
we get,

Taking Laplace Transform of the above equation

or,

Forcing Functions

I. Step Function

Page 12
Mathematically, it is represented as X(t) = Au(t), where u(t) is a unit step function. If we take laplace of this, we will get X(s) = A/s

Response

II. Impulse Function

Mathematically, it is represented as X(t) = Aδ(t), where δ(t) is an impulse function. If we take laplace of this, we will get X(s) = A

Response

Page 13
 III. Ramp Input

This function increases linearly with time and is described by the equations X = 0 at t < 0 and X = bt at t ൒ 0. If we take laplace of this, we will get
X(s) = b/s2

Response

Page 14
IV. Sinusoidal Input

This function increases linearly with time and is described by the equations X = 0 at t < 0 and X = Asin ωt at t ൒ 0. If we take laplace of this, we will
get X(s) = Aω/(s2 + ω2). where A is the amplitude and ω is the radian frequency. The radian frequency ω is related to the frequency f in cycles per
unit time by f = 2࣓࣊ .

Response

Page 15
or,

where,

At steady state, t → ∞. So,

On comparing this function with the input, we will observe:

a. The output is a sine wave with a frequency ω equal to that of input signal.

ۛۛۛۛۛۛۛۛۛ
b. The ratio of output amplitude to input amplitude is 1/ ඥ૚ ൅ ࣎૛ ࣓૛ . This ratio will always be less than 1.
c. The output lags behind the input by angle ࣐. So, it will always be negative.

Page 16
Linearization

Again consider the liquid - level in tank problem and this time, we assume that flow out of the tank follows a square root relation with the height as:

q0 = Ch1/2
where, C is constant.

For a liquid of constant density and a tank of uniform cross-sectional area A, a material balance around the tank gives

Putting the value of q0 in the above equation, we get

At this point, we cannot proceed as before and take the Laplace transform. This is due to the presence of the nonlinear term h1/2, for which there is
no simple transform. This difficulty can be circumvented by linearizing the nonlinear term.

By means of a Taylor series expansion, the function qo (h) may be expanded around the steady-state value hs ; thus

where qo'(hS) is the first derivative of qo evaluated at hS, qos'' (hS) is the second derivative, etc. If we keep only the linear term, the result is

where, qo'(hS) = C/2hs1/2. Putting this into the above equation we get,

Here, qos = qo(hS) and 1/R1 = C/2hs1/2

Putting the above equation in the following equation,

Page 17
We get,

At steady state the flow entering the tank equals the flow leaving the tank; thus qS = qos . Putting this in the above equation we get,

Introducing the deviation variables as:

Q = q - qS
H = h - hs
we get,

where, ࣎ = AR1 and R1 = 2hs1/2 /C

Response of first order systems in series

i. Non - interacting system

For tank -1,

Page 18
where, Q1 = h1/R and ࣎૚ ൌ ࡭૚ ࡾ૚

For tank - 2,

where, ࣎૛ ൌ ࡭૛ ࡾ૛
or,

In general,

The overall transfer function will be,

where, ∏ denotes product from i = 1 to i = n

ii. Interacting (loading) system

Page 19
Second Order System

They can be represented by the following differential equation,

And, the corresponding transfer function will be,

Example of Second order system

From force balance,

Page 20
The rate of change of momentum of the fluid may be expressed as:

The momentum correction factor 'β' accounts for the fact that the fluid has a parabolic velocity profile in the tube, and the momentum must be
expressed as βmV for laminar flow. The value of β for laminar flow will be 4/3.

So,

or,

Page 21
Here,

So, the transfer function will be,

Process Response

Description of an underdamped response

Page 22
 1.

2.

3. Rise time: This is the time required for the response to first reach its ultimate value and is labelled tr in above figure and tr increases with
increasing ζ.

4. Response time: The time required for the response to come within േ5 % of its ultimate value and remain there. The response time is indicated
above . The limits േ 5 % are arbitrary, and other limits can be used for defining a response time.

5. Period of Oscillation: The radian frequency (radians/time) is the coefficient of t in the sine term; thus,

or,

Page 23
where T is the period of oscillation (time/cycle). In terms of the above graph , T is the time elapsed between peaks.

If we put ζ = 0, we will get the natural time period of oscillation.

Transportation Lag

The transportation lag parameter ࣎ is simply the time needed for a particle of fluid to flow from the entrance of the tube to the exit, and it can be
calculated from the expression:

or,

Its transfer function will be ࢋି࢙࣎

They can be approximated by the Taylor series expansion as:

The transportation lag is quite common in the chemical process industries where a fluid is transported through a pipe. The presence of a
transportation lag in a control system can make it much more difficult to control. In general, such lags should be avoided if possible by placing
equipment close together. They can seldom be entirely eliminated.

Response of Transportation lag

Page 24
3. The Control Systems
12 August 2021 14:14

Consider the following process,

A liquid stream at a temperature Ti enters an insulated, well-stirred tank at a constant flow rate w (mass/time). It is desired to maintain (or control) the
temperature in the tank at TR by means of the controller. If the measured tank temperature Tm differs from the desired temperature TR, the controller
senses the difference or error ε = TR - Tm and changes the heat input in such a way as to reduce the magnitude of ε. If the controller changes the heat
input to the tank by an amount that is proportional to ε, we have proportional control.

It is indicated that the source of heat input q may be electricity or steam. If an electrical source were used, the final control element might be a variable
transformer that is used to adjust current to a resistance heating element; if steam were used, the final control element would be a control valve that
adjusts the flow of steam. In either case, the output signal from the controller should adjust q in such a way as to maintain control of the temperature in
the tank.

Components of control system

1. Process (stirred-tank heater).

2. Measuring element (thermometer).
3. Controller.
4. Final control element (variable transformer or control valve).

Block diagram

Page 25
Block diagram

The control system shown above is called a closed-loop system or a feedback system because the measured value of the controlled variable is
returned or “fed back” to a device called the comparator. In the comparator, the controlled variable is compared with the desired value or set point. If
there is any difference between the measured variable and the set point, an error is generated. This error enters a controller, which in turn adjusts the
final control element to return the controlled variable to the set point.

Important Terms

Closed loop: The measured value of the controlled variable is fed back to the controller.

Controller: A device that outputs a signal to the process or final control element based on the magnitude of the error signal. A proportional controller
outputs a signal proportional to the error.

Deviation variable: The difference between the actual value of a variable and its steady-state value. Block diagrams are always constructed using
deviation variables.

Error: The difference between the value of the set point and that of the measured variable.

Final control element: A device that provides a modulated input to the process in response to a signal from the controller. For example, this may be a
heater, a control valve, or a variety of other devices.

Load: The change in any process variable that can cause the controlled variable to change.

Measuring element: A sensor used to determine the value of the controlled variable and to send it to the comparator/controller. Examples include a
thermocouple (temperature), a strain gage (pressure), a gas chromatograph (composition), and a pH electrode (acidity). These sensors typically have
some dynamic behaviour associated with them and can affect the design of the control system.

Negative feedback: The error is the difference between the set point and the measured variable (this is usually the desired configuration).

Positive feedback: The measured variable is added to the set point. (This is usually an undesirable situation, and frequently it leads to instability.)

Regulator problem: The goal of a control system for this type of problem is to enable the system to compensate for load changes and maintain the

Page 26
Regulator problem: The goal of a control system for this type of problem is to enable the system to compensate for load changes and maintain the
controlled variable at the set point.

Servo problem: The goal of a control system for this type of problem is to force the system to “track” the requested set point changes.

Set point: The desired value of the controlled variable

Types of Controllers

1. Proportional (P) Controller

The proportional controller can reduce the error, but cannot eliminate it. If we can accept some residual error, proportional control may be the proper
choice for the situation. The proportional controller has only one adjustable parameter, the controller gain. The proportional controller produces an
output signal (pressure in the case of a pneumatic controller, current, or voltage for an electronic controller) that is proportional to the error ε. This action
may be expressed as:

where,
p = output signal from controller, psig or mA
KC = proportional gain, or sensitivity
ε = error = (set point) - (measured variable)
pS = a constant, the steady-state output from the controller

Introducing a deviation variable, P = p - pS and assuming that ε = 0 at t = 0. We have,

Page 27
With proportional-only control there will always be some offset (set point of output - steady state value of output) and as KC ↑ offset ↓.

This controller is also known as the gain controller and its equivalent representation is proportional band. It is the amount of change in error that will
cause the control action to go from full OFF to full ON. The amount of change in error is calculated as % of full scale error.

• ON/OFF control

It is a special case of proportional control is on/off control. If the gain KC is made very high, the valve will move from one extreme position to the other if
the process deviates only slightly from the set point. This very sensitive action is called on/off action because the valve is either fully open (on) or fully
closed (off); i.e., the valve acts as a switch.

In practice, a dead band is inserted into the controller. With a dead band, the error reaches some finite positive value before the controller “turns on.”
Conversely, the error must fall to some finite negative value before the controller “turns off.

2. Proportional Integral (PI) Controller

The integral mode ultimately drives the error to zero. This controller has two adjustable parameters for which we select values, the gain and the integral
time. Thus it is a bit more complicated than a proportional controller, but in exchange for the additional complexity, we reap the advantage of no error at
steady state. PI control is described by the relationship:

where,
Kc = proportional gain
࣎ࡵ = integral time, min
p = constant (the bias value)

Page 28
pS = constant (the bias value)

Introducing a deviation variable, P = p - pS, we have

With proportional-integral control there is no offset. However, the addition of the integral action in the controller can cause the system response to
become more oscillatory and increase the maximum deviation (or overshoot) of the response.

PI Control = No offset, but may be oscillatory

3. Proportional Derivative (PD) Controller

Derivative control is another mode that can be added to our proportional or proportional-integral controllers. It acts upon the derivative of the error, so it
is most active when the error is changing rapidly. It serves to reduce process oscillations. This mode of control may be represented by:

where,
Kc = proportional gain
࣎ࡰ = derivative time, min
pS = constant (the bias value)

Introducing a deviation variable, P = p - pS, we have

Page 29
 4. Proportional Integral Derivative (PID) Controller

This mode of control is a combination of the previous modes and is given by the expression

Introducing a deviation variable, P = p - pS, we have

Closed Loop Transfer Function

The block diagram of stirred tank heater has been redrawn and incorporates some standard symbols for the variables and transfer functions, which are
widely used in the control literature. These symbols are defined as follows:

Page 30
Overall Transfer Function For Single Loop Systems

• Overall Transfer Function for CHANGE IN SET POINT

For this case, disturbance U = 0 and the above figure may be simplified or reduced as:

This can be proved as shown below:

C = Gε

Page 31
 C = Gε
where,
ε=R-B
and,
B = CH
Putting the value of C in above equation, we get
B = HGε
or,
B = HG(R - B)
or,
CH = HG(R - CH)
or,

• Overall Transfer Function for CHANGE IN LOAD (DISTURBANCE)

For this case, set point R = 0 and the above figure may be simplified or reduced as:

This can be proved as shown below:

C = G2(M + U)
where,
M = GCG1ε
where,
ε=-B
and,
B = CH
So,
C = G2(GCG1ε + U)
or,

Page 32
or,
C = G2(-CHGCG1 + U)
or,

Important Terms

Closed-loop process: Process in which the feedback loop is connected to the comparator.

Closed-loop transfer function: Transfer functions relating two variables in the process when the feedback loop is connected to the comparator.

Feedback path: The path that connects the controlled variable and the comparator.

Forward path: The transfer functions that lie between two signals in the block diagram moving left to right as drawn in the block diagram above. The
complete forward path consists of G = GCG1G2. The forward path between U and C is G2 only.

Open-loop process: Process in which the feedback loop is disconnected from the comparator.

Open-loop transfer function: Product of all transfer functions in the loop relating B and R when the feedback loop is disconnected from the
comparator.

Page 33
4. Processes with Inverse Response
13 August 2021 10:59

Let us assume two first order processes are arranged in the following manner with

Page 34
 Schematic of a process producing inverse response

Since ࣎૚ ൐ ࣎૛ , the Process-2 will be faster than Process-1. Hence, the initial direction of the overall process response will be guided by the
Process-2 in the negative direction.

On the other hand, K1 > K2, indicates that Process 1 will guide the ultimate steady state in the positive direction by virtue of its higher gain than the
Process-2.

As a result, the overall process response curve will initially transit towards in the opposite direction of its final steady state. However, the transition will
change its direction after some time and finally settle at the steady state value. This type of response is termed as inverse response.

Page 35
The open-loop response of the system is

This system has a positive zero

It is noted that any system having positive zero will demonstrate an inverse response.

Compensation for inverse response

Conventional

With Compensator

Page 36
In order to have a non-positive zero for this system, the value of k should be chosen in such a way that

Page 37
5. Frequency Response Analysis
13 August 2021 11:47

Consider a simple first-order system with transfer function

Putting s = jω in the above equation we get,

We may convert this expression to polar form by multiplying numerator and denominator by the conjugate of ࣎࢐࣓ ൅ ૚ we get,

To convert a complex number in rectangular form (a + jb) to polar form Rejφ, where φ is phase angle and R = magnitude expressed as:

ۛۛۛۛۛۛۛ
࣐ ൌ ‫ିܖ܉ܜ‬૚ ሺെ࣓࣎ሻ and ࡾ ൌ ඥࢇ૛ ൅ ࢈૛
So,

And, we know that,

where,

At steady state t→∞ , the first term in RHS will vanish and we will obtain,

Page 38
So, after sufficient time elapses, the response of a first-order system to a sinusoidal input of frequency ࣓ is also a sinusoid of frequency ࣓.

Characteristic of a steady state sinusoidal input

• After transients die out, the output is a sine wave.

• Input frequency = output frequency = ࣓ .

• In general, the output is attenuated, that is, AR < 1.

• The output is shifted in time (it lags the input by the phase angle φ).

• Amplitude ratio and phase angle are both functions of frequency.

Page 39
Bode Plots

It consists of two graphs: logarithm of AR versus logarithm of frequency and phase angle versus logarithm of frequency .

For first order systems, we know that

Page 40
For first order systems, we know that

or,

Bode Plot

Limiting conditions

• As ࣓࣎ → 0, AR → 1 Low Frequency Asymptote

• As ࣓࣎ → ∞, the above equation becomes asymptotic to

log AR = - log(࣓࣎)

which is a line of slope -1, passing through the point

࣓࣎ = 1 AR = 1 (High frequency asymptote)

ۛۛ
The frequency at which 2 asymptotes intersect is known as corner frequency (ωC = 1/࣎), AR = 1/√૛ and ΦC = ‫ିܖ܉ܜ‬૚ ሺെ࣓࣎ሻ = ‫ିܖ܉ܜ‬૚ ሺെ૚ሻ = - 450

Series of First Order Systems

• Overall log AR will be the sum of the individual log AR or overall AR will be the product of the individual ARs.

Page 41
 • Overall log AR will be the sum of the individual log AR or overall AR will be the product of the individual ARs.

or,

• Overall phase angle is the sum of the individual phase angles.

and,

Second Order Systems

The frequency response of a system with a second-order transfer function

or,

Bode Plot

Page 42
Transport Lag

G(s) = ࢋି࢙࣎
AR = 1
࣐ ൌ െ࣓࣎ rad or - 57.29࣓࣎ deg
Proportional (P) control
G(s) = KC
AR = 1
φ = 0 at every ࣓
Proportional Integral (PI) control
૚
G(s) = ࡷ࡯ ቀ૚ ൅ ۛۛቁ
࣎ࡵ ࢙


ۛۛۛۛۛۛۛۛۛ
૚
ࡷ࡯ ૚ ൅ ૛
ሺ࣓࣎ ሻ

Page 43


ۛۛۛۛۛۛۛۛۛ
૚
AR = ࡷ࡯ ට૚ ൅ ሺ࣓࣎
ۛۛۛۛۛ
ሻ૛ ࡵ

φ = ‫ܖ܉ܜ‬ ି૚
ሺെ૚/࣎ࡵ ࣓ሻ
Bode Plot

Proportional Derivative (PD) control

G(s) = KC(1 + ࣎ࡰ ࢙)
AR = ࣓࣎
φ = ‫ିܖ܉ܜ‬૚ ሺ࣓࣎ሻ
Bode Plot

Page 44
Bode Plot

Proportional Integral Derivative (PID) control

Page 45
6. Stability of Control System
13 August 2021 22:11

A stable system will be defined as one for which the output response is bounded for all bounded inputs. A system exhibiting an unbounded response to
a bounded input is unstable.

Stability Criteria

Consider the following closed loop system

To determine under what conditions the system represented by above is stable, it is necessary to test the response to a bounded input. Suppose a unit-
step change in set point is applied. Then,

where, r1, r2 and r3 are the roots of the characteristic equation:

1 + GOL = 0

If there are any of the roots r1, r2, . . . , rn in the right half of the complex plane, the response C(t) will contain a term that grows exponentially in time and
the system is unstable.

Therefore, the definition of stability for linear systems may be translated to the following criterion: A linear control system is unstable if any roots of
its characteristic equation are on, or to the right of, the imaginary axis. Otherwise the system is stable.

Page 46
 Stability of the typical roots of the characteristic equation

Routh Test of Stability

It is a purely algebraic method for determining how many roots of the characteristic equation have positive real parts; from this it can also be determined
whether the system is stable, for if there are no roots with positive real parts, the system is stable. The test is limited to systems that have polynomial
characteristic equations. This means that it cannot be used to test the stability of a control system containing a transportation lag.

The procedure for examining the roots is to write the characteristic equation in the form

where a0 is positive. (If a0 is originally negative, both sides are multiplied by -1.) In this form, it is necessary that all the coefficients:

be positive if all the roots are to lie in the left half-plane. If any coefficient is negative, the system is definitely unstable, and the Routh test is not needed
to answer the question of stability. (However, in this case, the Routh test will tell us the number of roots in the right half-plane.) If all the coefficients are

Page 47
to answer the question of stability. (However, in this case, the Routh test will tell us the number of roots in the right half-plane.) If all the coefficients are
positive, the system may be stable or unstable. It is then necessary to apply the following procedure to determine stability.

Arrange the coefficients into the first two rows of the Routh array as follows:

In general, there are n + 1 rows. For n even, the first row has one more element than the second row. The elements in the remaining rows are found
from the formulas:

Theorems of Routh Test

The necessary and sufficient condition for all the roots of the characteristic equation to have negative real parts (stable system) is that all elements of
the first column of the Routh array ( a0 , a1, b1 , c1 , etc.) be positive and nonzero.

If some of the elements in the first column are negative, the number of roots with a positive real part (in the right half-plane) is equal to the number of
sign changes in the first column.

If one pair of roots is on the imaginary axis, equidistant from the origin, and all other roots are in the left half-plane, then all the elements of the nth row
will vanish and none of the elements of the preceding row will vanish. The location of the pair of imaginary roots can be found by solving the equation:

Cs2 + D = 0

where the coefficients C and D are the elements of the array in the ( n – 1)th row as read from left to right, respectively.

Root Locus

Procedure

1. Write the characteristic equation , 1 + GOL = 0

Page 48
 2. Find the no. of zeroes (Z, root of numerator) (n) and poles (P, root of denominator) (m).

3. Find the coordinate of the centroid as:

∑

ࡼ ି ∑

ࢆ
Centroid ≡ ۛۛۛۛۛۛۛ
࢓ ି࢔
Note: Centroid always lies on x - axis only.

4. Find the no. of asymptotes = m - n

5. Find the angle between asymptotes as:

ሺ૛ࢗା૚ሻ࣊
Angle between asymptotes = ۛۛۛۛۛۛ
࢓ ି࢔
For example if m - n = 3 then, q = 0, 1, 2.

6. Find the break-away point (A point on real axis at which root locus moves away from it) by putting

7. Find the point of intersection of root locus with imaginary axis by putting s = ωj in the characteristic equation.

Bode Stability Criteria

It states that a control system is unstable if the open loop frequency response exhibits amplitude ratio AR more than unity at the frequency ω at which
phase lag is 1800 i.e. φ = - 1800 which is known as cross-over frequency(ωCO).

Gain margin & Phase margin

Page 49
Gain margin & Phase margin

• Gain Margin(GM) > 1.7 for a stable system in designing. So,

• The phase margin therefore represents the additional amount of phase lag required to destabilize the system, just as the gain margin represents
the additional gain for destabilization. Phase margin > 300 for a stable system in designing.

Page 50
7. Advanced Control Strategies
13 August 2021 15:31

Feedforward Control

A feedback controller responds only after it detects a deviation in the value of the controlled output from its desired set point. On the other hand, a
feedforward controller detects the disturbance directly and takes an appropriate control action in order to eliminate its effect on the process output.
The feed forward controller utilizes the value of the reference as well as the value of the disturbance to compute the input. On observing it carefully, we
can see that it is a form of open-loop control, there is no feedback from the output but rather the two inputs are sensed and a control action is provided.

Its effectiveness depends on how the set point and the disturbance affects the output i.e. for accurate measurement of the result, we should know the
process models very accurately because any kind of modelling error will affect the output directly.

But, in feedback control, the accuracy of the models is not highly required because the output is measured actually and then, the controller takes the
action. That is why feedback controllers are mostly used in plants.

Consider a tank and there is some fluid which is flowing at the rate of Fi into the tank at temperature Ti and there is an outlet flow F and there is a
temperature T which we want to maintain by changing the steam flow through valve.

Now, we want the temperature T inside the tank to be exactly constant. We should manage the temperature Ti (disturbance) so that, T = TSP inside the
tank which is governed by the following model:

So, if this model contains some error in any of its parameter either F , ρ or C then, the corresponding error will be reflected in the output also due to

Page 51
So, if this model contains some error in any of its parameter either Fi, ρ or CP then, the corresponding error will be reflected in the output also due to
which the set point will not be maintained.

Consider the following system,

For a perfect feed forward control eliminating complete disturbance,

ࡳࢊ
ࡲࢋࢋࢊ ࢌ࢕࢘࢝ࢇ࢘ࢊ ࢉ࢕࢔࢚࢘࢕࢒࢒ࢋ࢘ ࢌ࢛࢔ࢉ࢚࢏࢕࢔ሺࡳࡲࡲ࡯ ሻ ൌ ۛۛۛۛۛۛ
ࡳ ࡳ ࡳ ࢓ ࡲ ࡼ

Cascade Control

Consider the single-loop control of a jacketed kettle as shown below. The system consists of a kettle through which water, entering at temperature Ti is
heated to T0 by the flow of hot oil through a jacket surrounding the kettle. The temperature of the water in the kettle is measured and transmitted to the
controller, which in turn adjusts the flow of hot oil through the jacket. This control system is satisfactory for controlling the kettle temperature; however, if
the temperature of the oil supply should drop, the kettle temperature can undergo a large prolonged excursion from the set point before control is again
established because the controller does not take corrective action until the effect of the drop in oil supply temperature has worked itself through the
system of several resistances to reach the measuring element.

Page 52
 Single loop control

The reason is that the controller does not take corrective action until the effect of the drop in oil supply temperature has worked itself through the
system of several resistances to reach the measuring element. To prevent the sluggish response of kettle temperature to a disturbance in oil supply
temperature, the control system shown in below is proposed.

Cascade Control

In this system, which includes two controllers and two measuring elements, the output of the primary controller is used to adjust the set point of a
secondary controller, which is used to control the jacket temperature. Under these conditions, the primary controller indirectly adjusts the jacket
temperature. If the oil temperature should drop, the secondary control loop will act quickly to maintain the jacket temperature close to the value
determined by the set point that is adjusted by the primary controller. The primary controller is also referred to as the master controller, and the
secondary controller is referred to as the slave controller.

Page 53
secondary controller is referred to as the slave controller.

A simplified block diagram of the single-loop system is shown below as:

Below is a block diagram representation of the cascade control system, shows clearly that an inner loop has been added to the conventional control
system.

Analysis of Cascade Control

Consider the following closed loop system,

So,

Simplifying,

Page 54
Similarly,

or,

If secondary loop is absent,

and,

Simplifying,

Response Time of a cascade controller

In absence of secondary loop and the slave controller,

Time constant ࣎࢖૛

Page 55
In presence of the secondary loop and the slave controller (say a pure gain controller)

So, there is a decrease in time constant in presence of slave loop. This will guarantee that the time constant of the overall process decreases when a
secondary loop is introduced to the system. Eventually that decreases the overall response time.

Page 56
8. Controller Tunning
14 August 2021 16:49

The selection of a controller type (P, PI, PID) and its parameters (K, ࣎ࡵ , ࣎ࡰ ).The adjustment of the controller parameters to achieve satisfactory control is
called tuning. The selection of the controller parameters is essentially an optimization problem in which the designer of the control system attempts to
satisfy some criterion of optimality, the result of which is often referred to as “good” control.

For each response curve, the process was subjected to a unit-step change in load (U = 1/s), and the controller parameters were selected by tuning
rules to be presented later.

PROPORTIONAL CONTROL. As shown above, proportional control produces an overshoot followed by an oscillatory response, which levels out at a
value that does not equal the set point; this ultimate displacement from the set point is the offset.

PROPORTIONAL-DERIVATIVE CONTROL. For this case the response exhibits a smaller overshoot and a smaller period of oscillation compared to
the response for proportional control. The offset that still remains is less than that for proportional control.

PROPORTIONAL-INTEGRAL CONTROL. In this case, the response has about the same overshoot as proportional control, but the period is larger;
however, the response returns to the set point (offset = 0) after a relatively long settling time due to which we obtain a sluggish response as the order of
the system gets increased. The most beneficial influence of the integral action in the controller is the elimination of offset.

Criteria for good control

• Integral of Square of Error (ISE)

The objective of the designer is to obtain the minimum value of ISE by proper choice of control parameters. A response that has large errors and
persists for a long time will produce a large ISE.

The objective of the designer is to obtain the minimum value of ISE by proper choice of control parameters. A response that has large errors and
persists for a long time will produce a large ISE.

Page 57
 • Integral of Absolute value of Error (IAE)

• Integral of Time weighted Absolute Error (IATE)

Zeigler - Nichols Tuning

Here,
૛࣊
PU = ۛۛۛ (sec/cycle)
࣓࡯ࡻ
૚
ۛۛ
KU = ࡭ࡾ

Page 58
9. Control Valves
14 August 2021 13:50

There are 2 types of valves based on the action:

• Air to Close or Fail to Open (Direct action)

 As input ↑, output also ↑

 As pressure of air ↑, stem will move down and the valve will close.
 In case of power cut, the valve will get OPENED.

Example: Coolant flow in a reactor

 Air to Open or Fail to Close (Indirect action)

 As input ↑, output ↓

Page 59
  As pressure of air ↑, stem will move down and the valve will open.
 In case of power cut, the valve will get CLOSED.

Example: Fuel supply in a reaction.

Note: The controller action will be just reverse of the valve action

To carry out smooth motion of the valve stem, A valve positioner is a device that can be attached to a valve that drives the valve to the desired
position, in spite of friction or hysteresis.

Valve Sizing

To specify the size of a valve in terms of its capacity, the following equation is used:

where,

q = flow rate, gal/min

f (x) = fraction of maximum flow ( = 1 for fully open)
x = fractional stem position (i.e., fraction open)
∆Pvalve = pressure drop across the valve, psi
sg = specific gravity of fluid at stream temperature relative to water; for water sg =1
CV = Factor associated with valve capacity

The following is also used by some manufacturers:

where,

q = flow rate, gal/min

∆Pvalve = pressure drop across the valve, psi
sg = specific gravity of fluid at stream temperature relative to water; for water sg =1
KV = Valve coefficient = 0.856CV

Valve Characteristics

Page 60
Valve Characteristics

Inherent valve characteristics (pressure drop across valve is constant).

I: Linear
II: increasing sensitivity (e.g., equal-percentage valve),
III: decreasing sensitivity (e.g., square root valve)

Let,

where,
qmax = maximum flow when valve is fully open (stem is at its maximum lift Lmax)
x = fraction of maximum lift
f = fraction of maximum flow

In general,

The types of valve characteristics can be defined in terms of the sensitivity of the valve, which is simply the fractional change in flow to the fractional
change in stem position for fixed upstream and downstream pressures; mathematically, sensitivity may be written as:

Page 61
In terms of valve characteristics, valves can be divided into 3 types:

• Decreasing sensitivity (square root or quick-opening)

ۛۛ
f(x) = √ ࢞
• Constant sensitivity (linear opening)

Assuming that the valve is shut tight when the lift is at lowest position, we have f = 0 at x = 0 and fully open when the lift is at highest position, f = 1 at
x = 1. So,

Integrating this equation and inserting limits give

So,
f(x) = x
• Increasing sensitivity (equal-percentage)

where, β = constant
So,

or,

where, f0 is the flow at x = 0. On rearranging we get,

f = f0eβx
The term β can be expressed in terms of f0 by putting f = 1 at x = 1 as:
β = ln(1/f0)
So,

Page 62
To express the range over which an equal-percentage valve will follow the equal-percentage characteristic, the term rangeability is used.
Rangeability is defined as the ratio of maximum flow to minimum controllable flow over which the valve characteristic is followed. It is also known as
Turn-down ratio (= 0.7 X Rangeability).

Types of plunges of a pneumatic valve

Page 63
10. Instrumentation
14 August 2021 18:13

1. Flow measurement
• Orificemeter • Turbine flowmeters • Hot wire/film Anemometer
• Venturimeter • Vortex flowmeters • Laser Doppler Velocimetry
• Variable area flowmeters (Rotameter) • Ultrasonic flowmeters • Electromagnetic flowmeters

2. Temperature measurement
• Thermocouple: Works on Seebeck effect • Optical Pyrometers
• Thermometer • Radiation Pyrometers
• Bimetallic Strips • Resistance Temperature Detector(RTD)
• Thermistor

3. Pressure measurement
• Strain gauge • Bellow Element • Piezoresistive Transducer
• Differential Pressure Cells • Diaphragm Element • Piezoelectric Transducer
• Liquid Column Devices • Bourdon Tube Element • Bellow Resistance Transducer

4. Level measurement
• Ball float • Linear Variable Differential Transducer(LVDT)
• Capacitance type level meter • Magnetic bond level meter

5. Composition measurement
• Polarograph • Paramagnetism • Psychrometer • Electrical Conductivity
• pHmeter • UV & Visible Radiation analyzers • Interferometer • Infrared analyzers
• Mass Spectroscopy • X ray diffraction • Hygrometers • Gas Chromatography

Sensors & Transducers

• A sensor is a device which detects the change in a physical quantity and gives the output in the same format

Page 64
 • A sensor is a device which detects the change in a physical quantity and gives the output in the same format

• A transducer is a device which converts the non - electrical signals (temperature, pressure, level, displacement etc.) into electrical signals
(current, voltage, etc) which can be measured.

Note: All of the above measuring instruments possess second order dynamics.

Signal conditioning: It is the conversion of the sensed signal into an analogue or digital value.

 Specifications of Sensors & Transducers

1. Range: It indicates the limits within which the input can vary. For example: A thermocouple can work in the range of 25 0C - 225 0C.

2. Span: It is the difference between the maximum and minimum values of the input. For example: the above thermocouple has a span 200 0C.

3. Error: It is the difference between the result of the measured value and the true value of the quantity being measured.

4. Accuracy: It is the closeness of the agreement between the measured value and the actual value. It is often expressed as the % of full scale
deflection or full range output.

5. Sensitivity: It is the change in output for a given change in input. It represents the slope of a transducer.
࡯ࢎࢇ࢔ࢍࢋ ࢏࢔ ࢕࢛࢚࢖࢛࢚ ࢜ࢇ࢒࢛ࢋ ࢕ࢌ ࢙ࢋ࢔࢙࢕࢘
Sensitivity = ۛۛۛۛۛۛۛۛۛۛۛۛۛۛۛۛۛۛۛۛۛۛۛۛۛۛۛۛۛۛۛۛۛۛۛۛ
࡯ࢎࢇ࢔ࢍࢋ ࢏࢔ ࢏࢔࢖࢛࢚ ࢜ࢇ࢒࢛ࢋ ࢊ࢛ࢋ ࢚࢕ ࢝ࢎ࢏ࢉࢎ ࢕࢛࢚࢖࢛࢚ ࢉࢎࢇ࢔ࢍࢋ࢙

6. Non - linearity: It is the maximum deviation of actual measured curve of a sensor from the ideal. Linearity is often expressed as % of non -
linearity which is defined as:
ࡹࢇ࢞࢏࢓࢛࢓ ࢊࢋ࢜࢏ࢇ࢚࢏࢕࢔ ࢏࢔ ࢏࢔࢖࢛࢚
% non - linearity = ۛۛۛۛۛۛۛۛۛۛۛۛۛۛۛۛۛۛۛۛ
ࡹࢇ࢞࢏࢓࢛࢓ ࢌ࢛࢒࢒ ࢙ࢉࢇ࢒ࢋ ࢏࢔࢖࢛࢚

7. Hysteresis: It is the error/deviation of the sensor's output at any given point when approached from 2 different directions. It is caused due to
loose linkages, magnetization and thermal properties.

8. Resolution: It is the smallest detectable incremental change of input parameter that can be detected in the output signal. It can be expressed in
absolute terms or as a proportion of the full - scale reading.

9. Stability: The ability of a sensor to give same output when used to measure a constant input over a period of time. The term 'drift' is used to
indicate the change in output that occurs over a period of time . It is expressed as % of full range output.

10. Dead Time: It is the time duration from the application of an input until the output begins to respond or change.

Page 65
11. Piping & Instrumentation Diagram Symbols
14 August 2021 21:44

Page 66
Page 67
Page 68