Process Control Theory and

Instrumentation
Lecture-1

MM 451
 MM401 course coverage
1. Control theory
– Introduction to classical control theory
– Behaviour of control systems
2. Discrete State Control or Automation
3. Instrumentation (process instrumentation)
– Sensors
– Instrumentation examples from process plants
– Piping and Instrumentation Diagrams (P&ID)
 Control systems
Controller + Controlled process  Control system
A control system provides the desired response by
controlling the output of a process.

Block diagram representation

Transforms input to desired output

 Type of Control Systems
Manual control - Example
Automatic control - Examples
Open loop control – Toaster, Heat treatment
Closed loop control - Structure
Feedback control – Example Income model
1. Closed loop negative feedback Manual
control
2. Closed loop negative feedback Automatic
control
 Level control of a water tank
Manual control
Manual level control Automatic control
Open loop control
Closed loop control
Feedback control

Purpose of control
Maintain desired operation
Desired changes:
• flow rate requirement
Unknown changes:
• Changing input flow rate
• Valve characteristics
• Change in back pressure
• Fluid viscosity
Automatic control of level in a tank
Manual control
Automatic control
Open loop control
Closed loop control
Pivot Lever arm Feedback control

Valve
Float

Automatic control with Float and Valve

 Examples of open loop control
Manual control

• Bread toaster Automatic control

Open loop control
Closed loop control
Heat for Feedback control
Insert Toasted
a fixed
bread bread
time
• Heat treatment

Heat to Hold temp Cool at a

Insert Treated
a given for a fixed fixed
object object
temp time rate
 Closed loop control
Manual control
Automatic control
Open loop control
Closed loop control
Feedback control

Desired Actual
Controller Process
Output Output
Respon Response
se
Measurement
Device

Block diagram representation

 Controlling direction of travel
Desired
+ error Actual
course of Steering
travel Driver Automobile course
mechanism
-
Actual course of Measurement
travel visual and
tactile
Manual feedback
control
Performance of controller

Direction of travel response

Feedback control (model of national income)
Model of national income
Model of national income
Common terms in control systems
• Desired value of controlled variable
or setpoint
• Measured variable
• Error from desired value
• Control action
• External disturbance
• Controller
• Process
• Feedback elements
 Process control principles
Desired Actual
course of course of
travel travel

External
Controller
disturbance Controlled
Error Control action Process
Set Measured
point variable

Feedback
Feedback elements
Block diagram representation of a generic control system
Type of control actions

Control action
 An on-off controller
Sensor

Controller Setpoint

Actuator

Heating Process
Power
supply

Observed temperature is lower than set point: Switch ON

Observed temperature is higher than set point: Switch OFF
 On/Off ON-OFF
with max. power control
control

Set
point
Controlled variable

100%

Maximum power = 50%

50%
Control
action
0%
Time
On off control with a step change

Setpoint

ON

OFF
 On off control Action (external
ON/OFF or Two step control
disturbance)
Overshoot
Set point
Undershoot

Control action
100%
Off time
Power

On time
0%

 
Duty cycle =
 Multimode On/Off control
• Multimode/multi-position controller is a logical extension
of on/off controller
• Provides several intermediate settings of controller output
• Reduces cycling and over/undershoot behaviour
Control action

100%

m m = 100% when e > e2

50% m = 50% when e2 < e < e1
m = 0% when e < e1
0%
e1 e2 Error(e)
On-off controller with hysteresis
Hysteresis or dead-band (overlap)
 More control actions
• Proportional control (P)
• Integral control (I)
• Derivative control (D)
• Mixed control action
• PI
• PD
• PID
 Control actions
Proportional
Control action µ Error
m(t) µ e(t)
m(t) = Kp . e(t)
Specifically, m(t) = Kp . e(t) + m(0)
Where m(0) is controller output for ‘no error’
Main characteristics:
• Smooth action (absence of abrupt switching on/off)
• Presence of offset error …………… described ahead
 Proportional control
Set point(SP)
Controlled variable(CV)

Offset error

100
e(t) = SP - CV
m m(t) = KP . e(t) + m(0)
m(0)
0
Time
Proportional control

Increasing gain
reduces offset

m(t) = KP . e(t)+ m(0)

 Control actions
•Integral
 
Control action µ Error

Main characteristics:
• Oscillatory behaviour
• Long settling time
• No offset error
 Integral control
2.5

2
Temperature

1.5

0.5

0
0 10 20 30 40
Time50 60 70 80 90
 Control actions
Derivative control

Main characteristics:
• Can not be used in isolation
• Why?
 Mixed control actions
Proportional + Integral (PI)

Main characteristics:
• Reaches final value quickly
• Reduced overshoot/undershoot compared to
Integral control action
• No offset error
Behaviour of error in PI controller
Error reset rate
e(t) = Set point – controlled variable

Offset error

When KI is large control action

approaches that if integral control
 Mixed control actions
Proportional + Derivative (PD)

+ KP .e(t)

Main characteristics:
• Reaches final value slowly
• None or little overshoot/undershoot
• Offset error
PD control

+ KP .e(t)
 Mixed control actions
Proportional + Integral + Derivative (PID)

Main characteristics:
• Reaches final value quickly
• Reduced overshoot/undershoot compared to I, or PI
• No offset error
PID control
Controller output in Proportional control

(+) e(t)
Error

t
0
Time
(-)

m = KP.e(t) + m(0)
m (%)
Controller output

100 m(0) = 50%

KP.e(t)

50

t
0
Time
Controller output in Integral control
10

(+) 5
Error

00 0 5 10 15 20 25 30 35 40 45

(-) -5
-10 Output floating at 0 error
Controller output (m)

-15

 
m=
𝐾 𝐼∫ 𝑒 ( 𝑡 ) .𝑑𝑡
 

0 5 10 15 20 25 30 35 40 45
Controller output in Derivative mode control

(+) e(t)
Error

t
0
Time
(-)
Reverse acting control action
 
m = + m(0)
m (%)
m(0) = 50%
Controller output

100

50
 
¿¿
t
0 0 Time
Controller output in a Reverse acting PD
controller
e(t) (%) =setpoint - controlled output
(-)
Error
t
0
Time
(+)
At zero error Offset
why control
 
action is finite m = KP.e(t) + + m(0)
m (%)
m(0) = 50%
Controller output

100

50 Kp. Offset
Proportional response
Derivative response t
0
Time
Structure of part-1 of the course
3
1 2
Controller Process

Measurement
Device

1. Learn about various control actions

2. Learn about dynamic process models
3. Learn about behavior of control systems
 Next PPT on systems modelling
(File name: Mechanical components)