Practical Loop-Tuning in Mineral Processing
www.mipac.com.au
Module 1
An Introduction
1
Module 1
Practical Loop Tuning (An Introduction)
www.mipac.com.au
What is a Control Loop?
Fly Ball Governor
(Watt, 1788)
Watt Steam Engine
Not shown in Notes (Watt, 1781)
2
Module 1
James Watt
www.mipac.com.au
The Industrial Revolution in Europe could not have taken place
without the work of James Watt, who is commonly credited with
inventing the steam engine……..
In 1788 Watt unveiled another major contribution to the realm of
instrumentation and machine control by introducing the centrifugal
governor. Since a similar device had been in use in windmills, Watt
made no effort to apply for a patent; still, it became commonly known
as the Watt governor. Its importance centered around its automatic
control of steam output. The governor consisted of two metal spheres
mounted on a vertical rod that was spun by the engine's output of
steam. The faster the rod spun, the farther the two spheres were
thrown outward by centrifugal force. The farther the spheres were
thrown, the more they choked off the steam outlet. As steam output
decreased, so did intake as well as the engine's power output. As
power and steam output decreased, the slower the spheres rotated, re-
opening the steam outlet and beginning the reverse of the process.
Born 1736 Engine output and speed could be controlled and adjusted between
Died 1819 these two limits.
Source: http://www.madehow.com/inventorbios/68/James-Watt.html
3
Module 1
General Process Hierarchy
www.mipac.com.au
Plant
Optimisation
Advanced Control Optimise
Process
Optimising Control
Loop Control Regulatory Control
Field/Panel/DCS/PLC
Stabilise
Measurement
Manual Control
Instrumentation – Inputs/Outputs
Processes
Economic Return
4
Module 1
Mineral Processing Example
www.mipac.com.au
Level Setpoint
Controlled Manipulated
Variable Variable
FI LC
DCS / PLC or
Single Loop Level
Disturbance or Controller Controller
Load Variable
Magnetic Level
FT Flowmeter Transmitter LT
(Measurement)
Feed
P
Tail
Pinch or Dart
Valve with
Flotation Bank Positioner
(Actuator)
(4 Cells) Concentrate
5
Module 1
Essential Control Loop Elements
www.mipac.com.au
Five Essential Control Loop Elements
1. Sensing Element
2. Transmitter
3. Controller
4. Final Control Element
5. Process
Only when all five elements are performing at
their best will the control system meet
expectations.
6
Module 1
Feedback and FeedForward Control ?
www.mipac.com.au
Temperature
Setpoint = T0 Watch that Feed
Flow
Temperature Temperature
Feedback Feedback Temperature
Temperature
Indicator Indicator
Steam Hot Steam Hot
Water Feedforward Water
Water Signal Water
Heater Cool Heater
Cool
Water Flow
Water Indicator
Feedback Control FeedForward Control
(with Feedback Trim)
Control loop response must, essentially, reflect process response
for successful control of the process
7
Module 1
Proportional-Integral-Derivative
Independent Controller Structure
www.mipac.com.au
𝑑𝐸
𝑂𝑢𝑡= 𝐾 𝑃 𝐸+𝐾 𝐼 ∫ 𝐸𝑑𝑡+ 𝐾 𝐷
𝐷𝑡
Proportional
Output
Setpoint Gain
E (0-100%)
+ +
n KP n
+
- +
f
Integral
Process Variable
K
Derivative
8
Module 1
Proportional-Integral-Derivative
Dependent Controller Structure
www.mipac.com.au
𝑂𝑢𝑡= 𝐾 𝐶 𝐸+
[ 1
𝐼
∫ 𝐸𝑑𝑡 +𝐷
𝑑𝐸
𝑑𝑡 ]
Proportional
Setpoint Gain Output
+ E +
(0-100%)
n KC n
+
- +
Integral
Process Variable 𝑑
𝐷
𝑑𝑡
Derivative
9
Module 1
Proportional-Integral-Derivative
PCS7 PIDConL
www.mipac.com.au
Parameters
Proportional Gain
Integral TI (sec)
Derivative Td sec)
10
Module 1
Proportional-Integral-Derivative
PCS7 PIDConL Useful Parameters
www.mipac.com.au
Parameter Description Default
PropSel 1 = Activate P component 1
IntSel 1 = I component activated 1
1 = D component is placed in
DiffToFbk 1
the feedback
Applying the P component to
PropFacSP 1.0
the feedback [0..1].
0 = P component fully in
feedback
Gain of differentiator [1..10]
DiffGain 5.0
DiffGain = TD /(delay time of
D component)
Process value range (PV) for
NormPV 100.0, 0.0
standardizing the proportional
gain (GAIN)
STRUCT
● High: REAL
● Low: REAL
Manipulated variable range
NormMV 100.0, 0.0
(MV) for standardizing the
proportional gain (GAIN)
PropSel, IntSel, DiffSel, DiffToFbk,
STRUCT PropFacSP, DiffGain, NormPV, NormMV
● High: REAL
● Low: REA
11
Module 1
External Reset Controller Structure (PID)
www.mipac.com.au
External Integral Feedback version
(Bias or Manual Reset)
Error Manipulated
Setpoint Variable
+ +
G
- (Proportional + High / Low Integral loop can be broken here for:
Gain) Limits + A/M • Deadtime compensation
• Cascade control
• Override control etc.
I
(Integral Time)
Controlled
Variable
D
(Derivative Lead Time) ( G, I, D, Bias) - Controller Tuning Constants
12
Module 1
Proportional-Integral-Derivative
PCS7 PIDConR External Reset
www.mipac.com.au
Parameters
Proportional Gain
Integral TI (sec)
Derivative Td (sec)
13
Module 1
Why the Different Controller Structures?
www.mipac.com.au
• For simple loops both controllers produce identical results:
- Same tuning constants
- Same tuning methods can be used
• The External Integral Controller has advantages, for example:
- Control of deadtime dominant processes
- Cascade control
- Override control
• Module 4 uses the External Integral Controller for deadtime
compensation
14
Module 1
Proportional Only Control
(Direct Acting)
www.mipac.com.au
Manipulated Variable
G=1
Step Change
Controlled Variable
Fixed Bias Setpoint
Time
15
Module 1
Proportional+Integral Control
(Direct Acting)
www.mipac.com.au
Manipulated Variable
Proportional Action
is repeated
Integral Time
Proportional
Action
Step Change
Controlled Variable
Setpoint
Time
NB: The Proportional Gain for a PI controller does not equal the Proportional Gain for a P only controller
16
Module 1
Proportional+Derivative Control
(Direct Acting)
www.mipac.com.au
Manipulated Variable
PD Control Proportional Action
is “Leaded”
P only Control
Derivative
Action Controlled Variable
Derivative Lead Time
Ramp Change
Setpoint
Time
17
Module 1
Proportional+Integral+Derivative Control
(Direct Acting)
www.mipac.com.au
RAMP Manipulated Variable
STEP Manipulated Variable
Controlled Variable Controlled Variable
Ramp Change
Step Change
Setpoint Setpoint
Time Time
18
Module 1
Process Types - Self Regulating
www.mipac.com.au
Self Regulating Process – Reaches an equilibrium state
19
Module 1
Process Types - Integrating
www.mipac.com.au
Integrating Process – Increases at a steady rate
20
Module 1
Process Response - First Order
www.mipac.com.au
First Order System
Process Gain = , Dead Time = , Time Constant =
Gp = p%/m%
63.2% p
m
td
21
Module 1
Laplace Transform
www.mipac.com.au
• Laplace transforms are used to simply
define process responses.
• For example a first order response is defined
as:
Lead Time
Dead Time
Gain
Lag Time
22
Module 1
Laplace Transform
www.mipac.com.au
• Process Gain (G or Kp) is the final change in
the process variable (PV) per increment in
the manipulated variable (MV)
• Dead Time dt is the time elapsed between
the MV change and the start of the PV
response.
• Lag time or time constant and reflects the
time taken for the PV to reach 63.2% of the
final value.
23
Module 1
Process Response - Multiple Order
www.mipac.com.au
Second Order System
Third Order System
1 = 5, 2 = 10
=5
1 = 5, 2 = 10, 3 = 15
Apparent Dead Time
24
Module 1
Process Response - Integrating
www.mipac.com.au
Integrating System
∆𝑝
𝑟 𝑖=
∆ 𝑚×∆ 𝑡
p
m
t
td
25
Module 1
Regulatory Control Techniques
www.mipac.com.au
1. Cascade Control
2. Ratio Control
3. Feed Forward Control
4. Override Control
5. Linearisation
6. Split Range Control
7. Filtering
26
Module 1
Regulatory Control Techniques
www.mipac.com.au
pH Setpoint
Non-linear Error (9.4 - 9.8)
Characterisers
Ratio AC
pH Controller
Ratio ( master loop)
Controller pH Setpoint
Ratio Cascade
X Setpoint
AC
Linearisation pH Controller Agitator Motor
(slave loop)
pH pH
Tank
Level
Feed
Valve Setpoint
Characteriser
FT Feed
Slurry Feed LC Well
Tank
Discharge
P
LT LAG
Lime Slurry
V/S
Pinch Valve
with Positioner
(Linear Characteristic) Leach Tank or
Variable Speed Baffles Flotation Cell Baffles
Pump
For a flotation cell the pH
probe is located in the feed box
27
Module 6
Regulatory Control Techniques
www.mipac.com.au
FR
Ratio Flocc. Ratio
WC
f(t) FC
Flocc. Rake
Flow Torque
FT NT
Flocculant LC
Slurry Flow Bed
FT
Level
Slurry Feed LT
Overflow
Clarified Water
Turbid Water Override
Bed Layer
Bed
Feedforward Mass
PT PC > DC Underflow
Density
f(t) Underflow
Flow
Feedforward FC
Model
SC FT DT
Underflow
28
Module 6
Proportional-Integral-Derivative
PCS7 Override Control PidConL
www.mipac.com.au
29
Module 1
Proportional-Integral-Derivative
PCS7 Override Control PidConR
www.mipac.com.au
30
Module 1
Proportional-Integral-Derivative
PCS7 Cascade Control
www.mipac.com.au
31
Module 1
Proportional-Integral-Derivative
PCS7 Cascade Control PidConR
www.mipac.com.au
32
Module 1
Controller Tuning
Overview
www.mipac.com.au
Prerequisites to commencing Loop Tuning
1. Know your Control System.
PID algorithm type, PV filtering, Integral Windup prevention, Output limiting.
2. Know the Process.
Load response or Setpoint response, Interacting Loops, Measurement
issues.
3. Review Control Strategy.
Opportunity for Feedforward, Ratio Control, Cascade, Output Splitting.
4. Check Instrumentation.
Instrument installation, Instrument calibration, Actuator condition, VSD setup.
5. Linearisation
Response to equal step sizes across the operating range, Valve
Characterisation, Gain scheduling.
33
Reference Books and Papers
www.mipac.com.au
1. Shinskey, F. G., Process Control Systems – Application Design and Tuning, McGraw-Hill
New York, 1996.
2. Lipták, B. G., Instrument Engineers Handbook 3rd Edition – Process Control,Chilton,1995.
3. St. Clair, D. W., Controller Tuning and Control Loop Performance, Straight-Line Control
Company inc, 1996.
4. Ziegler, J. G., and Nichols, N. B., “Optimum Settings for Automatic Controllers,” Trans
ASME, November 1942.
5. Rockwell Automation, Using the PIDE Instruction, White Paper
6. Jacques F. Smuts, Process Control for Practitioners, OptiControls, 2012
34
Module 1
Terminology for Controller Tuning
www.mipac.com.au
• Terminology is very confusing .
• No real attempt to standardise.
• One of the biggest barrier to learning process control or loop-tuning
• Table in notes attempts to list and explain commonly used terms (See
Page 11 of Module 1).
• Reference 3 has an excellent list of terms.
• Prior to tuning, ensure the instrumentation is in good working condition
and properly calibrated.
• PID Structure differs between control systems. It is Important to know
what you are dealing with!
35
Module 1