1. PSE and PROCESS CONTROL
Sigurd Skogestad
Department of Chemical Engineering
Norwegian University of Science and Tecnology (NTNU)
Trondheim, Norway
PSE Education Session
AMIDIQ 2012, Mexico, May 2012

2. Process control is an important subject in the chemical engineering education. In addition to covering feedback control, it is often the only course in the curriculum that includes aspects of process operation and process dynamics. It can also be a difficult course to teach, especially if this is the only course, because there are many topics and concepts that one would like to include, some of which are Control crash course (3 weeks): 1. Process operation: Why do we need process control? 2. Classification of variables (inputs, outputs, disturbances, measurements) 3. Feedback versus feedforward control 4. Block diagram representation (information diagrams, causality) 5. Flowsheet representation (process & instrumentation diagrams) 6. Single-loop control: Pairing of input and outputs 7. More advanced control: Ratio control, Cascade control, 8. The control hiearchy (optimization, advanced control, basic control) 9. Process dynamics (basics): first- and second order systems, time delay, identification 10. Process modelling: balance principle 11. PID control and tuning 12. Simulation Control theory (10 weeks): 13. Laplace transforms, transfer functions 14. Closed-loop response, derivation of PID tuning rules 15. Pros and cons of high gain feedback. Stability. Change dynamics. Biological systems 16. Dynamic systems (theory). poles, zeros, state space, observability, controllability 17. Control systems (theory), frequency analysis, stability conditions, robustness 18. Controller implementation: discrete control, windup, bumpless transfer 19. Identification (theory) 20. Multivariable control: interactions, MPC I have listed the topics approximately in the order I teach them in my course. To make sure the students understand what the theory is going to be used for, I teach the 12 first topics as a 3-week "crash course". Usually, some of the theoretical material (topics 13-20) has to be deleted, and the question is which? For example, can we omit Laplace transforms and frequency analysis? This is tempting, but it also makes it difficult to derive tuning rules analytically and to really understand feedback

3. PROCESS CONTROL
Theory
Left side of brain = logical

4. PROCESS CONTROL Control structures + Practise
Right side of brain = creative

5. PROCESS CONTROL
Theory & practise
Combine both sides!

6. Process control course. Four main elements:
Process dynamics: Step responses, simulation
Process control structures: Flowsheet (P&ID*). PID tuning
CONTROL
theory: Feedback idea, block diagrams, stability, transfer functions (Laplace), feedforward/cascade/frequency response, identification, multivariable control (MPC)
PRACTISE
Laboratory
Simulation (Aspen, Hysys/Unisim..)
SYSTEMS
Modelling principles, Solution. State space models, linearization (ABCD), optimization
*P&ID: Process and Instrumentation Diagrams

7. Difficult course Inputs and outputs, causality Feedback Stability
Many new concepts
Inputs and outputs, causality
Feedback
Stability
New mathematics
Laplace
Frequency analysis
System theory (ABCD)
And all of this combined with practise: operation of real plants
Too much for one course?

8. I teach the course in two parts
”Process control” crash course (3 weeks)
Focus on process control structures (P&ID)
Standard process control course (11 weeks)
Focus on theory

9. Crash course process control
Sigurd Skogestad
Institutt for kjemisk prosessteknologi
Rom K4-211
More information (literature, old exams, etc.):

10. Why control? Actual value(dynamic) Steady-state (average) time
Until now: Design of process. Assume steady-state
Now: Operation
Actual value(dynamic)
Steady-state (average)
time
“Disturbances” (d’s)

11. Example: Control of shower temperature
MVs, CVs and control

12. CLASSIFICATION OF VARIABLES
flow in Hs H LC
flow out
OUTFLOW: INPUT FOR CONTROL
INFLOW: DISTURBANCE

13. BLOCK DIAGRAMS u y d ys ys-ym ym measured output
Process
(shower) u
input (MV) y
output (CV) d
Controller
(brain)
Measurement
device ys
Desired value
Setpoint
ys-ym
error ym
measured output
FEEDBACK (measure output):
Controller
(brain)
Process
(shower)
Measurement
device
FEEDFORWARD (measure disturbance): dm
measured disturbance d u
input (MV) y
output (CV)
All lines: Signals (information)
Blocks: controllers and process
Do not confuse block diagram (lines are signals) with flowsheet (lines are flows); see below

14. Most important control structures
Feedback control
Ratio control (special case of feedforward)
Cascade control

15. Process and instrumentation diagram (P&ID) (flowsheet)Ts
(setpoint CV) T
(measured CV) TC
MV (could be valve)
2nd letter:
C: controller
I: indicator (measurement)
1st letter: Controlled variable (CV). What we are trying to control (keep constant)
T: temperature
F: flow
L: level
P: pressure
DP: differential pressure (Δp)
C: composition
X: quality
H: enthalpy/energy

16. Typical distillation control:
Two-point composition control
LV-configuration with inner T-loop LV CC xD Ts TC CC xB

17. Process dynamics (response)
“Things take time”
Step response (step in u):
k = Δy(∞)/ Δu – process gain
 - process time constant (63%)
 - process time delay
Time constant : Often equal to residence time = V[m3]/q[m3/s] (but not always!)
Can find  (and k) from balance equations:
Rearrange to match standard form of 1st order linear differential equation:

18. Pairing of variables Main rule: “Pair close”
The response (from input to output) should be fast, large and in one direction. Avoid dead time and inverse responses!

19. Model-based tuning (SIMC rule)
k = Δy(∞)/ Δu
From step response
k = Δy(∞)/ Δu – process gain
 - process time constant (63%)
 - process time delay
Proposed SIMC controller tunings

20. Process Control crash course (3 weeks):
1. Process operation: Why do we need process control?
2. Classification of variables (inputs, outputs, disturbances, measurements)
3. Feedback versus feedforward control
4. Block diagram representation (information diagrams, causality)
5. Flowsheet representation (process & instrumentation diagrams)
6. Single-loop control: Pairing of input and outputs
7. More advanced control: Ratio control, Cascade control,
8. The control hiearchy (optimization, advanced control, basic control)
9. Process dynamics (basics): first- and second order systems, time delay, identification
10. Process modelling: balance principle
11. PID control and tuning
12. Simulation

21. Control theory (11 weeks) “standard course”
13. Laplace transforms, transfer functions
14. Closed-loop response, derivation of PID tuning rules
15. Pros and cons of high gain feedback. Stability. Change dynamics. Biological systems
16. Dynamic systems (theory). poles, zeros, state space, observability, controllability
17. Control systems (theory), frequency analysis, stability conditions, robustness
18. Controller implementation: discrete control, windup, bumpless transfer
19. Identification (theory)
20. Multivariable control: interactions, MPC

22. + 3. Practise At least have demonstration Time consuming LAB ?!!
SIMULATIONS ?!!
Time consuming

23. + 4. Systems engineering General modelling principles, DAE-system
Solution of dynamic models (integration)
Linearization, State space models (deviation variables)
Optimization

24. Conclusion: Process systems engineering (PSE) and process control
Process control is a key course
Engineers must know some control!
Usually too little time to focus on systems issues
Need advanced course to cover process systems aspects of process control