CHEE319

1
CHEE319

• Process Dynamics and Control
• J. McLellan
• Winter 2005

2
Outline - Module 1

• definitions and examples
• components of a control system
• why is control required?
• the systems approach
• defn of feedback control
• control engineering
• control objectives
• improving operation - the operating window and
  control benefits analysis

3
Dynamics and Control

• dynamics - time varying behaviour of a process
  vs. static, steady-state behaviour
• control - maintain desired conditions in a system
  by adjusting selected variables in the system

4
Example - Shower
Hot
Cold
5
Example - Shower
FI
TI
6
Example - Shower

• controlled variables - variables to be regulated
• total flow, temperature
• manipulated variables - variables which are
  adjusted
• hot water tap
• cold water tap

7
Example - Tank Lab Experiment

• controlled variables
• level
• concentration
• inferred by conductivity
• manipulated variables
• concentrated salt flowrate
• exit flowrate
• fractional pump rates

8
Example - Tank Lab Experiment
water
brine soln
LI
effluent
AI
9
Example - Tank Lab Experiment
water
brine soln
LI
effluent
AI
10
Outline - Module 1

• definitions and examples
• components of a control system
• why is control required?
• the systems approach
• defn of feedback control
• control engineering
• control objectives
• improving operation - the operating window and
  control benefits analysis

11
Components of the shower example

• desired values - setpoints - target values for
  flow, temperature
• sensors - measure process conditions - e.g., hand
• control algorithm - determine corrective action
  required - intuition ?
• final control element - handles adjusted to
  maintain values - taps

12
Example - Shower
setpoint
control algo.
I want a hot shower
hotter shower turn HW tap to right
sensors
FI
final control elements
TI
13
Outline - Module 1

• definitions and examples
• components of a control system
• why is control required?
• the systems approach
• defn of feedback control
• control engineering
• control objectives
• improving operation - the operating window and
  control benefits analysis

14
Control is necessary in order to...

• reject disturbances protect against the flush
• adjust manipulated variables to restore and
  maintain controlled variables at setpoint
• if we know disturbance is coming, we can also act
  in advance - feedforward control
• also known as the load problem

15
Control is necessary in order to ...

• follow changes in setpoint
• move to new target values - e.g., warmer
  temperature
• frequently for economic reasons
• also known as the servo problem
• important in mechanical systems - e.g., motor
  control in printers

16
Outline - Module 1

• definitions and examples
• components of a control system
• why is control required?
• the systems approach
• defn of feedback control
• control engineering
• control objectives
• improving operation - the operating window and
  control benefits analysis

17
The Systems Approach

• look at flow of information in process
• identify generic quantities - inputs/outputs/state
  s
• generic - common to all dynamic systems

18
The Systems Approach

• inputs - variables in a process which cause
  changes in other variables
• e.g, manipulated variable, disturbance
• outputs - variables which respond to inputs and
  are measured
• states - internal states of the process -
  complete picture

19
Systems Approach

• key issue - causality - cause and effect
• outputs and states respond to changes in inputs
• manipulated variables are some of the inputs to a
  process
• controlled variables are some of the outputs in a
  process

20
Systems Approach - Tank Lab Example

• inputs - concentrated salt flow, exit flow
• outputs - level, salt concentration
• states - level, salt concentration, temperature,
  pressure
• note that input/output designation does not
  always match physical direction of flow

21
Systems Approach - Tank Lab Example
INPUTS
OUTPUTS
STATES
Salt Tank
brine flow
exit salt conc.

manipulated variable inputs
salt conc. level temperature pressure
exit flow
level

tap supply P
disturbance input
22
Summarizing Information Flow - the Block Diagram

• Summarize the flow of information in the process
  using schematics -
• Example - how is exit brine concentration related
  to the brine feed and exit flowrates?

salt concentration in effluent
Gbrine
brine feed flowrate


exit flowrate
Gexit
23
Block Diagrams

• each block represents a dynamic component
• each block has only one input and only one output
• signals can be summed together (summing
  junction), or sent to several different blocks
• the dynamic elements are typically summarized in
  the Laplace transform domain, as transfer
  functions
• there is a block diagram algebra which can be
  used to derive expressions describing process
  behaviour

24
Outline - Module 1

• definitions and examples
• components of a control system
• why is control required?
• the systems approach
• defn of feedback control
• control engineering
• control objectives
• improving operation - the operating window and
  control benefits analysis

25
Feedback Control

• use process outputs to determine changes to
  process inputs of same process
• negative feedback - act to reduce differences
  between setpoint and actual values

26
Example - Shower
setpoint
control algo.
I want a hot shower
hotter shower turn HW tap to right
sensors
FI
final control elements
TI
feedback loop
27
Outline - Module 1

• definitions and examples
• components of a control system
• why is control required?
• the systems approach
• defn of feedback control
• control engineering
• control objectives
• improving operation - the operating window and
  control benefits analysis

28
A Tale of Two Processes

• Its time to wash your hands you want to
  control
• Flow
• Temperature
• Which process design is easier to use?
• ease ability to enforce and change targets -
  setpoints

29
A Tale of Two Processes

• One of these faucets lets you set flow and
  temperature independently
• Changing temperature doesnt cause a change in
  flow
• Changing flow doesnt cause a change in
  temperature

30
Control Engineering addresses...

• Process Design
• controllability and operability
• how quickly do disturbances pass through
• responsiveness to changes
• Measurements
• selection and location of sensors
• accuracy and speed
• sufficient to characterize process status?

31
Control Engineering addresses...

• Final Control Elements
• location, type and variable
• flexibility, speed of action
• Control Structure
• pairing of controlled and manipulated variables
• favourable response and reduced interaction

32
Control Engineering addresses...

• Control Calculation
• algorithm to be used to determine changes in
  manipulated variables required to remain on target

33
Outline - Module 1

• definitions and examples
• components of a control system
• why is control required?
• the systems approach
• defn of feedback control
• control engineering
• control objectives
• improving operation - the operating window and
  control benefits analysis

34
Control Objectives

• Safety - paramount
• maintain proper operation
• emergency systems - e.g., interlocks, relief
  valves
• Environmental Protection
• via proper operation and containment

35
Control Objectives

• Equipment Protection
• proper operation and shutdown at limiting
  conditions
• e.g., loss of pump suction
• Smooth Operation
• in inputs, outputs - be a good neighbour
• minimize disturbances to integrated units, avoid
  stressing equipment

36
Control Objectives

• Product Quality
• composition, physical properties, performance
  propertiesOnce these objectives have been
  satisfied, we can focus on...

37
Profit

• reduce costs - improve efficiency
• raw materials, energy
• models - profit centre vs. cost centre
• frequently involves optimization

38
Monitoring and Diagnosis

• Immediate and short-term - assess process
  operation
• operators/supervisors/plant engineers
• identify potentially dangerous situations
• Longer-term - identify opportunities for
  improvement - Total Quality Management
• operators/supervisors/plant engineers/management

39
Outline - Module 1

• definitions and examples
• components of a control system
• why is control required?
• the systems approach
• defn of feedback control
• control engineering
• control objectives
• improving operation - the operating window and
  control benefits analysis

40
Justifying Control Applications

• Operating Windows and Controller Benefits Analysis

41
Operating Window

• region of feasible (allowable) process operation
• factors
• physical limitations
• e.g., mass fractions between 0 and 1
• safety/environmental/equipment limits
• e.g., maximum pressure, maximum temperature

42
Operating Window

• factors (continued...)
• equipment capacity
• e.g., max. pump throughput, tank size
• product quality
• e.g., minimum octane, maximum vapour pressure,
  maximum basis weight

43
Operating Window - Exercise

• What is the operating window for your shower?

44
Operating Window - Example

• reactor for producing octane components
• catalyst deactivates - higher temperature
  required
• metallurgical limits

Photo from www.grimmengineering.com
45
Profitability is maximized by operating at the
point of highest profit within the operating
window

• - constrained optimization problem (CHEE422)

46
In defining operating window, we encounter...

• Soft Constraints - limits which can occasionally
  be violated
• e.g., product quality
• e.g., small deviations over equipment guidelines

47
In defining operating window, we encounter...

• Hard Constraints - limits which cannot be
  exceeded under any circumstances
• safety and environmental limits
• e.g., maximum pressure in a vessel
• extent to which process can be operated close to
  these constraints depends on the degree of
  variability in the process

48
Process control reduces variabilityenablingopera
tion closer to operating limits

• - frequently a source of financial incentive to
  implement control

49
Assessing Benefits for Control

• given that primary objectives have been
  satisfied, control benefits accrue from reduced
  variation
• compare operation before (new) control strategy
  to operation with the (new) control strategy

50
Control benefit analysis requires...

• performance function for process
• how does profitability vary with operating point?
• summary of process variability with and without
  the (new) control strategy

frequency
profit
operating conditions
operating conditions
51
Control benefit analysis

• calculate average or expected performance
  with/without control
• Fj - frequency of occurrence of range
• process statistical distribution
• e.g., from histogram
• calculating Eprofit over process distribution

52
Reduced variability can improve profitability
by...

• allowing mean operation closer to constraints
• e.g., increased throughput, conversion

frequency
profit
constraint
constraint
old
operating conditions
new
operating conditions
53
Reduced variability can improve profitability
by...

• reduced variation about fixed target
• e.g., excess O2 and heater efficiency

frequency
profit
operating conditions
operating conditions
54
How close can we move to constraints?

• think of statistical variation - 2 or 3limits
• soft constraints - mean can be 2
  fromconstraint
• hard constraints - mean must be greater than 3
  from constraint

55
Look ahead...
Controller Design Procedure
Objectives
Modeling
Process Characterization
Controller Design
Implementation