Practical Advanced Process Control for Engineers and Technicians

THE WORKSHOP:

In today's environment, the processing, refining and petrochemical business is becoming more and more competitive and every plant manager is looking for the best quality products at minimum operating and investment costs. The traditional PID loop is used frequently for much of the process control requirements of a typical plant. However there are many drawbacks in using these, including excessive dead time which can make the PID loop very difficult (or indeed impossible) to apply.

Advanced Process Control (APC) is thus essential today in the modern plant. Small differences in process parameters can have large effects on profitability; get it right and profits continue to grow; get it wrong and there are major losses. Many applications of APC have pay back times well below one year. APC does require a detailed knowledge of the plant to design a working system and continual follow up along the life of the plant to ensure it is working optimally. Considerable attention also needs to be given to the interface to the operators to ensure that they can apply these new technologies effectively as well.

WHAT IS INCLUDED?

  • Receive a certificate of attendance in support of your continuing professional commitment
  • All workshops include the associated hardcopy technical manual
  • Printed workshop handouts
  • Lunch and refreshments
  • Interact and network with workshop attendees and experienced instructors
  • Practical, industry driven content to assist you in your continuing professional development (CPD)
  • Attendees automatically become IDC subscribers and receive exclusive deals and technical content every month

WHO SHOULD ATTEND?

  • Automation engineers
  • Chemical engineers
  • Chemical plant technologists
  • Electrical engineers
  • Instrumentation and control engineers
  • Process control engineers
  • Process engineers
  • Senior technicians
  • System integrators

CONTENT SUMMARY

JUSTIFICATION OF ADVANCED CONTROL

  • Advanced versus classical control
  • Advanced on-line control versus statistical process control
  • Comparison of pay back time on various examples of applications in real plants
    Practical exercise 1: Model representation

FUNDAMENTALS OF PROCESS CONTROL

  • Processes, controllers and tuning
  • PID controllers – P, I and D modes off operation
  • Load disturbances and offset
  • Speed, stability and robustness
  • Gain, dead time and time constants
  • Process noise
  • Feedback controllers
    Practical exercise 2: PID loop tuning parameters refresher

FUNDAMENTALS OF TUNING PID LOOPS

  • Open and closed loop tuning
  • Ziegler Nichols
  • Fine tuning for different process types
  • Lambda tuning
  • Ten different rules compared
  • Cascade systems
  • Feedforward control
  • Deadtime
  • Models and disturbances
    Practical exercise 3: Loop tuning refresher (both open-loop and closed-loop)

INTERNAL MODEL CONTROL (IMC)

  • Open loop model of the process in parallel with the process
  • Control system in two blocks
  • Equivalence with a classical controller
  • Disturbances rejection and control
  • IMC and delays
  • IMC and feed forward (measured disturbances rejection)
    Practical exercise 4: IMC controller

MODEL PREDICTIVE CONTROL (MPC)

  • Single input / output versus multivariable control
  • Example on a binary column causality graph
  • Constraints and planning ahead before acting
  • Different notions of models
  • Action model - measured disturbances model
  • Unmeasured disturbance models
  • Reference trajectories
  • Example of a quality blender control system
    Practical exercise 5: MPC controller representation

MPC: MODEL REPRESENTATIONS

  • State space representation
  • Transfer function representation
  • Impulse response representation
  • Various mathematical formulations
    Practical exercise 6: MPC controller interaction calculation

MPC: MODEL IDENTIFICATION

  • Identification require a good knowledge of the unit
  • Black box models / grey box models
  • Causality graph of the unit
  • What to identify?
  • How? Step responses - pseudo random binary signals
  • Exercises of identification on various types of petrochemical units
    Practical exercise 7: MPC controller calculation programming and setup

MPC: OBSERVERS

  • Overall formulation
  • Purpose of observers in control algorithm based on state / space representation
  • Innovation on measured output - estimation of the state
  • Study of Kalman algorithm
    Practical exercise 8: Gain scheduling

MPC: CONTROL

  • Overall formulation
  • Hard constraints on manipulated variables
  • Set values and soft constraints on control variables
  • The notion of horizon
    Practical exercise 9: Feed forward

REFERENCE MODELS

  • Handling setpoints on controlled variables
  • Measured disturbances rejection
  • Unmeasured disturbances rejection
  • Handling soft constraints on controlled variables
  • Rejection of disturbances
    Practical exercise 10: Ratio control

CONTROL FORMULATION PROBLEM

  • Quadratic criterion versus geometric control
  • Importance of the horizon length
  • Use of the weight matrix
  • Handling output constraints along the horizon
  • Projection of measured and unmeasured disturbances along the horizon
  • Final quadratic problem formulation and resolution
  • Off-line pre-processing
  • On-line calculations
    Practical exercise 11: Decoupling circuits (both feed forward as well as inverting)

MPC STEADY STATE OPTIMISATION

  • Degrees of freedom and rationale for optimisation
  • Economic output submitted to setpoint
  • Slogans to maximise or minimise
  • Bridge from optimisation to control
  • Reachable targets for economic variables
  • Interpretation of the horizon for economic variables
  • Change of the control formulation problem
    Practical exercise 12: Dead time compensation (using formulae as well as a Smith Predictor)

APPLICATION OF THE THEORY TO THE CONTROL OF TWO DIFFERENT UNITS ON A PROCESS SIMULATOR

  • Complete application (identification, controller design, control and optimisation)
    Practical exercise 13: Cascade control, using PV tracking and initialisation

Book Now