Advanced Temperature Measurement and Control

Unit 4 - Temperature Look Analysis

UNIT 4

Temperature Loop Analysis

This unit describes the basic tools needed to analyze and improve the performance of all temperature control loops.

Learning Objectives When you have completed this unit you should:

Understand three key variables that determine the performance of all temperature loops.
Be able to estimate controller tuning settings and the associated control errors from the three key variables.
Know the structure and performance advantages of feedforward and cascade control
Recognize when model predictive control is needed.

4-1. Key Variables

The tuning settings and performance of all temperature loops can be estimated from the knowledge of three key variables called dead time, time constant, and steady-state gain.

Of the three variables, dead time is the most important. Infinite controller gain and, consequently, perfect temperature control are possible if the dead time is zero. If the controller can immediately see and correct for load upsets and set point changes, deviations from set point become zero. Changes in the time constant and the steady-state gain (nonlinearities) would have no effect. While zero dead time is not possible, as the dead time approaches zero, tuning and performance problems become negligible. In fact, temperature control loops on large backmixed liquid volumes have a dead time that is relatively small compared to the loop time constant. Consequently, the temperature controller gains can be high enough if properly tuned to make the control error less than the measurement error (Ref. 1).

There are many sources of dead time, as summarized in Table 4-1. There is a pure and equivalent dead time characterized by sharp and smooth transitions, respectively Pure dead time originates from transportation delays through pipes, tubes, and jackets; plug flow in volumes due to imperfect mixing; control valve dead band and prestroke delay; algorithm computation, times; and digital scan times. Equivalent dead time develops from time constants in series. The largest ones are generally in the process and are interactive thermal time constants. Since they are interactive, the dead time formed does not exceed 20% of the net effective largest time constant. The sources of equivalent dead time are heat transfer surfaces, reaction rates, volumes in series, mixing, measurement lags, control valve or variable speed drive slewing rate, and signal filters (Ref. 1).

< Previous Excerpt Next Excerpt >

Purchase This Book

TABLE OF CONTENTS

Unit 4 - Temperature Look Analysis

Learning Objectives When you have completed this unit you should:

Understand three key variables that determine the performance of all temperature loops.

Contact Preferences

This is embarrasing...

Customize Your GlobalSpec Experience

Select Your Free Newsletters

Industry Newsletters

Select Your Free Product Alerts

This is embarrasing...