Tuning temperature control loops
Temperature loops are slow, lag-dominated, and patient. They are also the one loop type where derivative action is regularly worth having — the measurement is usually quiet enough for it not to cause trouble, and the lag is large enough for it to help.
The main hazard is impatience. A temperature loop can take a long time to settle, and changes made before the previous one has finished working produce a confusing picture and a worse loop.
Lag versus dead time
Both make a loop slow, and they need opposite treatment. Lag is a gradual response — the measurement starts moving straight away and takes time to get there. Dead time is a period where nothing happens at all.
Derivative helps against lag, because it acts on the fact that the measurement is already moving. It does nothing useful against dead time, because during dead time there is nothing to act on. A temperature loop whose slowness is mostly thermowell lag benefits from derivative; one whose slowness is transport delay through a long line does not.
Check the thermowell before adding derivative. A heavy well, a poor fit, or an air gap between sensor and well can add far more lag than the process itself, and it is a mechanical fix rather than a tuning one.
Cascade is usually the right answer
Most well-designed temperature control is cascaded onto a faster inner loop — jacket temperature, steam flow, coolant flow. The inner loop absorbs supply disturbances long before they reach the temperature.
If a temperature loop is fighting supply pressure variation directly, adding a cascade will do far more than any amount of retuning.
Questions that come up
How long should I wait before judging a change?
Long enough for the loop to have finished responding to the previous one, which on a temperature loop can be many minutes. Judging a change early is the most common way to conclude that a good change was a bad one.
Related
Free tools
Glossary
Last reviewed 2026-08-01.