PID Tuning for Pressure loop

  • Thread starter Ashutosh Chandel
  • Start date

Thread Starter

Ashutosh Chandel

I am facing a problem in tuning a PID controller in a pressure loop, I would like to know the settings for P I and D which will give me the best results, also I need help on anti reset windup for pressure loop. As the controller output can only build pressure but cannot release it.
Looks like you are faced with an integrating process. On integrating process, the gain needs to be high and the reset slow (derivative is typically a bad idea except if you are using a first-principle method such as IMC - Lambda tuning).

Altough you might want to use a IMC-LAmbda tuning methid, you can try to tune by feel to get a simlar response:

For an integrating process:
Kc= high
Tr= 2*speed of reponse.

A good start would be a reset of 20-30 seconds/rep and a gain of 1.

note that with an integrating process you may have an oscillatory system if your gain IS TOO LOW. So don't be shy to increase the controller gain. (typical fast integrator suach as a turbine governor will have a gain of 2-5 and a reset of 10-30 seconds)

Hope this helps,

Ben Janvier,
Senior process control consultant
[email protected]


It is difficult to give P,I,D setting without knowing the exact process dynamics. In specific cases like pressure header vent control systems,
it is quite common to use anti reset windup controllers. In these cases ,keep the Gain
always higher and Preferably use only PI mode, instead PID.
For starting settings on common control loops, see:

However, for the best solution, I would suggest using a good PID Tuning, modeling and simulation software package and:
1) model the loop. Use plant data from the CO moving in the up direction only. (i.e. use data for modeling with a small increase in the CO)
2) Then tune the controller using the same software.
3) Then simulate the response using the same software. Adjust the tuning to give NO overshoot on the CONTROLLER OUTPUT.


J. Gerry, ExperTune