Refrigeration Level Control Valve Tuning

[StraighOuttaND]

Hello All,

I am in the process of doing some PID tuning on the level controllers for our refrigeration system.

I am trying to tune the Level controllers for the Economizer and Shell/Tube Chillers. The level controllers are all upstream on the vessels. The Economizer feeds 3 Chillers that maintinan a propane level to cover the tubes. The Chillers than have a back pressure controller that discharge gas back to the compressors.


Surge Vessel > Level Control Valve > Economizer >
Level Control Valve > Chiller
Level Control Valve > Chiller
Level Control Valve > Chiller

With the current tuning parameters, the system goes into a perpetual swing when there is any kind of sudden flow change in the plant. (ie surge of gas due to a PIG or dehy beds switching. The surge of Gas flashes propane at the Chillers and the level controller swings open. This will than cause a constant swing of all the level controllers until their is operator intervention.

In this kind of system does the Economizer need to have the fastest tuning to accommodate the actions of the chillers?

Control Action is Reverse
P= Gain Multiplier
I= Integral Time in Minutes
D=Derivative Time in Minutes

 

[georgeverghese]

Sounds like the LCV for each chiller is pulling in too much propane when there is a gas surge in order to keep up with the required high rate of vaporisation. And hence the control loop interaction between these level controllers connected in parallel as they compete for the limited feed ex the economiser.

Try modifying each level loop to a cascade LIC - FIC loop with a max flow limit SP on the FIC. Also add a ramp up rate limit on the FIC setpoint reset signal from the LIC to smoothen out the demand rate rise.

But this change will mean that during a gas surge / overload, you may need to dump the excess gas that the chillers cannot handle ( as these will be locked in at max feedrate of propane ) which would otherwise cause a temp spike in the combined exit gas from these chillers.

 

[georgeverghese]

Is there a master temp controller ( sensing exit gas temp from the chillers) that resets the LC setpoint on each chiller ? Temp controllers are prone to cause fluctuations in the controlled variable, and this is due to the long sensing lag inherent in the RTD and thermowell.