Process change: New tuning required?

[controlnovice]

We just increased the size (HP, thus flow and pressure) of our cooling water pumps.

We have several users, each controlling the amount of cooling water to themselves with a control valve (all temperature control) - approximately 10 users.

Because of the increased pressure/flow, do we have to re-tune each of the user's cooling controllers due to the change in process conditions? Or should the existing tuning suffice?

If we need to change, can you explain why? Shouldn't the existing controller determine that it needs more or less cooling and thus change the valve output "as-is"?

 

[ScottyUK]

I suggest that the tuning be looked at. You have effectively increased the gain of the system because a given change in valve demand will now produce a greater change in flow through the valve than it did previously at the lower pressure. Your system will become more 'lively' with the higher delivery pressure. At the extreme it could possibly become unstable if it was already marginally tuned. You may also see higher order effects due to the higher velocity of fluid in through system which will affect the lag inherent in the system, thus requiring a change in PID loop parameters.


----------------------------------

One day my ship will come in.
But with my luck, I'll be at the airport!

 

[ilan7]

Due to the pressure change at the inlet to each valve, the valves has to change the position in order to keep the same flow. If you have one or two valves in parallel, no re-tunning will be needed. As you said you have 10 valves in your system, I asume that whenever one of them open / close it effect the total pressure, and you might get into a situation that the valves will "fight" with one another. In this case, you will have to carfully retune the system. Otherwise - just let it run.

 

[rmw]

This assumes that there is a significant change in the head produced by the CW pump. If you are only talking a matter of a few feet of head, and your valves are sized on a high delta P, while all that has been said is true, and I agree with it, I don't think it would really matter that much.

However, if you increased the head signicifantly, or your valves are sized on very low delta P's so that your new head is a significant percentage differance of your sizing delta P, you should re-tune.

rmw

 

[FrancisL]

The existing tuning might well suffice.
The existing tuning is undoubtedly not perfect, and there are non linearities all over the place (control valves, pump response etc) so the loops are probably on compromise settings anyway.

If the loop is still stable and is fast enough then why change it?