RSLogix PIDE control question?

Hello all. I guess this question is more about a control scheme than completely about a PIDE. I have a blower system that has a VFD and a modulating recycle valve. Each on separate PIDE's.Blower tries to meet a pressure setpoint and also has a minimum speed it can run. Recycle valve recycles outlet overpressure back to inlet. Recycle is usually set 1 psi higher than blower SP so they don't fight each other. Customer has situations where his flow may stop abruptly. This usually trips our system and pops relief valves because we don't respond fast enough. We don't know when these conditions may happen and if we tune PID loop for high gain on the recycle, the system could become unstable. I'm looking for thoughts on how to get away from this? I even thought of using some kind of Low Select on the blower and have a very fast acting PID in in an upset condition. Any thoughts? Thanks in advance!

My two cents. Try responding fast on only one loop. Probably the recycle valve because you do not want to abuse the VFD.

Thanks- that was going to be my first attempt to rectify problem. I'm just worried about the pressure dropping so low that it goes under the VFD setpoint and makes that take off and then the recycle and VFD get into a battle to raise and lower pressure!

Well- this didn't work out too well. The fast acting recycle valve swung the outlet pressure so low and high that it affected the VFD pressure PID and begun to swing wildly. Back to the drawing board...

is there a signal that you could grab when his system stops abruptly? you could then use that signal to open the recirc valve 100% and stop the vfd, bypass the pids until the signal is gone.

I've just asked for a signal as you described. Here's the scenario- gas compressor running to a setpoint pressure feeding 4 Cat engines. One of the engines trips abruptly-recycle valve setpoint is 1 psi higher than compressor setpoint and has higher gain. The system pressure spikes causing the recycle to react hard which it kind of needs to so the relief valves don't pop. As pressure swings the VFD tries to jump in the act and compensate too... yada,yada...

eh, my idea wont work because of the other three engines that are still workin.....have you tried lengthening the pide update time for the vfd? this way the recycle valve takes care of the spike before the vfd pid notices?

The worst part of this whole adventure is that the system is out of state. I have thought about slowing the update rate and also possibly freezing the VFD PID until the disturbance is over.

that sucks, im currently going through the same thing with a customer in philly and the customer has no internet access for our vpn.....so 7 hour drive everytime to make changes.

Yup. That's my life! Going out there next week- I'll post a follow up.

Does the recycle valve have enough capacity to maintain the pressure at 1PSI above setpoint indefinitely?

You likely can't have the VFD trying to control pressure if the recycle valve is active. Assuming you have bumpless transfer active in the PIDE your best bet is probably to switch the VFD PIDE into manual as soon as the recycle valve opens and ramp the VFD manual command down slowly outside of the PID until the recycle valve closes. At that point put the PIDE back into auto and let it take over control of the VFD speed again.

Assuming you are using the incremental version of the PIDE you might alternatively be able to keep the PIDE active, drop your proportional gain very low when the recycle valve is open and see if the integral alone will bring down the VFD command at a reasonable rate.

This all assumes the recycle valve can keep things under control long enough for the VFD to react.

Keith

I kind of like Keith's suggestion but why not run the VFD in open loop all the time? I would run them in proportion to the number of gas compressors on-line. I would only use the recycle valve for trimming the flow from the VFD.

I see two huge problems.
1. It isn't wise to have 2 PIDs that fight each other.
2. I doubt your control will ever be fast enough unless you know what the compressors are doing so the control can react instantly instead of by pressure changes.

Peter- I haven't been a fan of the 2 PID's either. Especially because they never think about giving us feedback from the engines. I think a mechanical pressure regulator set 1psi above VFD setpoint would be better.

Originally posted by dginbuffalo:

I think a mechanical pressure regulator set 1psi above VFD setpoint would be better.

Click to expand...

It would have to be pilot operated regulator in order to track changes in setpoint. This isn't insurmountable but it is a consideration.

Keith

Keith- usually we have a certain setpoint and don't deviate from that. But if we did need to change it, that could be an issue.