PID: VFD Feedback to control a FCV

[MightyMoose]

Hello everyone, I have yet another thread about PID's coming your way.
Any PLC5 users in the house?

I'm currently writing a program and have little experience with PIDs. I'm attempting to use the feedback from 2 VFDs controlling screw conveyors to adjust the position of a Flow Control Valve.

The goal is to keep the VFDs running at near full power consumption. I think my process variable should be the kW% and the CV is the position of the FCV.

I think I've been staring at the problem for so long that it feels impossible.

If anyone can give me a little guidance or if I'm simply over thinking this because I only have a little knowledge of PIDs.

 

[drbitboy]

0) What model PLC5?

1) almost certainly there should not be two PIDs with the same Control Variable (N14:24); only the second PID will control that output. I suspect there is some way to combine the power values, and you can try to control that combined value.

2) When (how frequently) are these PIDs running? Are either N7:171/0 or N7:171/1 tied to some timer? Or is this rung in a program that is run via timed interrupt?

3) We will need a lot more information e.g.

• P&ID, at least a sketch
• The Setup Screens
• A description of the process e.g. does opening or closing the valve cause the power to increase and decrease, respectively, or vice versa?
  • How does the flow control valve effect changes to the VFDs' power values?
  • What controls the VFDs?
• What should happen when the flow control valve has reached its limit and the VFDs are not at target?

I am not even sure this is an application apropo a PID instruction: you are trying to maximize a process variable, not minimize its offset (error) from a setpoint.

Try sending a PM (Private Message) to @OldChemEng (click on Contact Info on that page); I know he has beaucoup experience in this area. Plus he'll be annoyed when you tell him I sent you, so it's win-win .

-

 

[MightyMoose]

0. PLC-5/40E.

1. Yeah, I had a feeling one would override the other. I'm probably going to be averaging the values or comparing and using the smallest one.
(FCV=N14:25)(Horizontal Conveyor=N7:130)(Vertical conveyor=N7:131)

2. It runs every 5 seconds. N7:171/0 and N7:171/1 are manual bits to decide if the PID should control the SP or a manual setpoint will be entered.

3. The cement runs through a chute and through the flow control valve. The valve gradual causes an increase in power consumption the higher % the FCV is opened due to the conveyors needing more power to move the higher flow. Therefore opening valve power increases, closing the FCV decreases the power consumption.

Thanks!

 

[ganutenator]

I'm liking where dr. Nibble's brain is going w/ the 'are you sure you a PID is needed here'?

 

[drbitboy]

How are you ensuring the PID runs exactly once, every 5s exactly? More importantly, how are you ensuring it does not run in between the 5s intervals?

Does the screw conveyor run at a constant speed? And/or what controls the screw conveyor (speed, torque, etc.)?

 

[MightyMoose]

Yeah, my manager basically started showing me how we wanted to set it up before running off on a vacation. Not sure it even needs a PID but that's what was decided on.

The screw conveyor is not at a constant speed/it runs slower when there isn't as much cement being conveyed, and is controlled by torque.

What would you use instead of a PID?

He comes back Monday so I'll discuss it with him then. Not urgent, just hoping I could figure it out.

 

[drbitboy]

What would you use instead of a PID?

An Integral-only controller, which the PID could be configured to do. Basically you increase the valve in steps as long as the HP is below the target, and vice versa, probably with a deadband.

There are various strategies to do this; one is floating control cf. here, which may or may not work with this process.



If you search for floating control on the forum you will find some rather lengthy discussions.