Replace Motor with VFD and controls

Since I'm sure many have done this here and this is actually my first time controlling a pump, I hope I could get some opinions from everyone.

I currently have a 40 HP pump with a valve that is throttled back to keep the discharge pressure at 50 psig (there is a transmitter at this point along with the gauge). As everyone knows, that is not a very effective control system. I have been approved to replace the motor starter with a VFD (not sure which make yet) and I want to do better controls. There is a PLC (AB PLC5) located about 75 feet away.

Would you get a VFD with a built in PID control and bring the 4-20mA to the VFD or would you take the 4-20mA signal back to the plc and bring a separate 4-20mA signal to the speed reference of the drive. I am not currently worried about cost (YET!!!), just want some feedback on which would be better and why.

Thanks

I would use the VFD with PID that the simple way.

About the cost,Most of the new VFDs have this option with out additional cost.
From other hand the valve analog output you have, supposse to have PID so you can connct the VFD instad of the Valve.
It supposse to do so.But I am not sure it will work fine.
I would buy VFD with PID Option.If you will have problems with the PLC PID.Goto the VFD.

All the best

Bruce you are better at this then I will ever be, that said use a VFD with PID. The first reason I would use this method is response time, whether critical or not the VFD works direct with the valve and feedback; therefore no intevening actions from another device have to occur. The second reason is when not functioning properly all you have to do is check that it is enabled and the troubleshooting is isolated to the drive, valve, and feedback i.e. no intervening devices in the equation.

Bruce,

Do you plan to vary the motor speed to maintain output pressure instead of using the throttling valve? I would assume so, since there aren't many other reason to buy an expensive 40HP VFD. The last time I did a similar project I used the PLC analog method. I ordered the VFD without the PID firmware by mistake, so I had to. I wanted to use the drives (PFLEX70) control to save from buying the analog modules. I'm glad I did it the way I did. With the pressure reading in the PLC I can display info to the operators & even have MAN & Auto modes. If you are heading down this route, you will need a pump performance curve from the pump manufacturer to determine the pumps limits & efficiency curves. I'm sure volumes have been written on this subject (RPM / Flow / Head). Do a Google search to find some.

BD

I think that the most important thing to consider is whether or not the PLC 5 needs to "know" what is going on with the pump. Analog I/O with a PLC 5 can become expensive, not to mention having to set up the cards.

If the PLC does not need to know what is going on with the pump then do everything in the drive.

If the PLC IS in control of the drive and you happen to have RIO, I would suggest communicating with a pflex 70 or 700 via RIO and taking your analog pressure input into the drive(kind of a free analog module). You can set up a BTR to read that value from the drive, do your PID in the PLC, and send back a command to the drive via the RIO adapter.

Hope this makes sense.

The trouble with VFD PID controls is tuning, this can be a bit awkward. I would normally use a seperate PID controller or PLC.

Where I have used a VFD PID control in the past tuning was a matter of trial and error.

Is the existing system closed loop ? You say the valve is throttled back to maintain 50 psig - what is controlling this ?

The second reason is when not functioning properly all you have to do is check that it is enabled and the troubleshooting is isolated to the drive, valve, and feedback i.e. no intervening devices in the equation

Click to expand...

This is the kind of things I am looking for - Reasons for using one way above the other. I never thought about this Ron. Thanks

Do you plan to vary the motor speed to maintain output pressure instead of using the throttling valve?

Click to expand...

Yes. I will leave the valve in the system but at 100% open.

I think that the most important thing to consider is whether or not the PLC 5 needs to "know" what is going on with the pump. Analog I/O with a PLC 5 can become expensive, not to mention having to set up the cards.

If the PLC does not need to know what is going on with the pump then do everything in the drive.

If the PLC IS in control of the drive and you happen to have RIO, I would suggest communicating with a pflex 70 or 700 via RIO and taking your analog pressure input into the drive(kind of a free analog module). You can set up a BTR to read that value from the drive, do your PID in the PLC, and send back a command to the drive via the RIO adapter.

Hope this makes sense.

Click to expand...

I have plenty of analog in but I'm not sure of the output. These are all on remote I/O. I have to check for outputs. The transmitter just displays the pressure on the interface computer (I think it is an intellusion -spelled right???) but nothing is done with it besides the operator just recording it on a shiftly basis. I didn't think of using the PF 70 with RIO. I'n not sure how easy that will be, but I like it (I think).

Is the existing system closed loop ? You say the valve is throttled back to maintain 50 psig - what is controlling this ?

Click to expand...

No, just someone open/closing the valve to get the correct pressure.

Good things to think about. I think I like the RIO but I was originally leaning towards the PID loop in the VFD. Now should I ask the next question; which would be a better VFD for this or should I just read the thread that talks about this?

If the pressure is being controlled by someone manually actuating a valve then anything that you do will be an improvement.


I would suggest a powerflex 70. Setting up the RIO is pretty straight forward. You can configure the adapter as discrete in which case it will be either a 1/4 or 1/2 rack in your input and output data tables. However if you want to pull in an analog from the drive like I had described you will need to set it up with BTR/BTW instructions.

This is no more difficult than setting up a remote analog card. I like this method because of the reduced wiring. All you will have for control wiring on the drive is a blue hose and probably wires for an E-stop.

You will find it much easier to set up a VFD's PID loop that using a PLC. You eliminate some response delay and can get smoother reactions. You can use the PLC to provide initial run/stop and base speed settings, then let the drive control the PID.

You asked what drives have built in PID, Yaskawa P7 drives are designed for VT, variable torque, applications with built in "PI proportional intergral" control as they call it. Yaskawa also has a PID loop tuning procedure that is fairly easy to follow and setup.

I normally use the PLCs inbuilt PID control rather than the drives.
This is mainly so that I can tune the loop using our SCADA system. Generally I find it easier to do it this way.
The exception is when I use positioning drives, in this case the PID control is done within the drive, requiring a direct connection to the drive to tune the loop.

Now days, we generally use Profibus to control and monitor the drives, with the potential to set them up for parameter updating. This can allow us to keep the PID control within the drive. This is something I plan to experiment with when I setup our new test bench.