Pressure control with VSD

[DAVIDFURUT]

Variable Speed Fan Control

The other variable here is the fan operating performance. Most spray booth fans {tubeaxial} typically produce and run in the 1 to 2" static pressure range. It should be checked that the fans would not run in an unstable region of the performance operating curves {cfm vs pressure, fan curves from manufacturer}. This generally results in pulsations in flow and chaotic readings on your sensors.

 

[Tom Jenkins]

If it was up to me, I wouldn't screw around with PID control. I would be very afraid that the system would go nuts if a maintenance door were opened or some other process upset occurred. After all, I suspect they just want to make sure air movement is from the outside in to eliminate fume escape.

I'd use a simple setpoint with dead band control. If the suction drops below setpoint increment the discharge fan speed, delay a few seconds, and repeat until the suction is above setpoint. If the suction is above a second level, setpoint + deadband, decrease the discharge fan speed, delay, and repeat until the suction is below deadband. In beteen do nothing. Easy to program, stable, and sufficiently accurate for the purpose. The response on increasing suction would look like:

decrease fan speed

------ High Suction (Setpoint + deadband)

fan speed unchanged

------ Low Suction (Setpoint)

increase fan speed

As Dick points out, you need a compound pressure transmitter - this measures from slight vacuum to slight positive pressure. I don't like Setra, but Foxboro and Invensys Online and Rosemont and lots of other guys have them.

 

[TamHo]

hI would like to thank you all for giving me your expert advice.

I have been out all day taking care of electrical field inspections so I apologize to everyone for not being able to reply sooner.

ArikBY
"The result came from your reference. it much esay to let the VFD to do the PID loop."

---I wish it was this easy Arik, unfortunately there are differentmodes of operation that require digital reference speeds and different modes that require PID control. My mistake for not mentioning this at the beginning of the thread, sorry about that.

Kevinrcase
"I hope that the input can be filtered either in the drive or transmitter since the static pressure would not be as tame as lets say level in a tank control loops."

---You are absolutely correct on this Kevin, in my particular setup (mechanically), there are fluctuations in pressure and this sometimes causes problems ie. the VSD continuously hunts for the optimal speed to run at. Filtering can be used in the PLC and also sampling time can be played with to address this problem.

DickDV
---That is correct Dick, the pressure is to be ambient at atmospheric or a bit +/- depending on the user.

"...you will need a sensor/transmitter that sends 4ma somewhere at negative pressure and 20ma somewhere at positive pressure."
---The transmitter that is installed is a differential pressure transmitter, the low side is connected to ambient, and the high side is connected to the cabin. I have read the values obtained from the P.T when the unit is off, we get a value, if the unit is running negative, we get a value less than ambient (when it is off), and the opposite when the unit is positive. Wouldn't this just be a matter of scaling these values? Once they have been digitized in the PLC, we can interpret the values any way we like. There is an HMI operator interface being used to change temperature and mode of operation.


Kevinrcase
---That is indeed correct, the scale will be in W.C.

DickDV
"As I think further about this, the drives I'm most familiar with do have the software option of setting up the input for 0ma to 20ma."

---That is a wonderful idea, never thought about doing this before. The PLC analog card has dip switches that I can change from either 4-20 or 0-20mA ranges. Great idea!

DavidFurut
---The fan is a reversed incline fanwheel and there are some pulsations in the pressure. The pulsations can be filtered out in the PLC.

Tom Jenkins
That is a very good point about the maintenance doors being opened. Fortunately, I have installed some limit switches on the doors that can react to the doors being opened, ie. hold the running speed until the doors are closed. The setpoint with dead band control sounds like the solution to the problem. I will check out Foxboro, Invensys and Rosemont to see what they have.

Thank you all for your great responses and your professional expertise, I think I have an idea of how to tackle this problem now and sounds like the setpoint with deadband control will be the solution.

TamHo

 

[Jim Dungar]

A quick glance at a Dwyer Instruments catalog shows their series 603A transmitter is available 4-20mA output based on -10 to +10"WC.