Pump Control Method

[ricketycricket]

Hey,
I am a newbie when it comes to process control and PLCs as I am just out of college and at my first job. I am trying to get a system set up that will pump air through two separate, independent systems using only one blower. Each leg of the system will require a different flow rate profile than the other (e.g. 20 min at one flow rate, ramp to another flow rate for 30 min). The profiles will need to be changed often and so I think each leg will need its own programmable controller that will have an input from an LFE and control the process via a butterfly valve. What I am unsure of is how to control the blower with these two independent control loops. My plan is to have both controllers output the valve signal to something that is capable of comparing the two signals and transmitting the larger of the two signals to a VFD to control the blower. So the leg of the blower system that requires the most flow will be the signal getting to the VFD. My question is can a PLC compare the two signals and output the higher of the two? Is there a simpler method? How ignorant am I when it comes to process control? Any advice is appreciated.

 

[leitmotif]

If mine I would
Have only one PLC.
It would sense "A" dP if valve A is open and "B" dP if B is open. I assume dP is proportional to flow.

I think if BOTH A and B are NEVER to be open I would have limit switches and an error logic to send alarm for either both valves open or dP (flow) in both A and B.

Depending on what flows you have ie are A and B matched (more than likely NO) then you could control the VFD based on dP ie flow. You may be able to have PLC control VFD (motor speed thus flow) depending on blower type - check your performance curves. Do not forget the affinity laws where power is the cube of speed change.

Dan Bentler

IF you have a lot of powder in the lines I would keep that in mind - tends to plug sensors.

 

[Calistodwt]

You say that you want to pump through two seperate systems yet you diagram shows sucking (extracting) out of two seperate systems which combine into the suction of your one blower? You also do not say whether your blower is a positive displacement or centrifugal type? Are both systems going to run at the same time, one at a time or any combination? What type of flow measuring device are you using?

 

[ricketycricket]

The pump is a centrifugal type. The flow will be measured using a Laminar Flow element and controlled using a butterfly valve. The two legs of the system will be running at the same time but different flow rates.

 

[KalleOlsen]

And the setpoint for the two legs controllers will be the flow, and for the VFD the flow sum?

Kalle

 

[ricketycricket]

My thought was that the leg that demanded the highest flow would have its valve set point transferred up to the VFD and the VFD would use a PID loop to maintain that valve at 50% open or something around there. The reason is that the leg with the lower flow rate can always close the valve further to reach a certain setpoint. The leg of the system that needs a higher flow rate is more important because it needs a greater pull from the blower to reach higher flow rates. So by having the higher of the two setpoints transferred up to the VFD both systems can operate independently.

 

[Tom Jenkins]

You need to sort out your mechanical design before you start to worry about controls. You should never throttle the suction side of a pump. If you control flow with a valve you don't need the VFD. you are using the terms pump and blower interchangeably.

Until you understand the mechanical system you aren't likely to do a good job on controls.

 

[KalleOlsen]

The flows from both legs are bound to pass through the pump.

If you could use one pump per leg, you could skip the valves and run the VFD energy optimized. Probably you would use considerably less time to prog and commission.

Is this a powder transport, a vacuum cleaner or a filter system? Can the pump work with closed inlet / vacuum ?

...
Kalle

 

[Calistodwt]

I agree with Tom that you need to sort you process / mechanical design requirements out first. Also as your process fluid is air then you must be using a blower NOT a pump. You say your blower is a centrifugal one and therefore when you get to design your control loops you probably will need to consider anti-surge control unless your blower cannot surge or else it is too small to worry about. It's OK to throttle the suction of a blower but you may have to have some form of suction pressure control in an arrangement with your flow loops control.

 

[ricketycricket]

Thanks for all of the responses! Sorry for not being clear with my terms and description. I will not be using a pump, I will be using a blower as a part of a filtration system. I do not care about the outlet air from the blower. The only thing that is important to me is the flow rate going into each leg of the system. Each leg will be used for separate tests. One blower is being used because of space constraints. I was using the VFD in addition to the valves because when both legs are running the VFD can optimize energy use.
I am nearly certain that the basic set-up I have will work with two manually adjusted butterfly valves and a VFD. It will just significantly increase test time and require adjustments to reach the right set points. My limited knowledge of PLCs is mostly why I am here. Is a PLC capable of receiving two signals and transmitting the larger of the two on?

 

[Mickey]

I would set the VFD to your desired total flow and set the side you want to flow the most wide open. Then set the low side ( pinch back) to its desired flow rate, the high side will take the rest ,both sides adding up to the total flow. One PLC is all you need.

 

[Ron Beaufort]

Is a PLC capable of receiving two signals and transmitting the larger of the two on?

yes ...

and I assure you that as you learn more about these things, you'll be amazed at how much they're capable of doing ...

 

[KalleOlsen]

I agree with Mickey.

An option will be to add a dP sensor and run the blower dP- controlled. Evtl. feed forward with the sum of the valve positions or flow setpoints.

Kalle