surge in multistage centrif blower?

[NoHoCa]

I searched the forums but couldn't find any topics. I'm looking for information on surge as it occurs in multistage centrifugal blowers (room temperature air service). It's been suggested that because the blower curve has a saddle in its upper end (as opposed to continuously rising to shutoff), the unit may surge back and forth across the peak in the operating curve. Apparently this phenomonon can lead to physical damage to the rotating equipment.

My question is for a system that might be controlled by AFD (adjustable frequency drive), such that the unit moves up and down its performance curve in response to changing process demands.

Any real world experience? Printed references? Cautions? Advice? Thanks...

 

[MintJulep]

You've summed it up nicely yourself.

 

[mayt4u]

Hi NoHoCa, I believe your running into a problem of "stone walling", this can be very bad for the compressor, as it starts to set up an oscillation and things start to fall/fly apart. I'll get back to on some things to try, but I believe one of the cures is to restrict the intake a little. Not quite sure. How many stages do you have and what are the outputs of each. I can work with CFM and psi. Go figure. Never heard of droop (saddle) in the output curve and wonder why anyone would design in such a problem. Did this happen before or did it just begin? What changes have you put into the system recently? Does it only happen when? Is there a drop in output when the oscillations start psi? Is there an increase of rpm, in the prime mover, once this starts?

I'll take this problem to my old hippie, bike building, banjo picking, cigar smoking, beer drinking buddy and talk to our engineering instructor and see what comes out.

Cheer up, it could be worse, it could be me trying to figure this out!!!

 

[NoHoCa]

Actually we're considering converting a fixed speed blower application to a variable speed app. And so the potential for surge has not yet been realized.

I'm wondering how close to the surge point (in psi) we can get, and curious if anyone's tuned a PLC control loop for such an application.

Feedback from fellow bike builders and beer drinkers, or others as well, greatly appreciated...

 

[MickyMan]

Generally, the operating point should be at least 4% below the fan's peak pressure.