Testing Fan for Unstable Operation

[OhDeeYou]

I need to develop a test procedure to prove that an AHU fan (FC centrifugal, dual-inlet, direct-drive) selected in the unstable region of the OEM fan curve won't operate erratically (stalling/hunting/surging). Since the fan is housed in a very tight cabinet with poor airflow conditions (potential systems effects at the inlet and outlet due to obstructions, tight turns, no discharge duct, etc) and the fan curve is based on optimal conditions, I don't really "know" that the fan is actually in the unstable region--that is, the OEM fan curve doesn't reflect the fan's performance in this application.
I want the test procedure to try to induce unstable operation by simulating any reasonable activity or operating condition that could take place in the field which could reveal instability. I'm thinking about things like gradual and abrupt changes in pressure drop (via a damper or obstruction) and sudden power loss/restoration.
Any thoughts or ideas? Selecting the "right" fan is no longer an option. (BTW, we are the end-user, not the designer of this AHU.) Thanks!

 

[MintJulep]

Well, as you no-doubt already know, you can't prove a negative.

As a starting point, how about developing the "as-installed" "fan unit" curve?

 

[OhDeeYou]

That's an important first step we plan to do. Assuming it shows that operation is in the unstable region, what kinds of things can/should we do to see if fan performance will become erratic? Or is just operating from shut-off to rated flow and back down to shut-off all that needs to be done?

 

[willard3]

If you have the fan curve, you only have to operate in the cfm/pressure range that is unstable for limits for the test, not necessarily wide open and not fully closed.