Why do parallel fans/blowers show an oscillatory behavior?

Question

I am analyzing an industrial ventilation system that at some point operates two industrial size blowers in parallel. They often show oscillating behaviour: one blower briefly draws more power / generates more flow, and the other performs less. Then the roles reverse. The only practical, known way to reduce the oscillation is to throttle down the vane valve in front of the fan and then open the valve again. The blowers are almost (but not completely) identical.

I seems that both blowers affect each other in some unstable way. Is this a known phenomena? I'm not finding anything when searching for fan/blower/pump oscillation/resonance/hysteresis.

Answer 1

The reason is due to slight instability of the fans, and bad fan stability selection. To begin with, all backwards inclined fans have, to some extent, an unstable range. All fans are designed to operate at an RPM using an RPM chart. This includes the stable and unstable ranges of the fan:

So long as the fan is in the "downward slope", this is considered acceptable. This has been discussed as fan surge with other answers on this site. One of the big notices with fan surge is that it stops at the local maximum of the fan, and goes no farther. Many times, a single fan is considered to operate at this point and this is fine for operation. However, operating multiple fans in parallel can cause problems if they are in the normally stable region, but above the maximum pressure at no flow point:

This is because the double fans can oscillate as you describe, with one fan operating at the left point of the curve, and one fan operating at the right point of the curve, then rapidly switching back and forth. Only by ensuring that the static pressure across the fans is less than the no-flow pressure, can you ensure completely stable operation.

For more research, see below some additional sources. Math Source Graphical Source