Pump Cavitation After a Wet-Well Drawdown
A vertical turbine pump in a 20-foot-deep wet well has begun making a rattling, gravelly noise and vibrating. Discharge pressure fluctuates about 8 psi. The pump ran quietly for months at a wet-well level of 12 feet. Last week a downstream storage tank was taken offline for cleaning, so the wet well now cycles between 14 feet and 4 feet. The noise appears only when the level drops below about 7 feet. The operator suspects a worn impeller and wants to pull the pump, which means a crane and a two-day outage.
1. What is the most likely cause, and how would you confirm it before pulling the pump?
Read the worked answer
The symptom pattern points to cavitation from insufficient submergence, not a worn impeller:
- Noise and vibration begin only below a specific wet-well level (7 ft).
- Discharge pressure fluctuates - classic for vapor pockets collapsing in the impeller.
- The pump was fine until the operating level range changed.
Confirm without pulling:
- Watch the level and the noise together. If noise starts reliably at ~7 ft and stops when the level rises, that is submergence, not wear.
- Check suction conditions. Is a vortex forming? Is the pump drawing air? A vortex at low level pulls air and mimics cavitation.
- Compare to the pump curve. Required submergence to avoid cavitation is a published value (NPSH required plus a safety margin). If 7 ft is below it, the case is made.
- Listen and feel. Cavitation sounds like gravel; a worn impeller usually shows as a steady loss of head at all levels, not a level-dependent event.
Pulling the pump first would cost a crane and two days and likely find a perfectly good impeller - the real fix is operating level.
2. What are your options to fix it, and which do you choose?
Read the worked answer
Options:
- Raise the low-level setpoint so the pump never runs below the submergence requirement (e.g., stop at 8 ft). Cheapest, immediate, but reduces usable wet-well storage.
- Restore the downstream tank to service so the wet well does not have to cycle so deeply. This is the root cause - the tank outage changed the duty.
- Throttle or trim the pump so it demands less NPSH. Slower, but reduces capacity.
- Modify the wet well (baffles, anti-vortex devices, deeper well) - capital work, not a today fix.
Choose 1 now, 2 as the real fix. Reset the low-level cutout to protect the pump immediately, accept the reduced cycle volume, and get the storage tank back online so the wet well can return to its original operating band. Then verify the pump's required submergence against the manufacturer's data and set the cutout with margin.
The judgment point: the pump did not fail - the operating envelope changed. Fix the envelope, not the equipment.