Booster Pump Noise and Falling Flow: Is It Really Cavitation?
Fairview Booster Station lifts finished water from a 500,000-gallon clearwell to a 1.0-MG elevated tank using two 700-gpm, 40-hp horizontal split-case pumps in parallel (one standby) against about 150 ft of total dynamic head. Suction is 12-inch piping with a basket strainer; the pump centerlines sit 4 feet below the clearwell floor. Two weeks ago, a new level-control float was installed and the low-level setpoint was lowered from 20 feet to 9 feet of depth to keep the clearwell in a tighter band. Since then operators have logged: discharge pressure swinging ±10 psi, flow down from about 700 to about 610 gpm per pump, motor current fluctuating 8–10%, and a rattling, gravel-like noise at the pump 1 volute, with vibration that loosened a flange bolt. Water temperature is 78°F, atmospheric pressure corresponds to about 33.0 ft, suction friction loss including the strainer is about 2.0 ft, and the pump's NPSH required at 700 gpm is 18 ft.
1. Calculate the net positive suction head available at the 9-foot clearwell level and decide whether cavitation explains the symptoms.
Read the worked answer
NPSH available calculation.
NPSHA = Ha (atmospheric head) + Hs (static head on suction) − Hf (suction friction, including strainer) − Hvp (vapor pressure head).
- Ha = 33.0 ft at this elevation.
- Hs: the pump centerline is 4 ft below the clearwell floor, and the water is 9 ft deep above the floor, so there is 13 ft of static head above the pump centerline.
- Hf = 2.0 ft.
- Hvp at 78°F is roughly 1.0 ft.
NPSHA = 33.0 + 13.0 − 2.0 − 1.0 = 43 ft.
Interpretation. NPSH required is 18 ft at 700 gpm, so the margin is about 25 ft - the pump has roughly 2.4 times the NPSH it needs. Classic NPSH-driven cavitation at this clearwell level is very unlikely. That does not mean cavitation-like symptoms are imaginary; it means the cause is almost certainly something that presents the same way, or a mechanical problem. Do not accept "cavitation" as the diagnosis just because the noise sounds like gravel in the volute. Also note what the change in setpoint tells you: the only thing that changed two weeks ago was the operating level, which points at conditions in the clearwell rather than at the pump's NPSH margin.
2. What other causes fit these symptoms, and what tests distinguish them?
Read the worked answer
Air entrainment from a vortex. This is the leading candidate. At a 9-ft depth with two pumps pulling about 1,400 gpm total, approach velocity and submergence are marginal; a vortex can form and draw air into the suction, producing exactly this picture - gravel-like noise, swinging discharge pressure, lost capacity, fluctuating current, and vibration. A vortex can pull air long before NPSH is limiting.
Test: restore the clearwell to the 20-ft level and watch. If the noise and pressure swing stop within minutes at the higher level but return at 9 ft, it is a vortex/air problem. Also look at the clearwell surface for a dimple or whirlpool, and check the suction intake arrangement.
Mechanical causes.
- Worn wear rings or impeller damage cause recirculation, loss of flow and head, noise, and vibration - and they get worse over time.
- Misalignment, a loose coupling, or a damaged bearing produces vibration and a loosened flange bolt, but usually not the flow loss pattern seen here.
- A partially closed or failing suction valve left over from maintenance, or a fouled strainer, would show up as a rising differential across the strainer.
Distinguishing tests.
- Run one pump alone; mechanical problems stay with that pump, and an air problem tied to the low clearwell level will change with level.
- Check the strainer differential pressure and the suction valve position.
- Check pump discharge pressure against the manufacturer's curve at a measured flow; a pump running well left or right of its curve tells a different story than worn internals.
- Measure vibration at the bearings with a vibration meter and compare to the vendor's limit.
3. What do you do right now, and what changes do you make permanent?
Read the worked answer
Right now:
- Do not leave a pump running with escalating vibration and a loosened flange bolt. Secure pump 1, run the standby, and treat the vibration as a safety issue - a failing coupling or bearing can throw parts, and the pump is in an occupied space.
- Restore the clearwell low-level setpoint to 20 ft immediately; that is the single variable that changed and the cheapest test you will ever run.
- Reinspect and retorque the flange, check alignment and coupling, and inspect the strainer and suction valve position.
- If the symptoms persist at the higher level on the standby pump, schedule a teardown to inspect the impeller, wear rings, and bearings.
Permanent fixes:
- Rewrite the level-control setpoints with the pumping rate in mind, and document the minimum operating level in the station operating procedure rather than leaving it to a float setting someone can change.
- Install a vortex breaker or baffle at the suction intake if the hydraulic analysis shows submergence is marginal at realistic low levels, and confirm the intake can supply both pumps with one out of service.
- Keep a running log of flow, discharge pressure, motor amps, vibration, and clearwell level so that a two-week trend like this one is caught on day one instead of day fourteen.
- Document the event, the diagnosis, and the corrective action; if the change was made without engineering review, fix the management-of-change process too.