Pump Cavitation After a Strainer Change
A 75 kW end-suction centrifugal pump at a hospital chilled-water plant is rated 200 m³/h at 32 m head, 1,450 rpm, with NPSHr 4.5 m. It draws water from an open 35°C break tank with the pump centreline 2.5 m above the tank water level. The suction pipe is 6 m of 200 mm internal diameter steel with two 90° bends, a fully open gate valve, and a strainer; design suction loss was 1.8 m, including 0.4 m for the strainer. After the strainer element was replaced with a finer mesh, the pump cavitates loudly at 180 m³/h; the suction gauge reads −35 kPa; the duty point has moved. Assume atmospheric pressure 101.3 kPa, water density 994 kg/m³, vapour pressure at 35°C 5.6 kPa, and g = 9.81 m/s².
1. Calculate NPSHa with the new strainer loss and decide whether cavitation is explained.
Read the worked answer
NPSHa = (p_atm − p_v)/(ρg) + z − h_L.
(p_atm − p_v) = 101.3 − 5.6 = 95.7 kPa. 95.7 × 10³ / (994 × 9.81) = 9.81 m. Static term z = −2.5 m because the pump is above tank level. New suction loss: design 1.8 m included 0.4 m strainer; replacing 0.4 m with 2.6 m adds 2.2 m, so h_L = 4.0 m. NPSHa = 9.81 − 2.5 − 4.0 = 3.31 m. NPSHr = 4.5 m, so NPSH margin = 3.31 − 4.5 = −1.19 m.
The pump is operating with insufficient suction head, so cavitation is fully explained. A margin of at least 0.5–1.0 m should be provided; this installation is unacceptable even before considering further strainer blockage.
2. The suction gauge reads −35 kPa. Is this consistent with the calculation, and what would you check?
Read the worked answer
Gauge pressure at the pump suction relative to atmosphere is approximately ρg(z − h_L) = 9.75 × (2.5 − 4.0) = −14.6 kPa, not −35 kPa. The −35 kPa reading corresponds to a total suction loss of about 6.1 m, or to a gauge error, a gauge at a different elevation, or air/gas in the suction line.
Check gauge calibration and tap location, measure the differential across the strainer, inspect for a partially closed valve or collapsed strainer, and vent the casing. The inconsistency does not change the conclusion: the calculated NPSHa is already below NPSHr, so cavitation is expected. It does mean the actual suction loss may be even higher than assumed, making the situation worse.
3. Recommend immediate and permanent remedies.
Read the worked answer
Immediate: replace the fine strainer with the original coarse basket or a low-loss Y-strainer, clean the suction line, vent the pump, and reduce flow to about 150 m³/h until the suction is corrected. Do not throttle the suction valve; throttle the discharge if necessary.
Permanent: increase the suction pipe to 250 mm, reduce bends and fittings, lower the pump or raise the tank to give positive suction head, or use a pressurised suction tank. Alternatively, select a low-NPSH pump with NPSHr below 3 m at duty, or a vertical multistage pump with better suction characteristics. Recalculate NPSHa for worst-case strainer blockage, such as 50% blocked, and require at least 1 m margin. Add a differential-pressure gauge across the strainer and a low-suction-pressure alarm. After modification, verify with a cavitation test and vibration measurement.