Centrifugal Pump Cavitation in Summer Operation
A dairy plant uses a 15 kW centrifugal pump rated at 25 m³/h and 28 m head, with a required NPSH of 3.2 m. The pump draws water from an underground sump whose minimum water level is 4.2 m below the pump centreline. The suction line is 8 m of 80 mm internal diameter pipe with two 90° elbows, a foot valve and a strainer; the total suction loss at duty flow is 2.1 m. The site atmospheric pressure is 100 kPa and the water density is 996 kg/m³. In winter the water is 30°C (vapour pressure 4.25 kPa); in summer it reaches 45°C (vapour pressure 9.6 kPa). Operators throttle the discharge valve when they hear noise from the pump, believing this protects it. During the last summer the pump cavitated, the mechanical seal failed twice, and the maintenance team replaced the seal without checking suction conditions. The manager wants to know whether the pump is installed correctly and what to change before the next summer.
1. Calculate the available NPSH at 30°C and 45°C. Decide whether cavitation is likely and explain why throttling the discharge is not a reliable cure.
Read the worked answer
NPSHa = (p_atm - p_v)/(ρg) - H_s - h_f. At 30°C: (100,000 - 4,250)/(996 × 9.81) = 9.80 m; NPSHa = 9.80 - 4.2 - 2.1 = 3.50 m. Margin over NPSHr 3.2 m is only 0.30 m, which is too small for normal control. At 45°C: (100,000 - 9,600)/9,770.8 = 9.25 m; NPSHa = 9.25 - 4.2 - 2.1 = 2.95 m, which is below the required 3.2 m. Cavitation is therefore expected in summer and the repeated seal failures are consistent with it. Throttling the discharge may reduce suction loss slightly at lower flow, but it does not reduce water temperature and can increase NPSHr at low flow because of recirculation; it is not a reliable cure and wastes energy. The installation is marginal in winter and inadequate in summer.
2. Propose two practicable modifications to achieve at least a 1.0 m NPSH margin in summer. Support each with a calculation.
Read the worked answer
Target NPSHa at 45°C of at least 4.2 m, i.e. a 1.0 m margin over NPSHr. Current NPSHa is 2.95 m, so the suction head plus loss must be reduced by at least 1.25 m. Option 1: lower the pump or raise the minimum sump level by 1.3 m; then H_s = 2.9 m and NPSHa = 9.25 - 2.9 - 2.1 = 4.25 m. Option 2: replace the 80 mm suction line with 100 mm and remove one elbow; if the loss falls from 2.1 m to 0.8 m, NPSHa = 9.25 - 4.2 - 0.8 = 4.25 m. Option 3: cool the water or pressurise the suction vessel, but that is costly. Do not throttle the suction line. Check the foot valve and strainer for blockage, and verify the actual duty flow before sizing. A combination of a 0.8 m lower pump and a larger suction line would give a safer margin.
3. What operating, instrumentation and maintenance changes would you make? Explain why discharge throttling can worsen low-flow cavitation risk.
Read the worked answer
Throttling the discharge shifts the operating point to lower flow, which reduces pipe friction and can raise NPSHa slightly, but it also raises the risk of low-flow recirculation, recirculation cavitation and temperature rise in the casing. It does not address the root cause: the suction lift and losses are too high for the summer vapour pressure. Fit a compound gauge on the suction, a pressure gauge on the discharge, a flow meter and a temperature indicator, and record NPSHa margin at the worst summer condition. Add a low-flow protection or VFD only as an operating measure, not as a substitute for suction-side correction. Keep the strainer clean, check for air ingress, and monitor vibration and seal leakage. The operating procedure should prohibit using the discharge valve to stop cavitation noise; the correct response is to stop, investigate suction conditions and correct them.