Cavitation in a Sump Pump: The Worst Case is the Minimum Sump Level
A 20 L/s centrifugal pump lifts water from a sump to a tank 28 m above the pump centreline. The pump sits on a platform 5.5 m above the minimum sump level, which occurs at the end of the dry season. The suction line is 100 mm diameter, 9 m long, with a foot valve (K = 2.5), two 90° bends (K = 0.3 each) and a friction factor of 0.024. The pump curve states NPSH required = 3.5 m. The site is at 300 m altitude; water temperature in May is 32 °C, when the vapour pressure is 4.8 kPa. In the second summer the operators report rattling, fluctuating discharge and pitting near the impeller eye. The suggestion on the table is to speed the pump up by 10% to recover the lost discharge.
1. Compute the available NPSH at the worst operating condition and compare it with the requirement. State clearly what condition governs.
Read the worked answer
Step 1 - Suction velocity and velocity head. A = π x 0.100²/4 = 7.854 x 10⁻³ m²; v = 0.020/7.854 x 10⁻³ = 2.55 m/s; v²/2g = 0.330 m.
Step 2 - Suction losses. hf (pipe) = f L v²/(2g d) = 0.024 x (9/0.1) x 0.330 = 0.714 m. Minor: foot valve 2.5 x 0.330 = 0.826 m; two bends 0.6 x 0.330 = 0.198 m. ΣhL = 0.714 + 0.826 + 0.198 = 1.74 m.
Step 3 - Pressure heads. Ha at 300 m altitude ≈ 10.0 m of water; Hv = 4.8 kN/m² / 9.81 = 0.49 m.
Step 4 - NPSHa. NPSHa = Ha − Hv − Hs − ΣhL = 10.0 − 0.49 − 5.5 − 1.74 = 2.27 m. (If velocity head at the flange is counted separately, as some texts do, it is even less - about 1.94 m - but the conclusion is the same.)
Step 5 - Comparison. NPSHr = 3.5 m; shortfall = 3.5 − 2.27 = 1.23 m.
Judgement. Cavitation is unavoidable at the end of the dry season. The governing condition is minimum sump level, not the normal or monsoon level at which the pump was evidently commissioned and tested. During the monsoon the sump is full, Hs drops by 2–3 m, NPSHa rises to about 4.5–5 m and the pump sounds normal - which is exactly why the problem appeared only in the second summer and was misread as a pump defect. The pitting at the impeller eye, rattling and fluctuating discharge are textbook cavitation symptoms, not wear.
2. Evaluate the three proposals on the table - lower the pump by 1.5 m, replace the suction line with a 150 mm pipe, or increase the pump speed by 10% - and quantify each.
Read the worked answer
Proposal 1 - Lower the pump platform by 1.5 m. NPSHa becomes 2.27 + 1.50 = 3.77 m against 3.5 m required: margin only 0.27 m. Lowering by 2.0 m gives 4.27 m, a usable margin. Cheapest per unit of benefit only if the platform can be cut down without structural work and the delivery head increase (28 m + 2 m) is acceptable to the pump curve.
Proposal 2 - Replace the suction with 150 mm diameter. v = 0.020/0.01767 = 1.13 m/s; v²/2g = 0.065 m. hf = 0.024 x (9/0.15) x 0.065 = 0.094 m; minor = 3.1 x 0.065 = 0.202 m; ΣhL = 0.30 m. NPSHa = 10.0 − 0.49 − 5.5 − 0.30 = 3.71 m: margin 0.21 m. The gain (1.44 m) is nearly all from reduced losses, and the answer is sensitive to the foot valve: a 150 mm foot valve of better design (K ≈ 1.5) would add roughly 0.1 m instead of 0.2 m. Combined with a 0.5 m lowering of the pump, NPSHa ≈ 4.2 m - the robust option, and it keeps the delivery head unchanged.
Proposal 3 - Speed the pump up by 10%. This is the wrong direction. NPSHr rises approximately with N²: 3.5 x 1.1² = 4.24 m, while the higher flow raises suction losses by about 1.21 x 1.74 = 2.10 m, so NPSHa falls to 10.0 − 0.49 − 5.5 − 2.10 = 1.91 m. The shortfall more than doubles. Speeding up will destroy the impeller in weeks.
Recommendation. Adopt 150 mm suction plus a modest lowering (0.5–1.0 m), fit a low-loss foot valve, and keep the speed unchanged. Verify NPSHa against the pump curve at the maximum summer water temperature and the minimum sump level before restarting.
3. Explain why the symptoms appeared in the second summer rather than immediately, and what operating rule you would issue to the pump house staff.
Read the worked answer
Why the delay. At commissioning the sump was almost certainly at a high level, so Hs was small - perhaps 2.5–3 m instead of 5.5 m - giving NPSHa of about 4.8–5.3 m, comfortably above 3.5 m. Cavitation inception is also progressive: the first bubbles collapse with little noise and only slight loss of head, and damage to the impeller eye accumulates over hundreds of hours before the hydraulic performance visibly degrades. By the second dry season the impeller surface had already been eroded, which further reduces NPSHr margin and accelerates the deterioration.
The physics in one line. Cavitation occurs when the absolute pressure at the impeller eye falls to the vapour pressure. NPSHa falls with (i) rising suction lift as the sump empties, (ii) rising water temperature and hence vapour pressure, and (iii) rising suction losses as flow increases. All three worsen in May.
Operating rule to issue.
- Fix a minimum operating sump level (marked on the gauge board) at which NPSHa = NPSHr + 0.5 m, computed with the summer vapour pressure; the pump must be interlocked to trip or throttle below this level.
- Record suction vacuum, discharge pressure and flow at every shift; a falling suction vacuum with rising flow is an early cavitation warning.
- Do not increase speed or throttle the suction valve to chase discharge; regulate on the delivery side only.
- Inspect the impeller eye and suction bell every pre-monsoon shutdown and record the pitting; replace the impeller rather than attempting weld repair on a cavitated eye.
- Re-commission with a measured NPSH test at the minimum sump level, and keep the sump free of floating debris that would raise the foot-valve loss coefficient.