As described in the previous topic, pumps need a certain pressure at the inlet to avoid cavitation.
Such pressure is made available by the system, is expressed in meters of water column, and is therefore called Net Positive Suction Head Available (NPSHa).
NPSHa must be compared with the NPSHr (the pressure needed by the pump). There is an extra safety margin to be added, to guarantee the safest possible conditions. The safety margin can be calculated in 2 ways: as a percentage of the NPSHr (add 25% margin), or as a fixed quantity (add 0.6m of water column), whatever is larger should be used.
NPSHa depends on four main factors:
The formula that describes their relationship is
with
As clarified at the beginning of this topic, NPSHa should be as large as possible, so that it is larger than the NPSHr over the working flow-rate range of interest. To maximize NPSHa, we can can observe in the equation above how increasing the pressure Pa above the free surface (i.e. by pressurizing the vessel) has a positive effect. Increasing the height difference between pump inlet and free surface level also helps, as the Z1 term will increase. The pressure losses Ploss term must be reduced as much as possible, as it increases with the square of the flow rate. To do so, the tubes at the inlet side must be as large as reasonably possible, as short as possible, and the components that cause resistance to the flow (e.g. elbows, fittings, filters, heat exchangers, regulation valves, etc) must be moved downstream of the pump. Lastly, the fluid itself should be as far as possible from the boiling point (Pvap as low as possible), meaning that the temperature should be kept as low as possible, compatibly with the process (no unnecessary heating upstream of the pump).
Graphically, the NPSHr appears as a set of curves (as shown in the previous topic) increasing with the flow rate. At the same time, the NPSHa has the shape of an upside-down parabola (due to the -Ploss term, offset by the Pa–Pvap and Z1 terms) as shown in Figure 2.
As shown in Figure 2, there is no risk of cavitation if the flow-rate is below 32 lpm, because the NPSHa is larger than the NPSHr + safety margin. Between 32 lpm and 35 lpm there a chance of cavitation, because the safety margin is violated. For flow-rates larger than 35 lpm, cavitation will occur.
If the system that produced the plot in Figure 2 gets improved (upstream vessel pressurized, elbows removed, larger tubes), it could produce a new combination of curves, such as in Figure 3. It is clear that now the cavitation will be very unlikely over most of the operative range.