Cavitation

I liked his graphics and it is a good basic introduction to the concept of cavitations.

I'd add that 99% of pumps operate with some degree of cavitation present. Just because cavitation is present does not automatically mean damage will occur. The critical thing is how much cavitation is present and the intensity of it.

Cavitation will also generate noise and vibration which may be the primary concern since this will reduce the seal and bearing life of the pump.

Users might want to familiarize themselves with the new Hydraulic Institute standard on NPSH margin, which is a useful guide to avoiding damaging cavitation.

Yes, agreed Simon. I suggest to size your pump for lower speed whenever possible. This however, calls for a bigger pump and more expensive unit but worth it. At lower speed, your pump NPSHr is lower.

Good video.

All good advices. Other references such as PUMP CAVITATION. VARIOUS NPSHR CRITERIA, NPSHA MARGINS, AND IMPELLER LIFE EXPECTANCY by Bruno Schiavello and Frank Visser (available from Amazon) and CAVITATION AND RECIRCULATION TROUBLESHOOTING METHODOLOGY also by Bruno (Proceedings of the 10th International Pump Users Symposium) are quite useful.

I know of an extraordinary case. Pump designed for 960 RPM was speeded up and used for larger capacity/head characteristics in a fertiliser plant in mid-1060s. There were four units in parallel operation. This split case pump developed severe cavitation problems in service, leading to vibrations, erosion at impeller inlet and unstable operation when impeller was worn. This was referred to a teaching institute, investigated and resolved by designing a new impeller on same casing.
Moral of the story is that this is a problem due to limitations of a given design. On some occasions, it can be solved by having an alternate impeller in same casing. This is not a universal solution.

All good comments but let's back up a little. First issue is designing the system. It seems this discussion is focused upon centrifugal pumps. Typically design flow is not an issue but when it comes to system design data. The fear of falling short leads to typically overstating the tdh value (by 5% or even more). It is generally advised to select the pump at or near best efficiency. The NPSHr curve is relatively flat moving to the right of shut-off head but typically begins to rise rather sharply approaching BEP. With slower speed the flow increases more dramatically with a relatively small head change. If head is overstated the operating point moves right and with it the NPSHr value increases. So determining NPSH and related safety margin begins with very accurate system head design data.

Absolutely Robert, system design, whatever the concept of pump within it, is of paramount importance. Many cases of noisy pumps diagnosed as cavitation result from aeration induced by the design of the suction pipework. This is an area where many installers fail because of a lack of pump knowledge. Pumps are placed in convenient positions which may not be the optimum position for trouble free operation.

In addition to the comment of Robert; centrifugal pump according DIN/ISO standard have the indicated pressure in the curve available on the SUCTION side of the pump and not on the discharge side. According the standard the internal losses are part of the delivery head. With better friction loss calculations and less oversized systems this is critical on total delivery and total pump performance.

I would like to express my thanks to all the contributors for not only the comments but for referring precisely to more sources of reference and information. God bless you all!