Tying this together with your second posting regarding vortexing which I just read - the noise you hear which you think is cavitation is actually mis-match of flow onto the impeller blades, very noisy / destructive as there is probably a lot of localised cavitation happening even though you have surplus NPSHa.
The pump is not suffering suction cavitation, it is being subjected to pre-rotation of the inlet flow.
You really should provide more details on the pump and system so we can provide meaningful comments. It is possible for a pump to cavitate even if it has a very high NPSH margin. The most obvious causes are pre-rotation or sump configuration for vertical turbine pumps, operation at excessively high or excessively low flow, poor suction line configuration for a double suction impeller, vortexing, dissolved gases, hidden suction line restrictions (start-up strainer, dropped valve gate), etc. And since NPSH(required) is usually defined as cavitation sufficient to produce a 1% or 3% head loss, the pump manufacturer never tells you what suction head is required to suppress all cavitation (insipient cavitation). I have pumps that cavitate constantly and destroy the impeller every 6 months even running with ideal suction piping and a 100 foot NPSH margin. On paper, these pumps should not cavitate at all. But they do. After battling for many years, we finally found a design mistake regarding the diffuser throat geometry. .
Is pre-rotation the same thing as recirculation? If so, what is the recommended lower limit? Is it 70% of the flow at BEP or do you base it off the NSSS?
Possible conclusion to my ordeal... I got an in between bearing pump which takes suction (2 impeller eyes) 16-20ft below the vessel. The suction piping has a total run of 20ft with 3 90deg ells. the pump is running below 70%of the flow at BEP. (This has never been an issue since the pump has run at lower flows). however, there has been changes on the process conditions. The vapor space pressure in the vessel was reduced and the process temp has gone up. the adverse affects of the change (vapor pressure of the liquid went up and vapor space pressure decreased) resulted in the reduction of the NPSHa.
Seems we need a lot more detail, you should post the full operating conditions, pump size, BEP flowrate / head, operating flowrate / head, temp, levels, post a picture of the pump operating curve etc etc.
As you haven't had any problems until changes were made in the operating conditions it seems obvious that is the problem.
Smack, in your inital post you mentioned ample NPSHa was provided, then later you mentioned a reduction in NPSHa due to process changes. If your running less then 70% of BEP flow, I doubt NPSH r or a is your problem. How far are the 90 deg elbows from the suction flange and where are you taking pressure measurments to calculate your NPSHa? If the 90's are too close to the pump suction flange, you probably have an uneven flow distribution through the two impeller eyes (one suction side of the impeller has a higher flow then the other).
