My NPSH (required) is better (lower) than yours

A good thing? At first glance the answer is obvious, of course! closer examination of the facts reveals a different story. In order to reduce the NPSHr, the pressure in the eye of the impeller will have to be increased. As NPSHr is a pump condition, none of the external factors associated with NPSH available come into play, enter one Mr Bernoulli. This gentleman stated that as the velocity of a liquid increases, its pressure will decrease and conversely, as the speed of the liquid decreases, then the pressure will increase. Problem solved, because all that needs to be done for a given flow rate is to increase the cross sectional area that the fluid has to pass through and this will ensure a drop in velocity and a concomitant increase in the pressure. From all of this, if the impeller eye area in a centrifugal pump is increased, then the pressure will increase, this means a reduction in the pump's NPSHr. Take a look at the two diagrams.

 Figures 1 & 2 show two impeller designs which deliver the same head and capacity. Impeller 2 has a larger eye than impeller 1 and from Bernoulli's law, impeller 2 will have a lower NPSHr due the the increased eye pressure. The lower NPSHr comes at a cost however. By opening the impeller eye,the designer has increased the risk of recirculation occurring at the vanes
(Figure 2). This is fertile ground for the formation of cavitation with all the vibration, noise and metal removal associated with this condition.

By solving one problem, a potentially far larger one has been created. The further left of the Best Efficiency Point (BEP) the pump is operated, the more exaggerated the problem becomes. So how does the calculation of suction specific speed help in avoiding the problem? As was shown in part one of this series, Nss numbers higher that 175 indicate a pump that is prone to generating suction recirculation cavitation. Nss numbers of 175+ mean that the designer and the end-user have to closely examine the TDH range that the pump will "see”. If there is a possibility that the duty point could move less than about 80% of the flow rate at BEP, then caution is advised.

It has been said that due to advances made in the Computational Fluid Dynamics (CFD) field, the problem of high Nss pumps failing due to recirculation cavitation has been largely done away with. This might have a measure of truth in it but how many pump ranges that are currently on the market were designed in the seventies and early eighties?

The case study covered in Part 1 concerned three large end suction pumps. My request to the local agent brought to light THE ACTUAL TEST CURVES OF THE THREE PUMPS IN QUESTION AS TESTED IN 1975! Truly amazing service! It can, however, be safely said that these pumps were probably designed in the middle sixties, in the days before CFD and fast computers became freely available.

If NPSHr is a major design criterion in your next pump system, beware of bearers of “good” (NPSHr) news!