Q - Does the NPSHR of an impeller change when its diameter is cut?
A - No. The impeller NPSHR is dependent on its suction geometry such as the
eye diameter, eye area, and vane inlet angle. Cutting the impeller diameter has no
effect on the suction geometry, and should have no effect on NPSHR.
Q - Most pump performance curves shows the same NPSHR curve regardless
of impeller diameter but there are also some that show two NPSHR curves: one
for maximum impeller diameter and another for minimum diameter. I noticed that
the NPSHR for minimum diameter is higher than for maximum diameter. You said
that NPSHR does not change with impeller cut diameter. Please explain this
A - You raised a very interesting point. Our sources think that this has more to do
with how NPSHR is determined rather than with actual change in NPSHR. These
explanations were given:
This is an anomaly that is caused by internal flow recirculation and how flow rate
is measured. In some instances cutting the impeller diameter will result in higher
internal flow recirculation. Radial flow impellers whose ratio of eye diameter to
impeller diameter is greater than 0.50 are more sensitive to this phenomenon.
For clarity, assume a hypothetical situation where a pump has:
Qd, flow at discharge nozzle = 100 GPM
Qi, internal flow recirculation = 10 GPM
Qs, flow at impeller suction, Qd+Qi = 110 GPM
During test the NPSHR is measured at 100 GPM discharge flow rate. That
NPSHR is considered the NPSHR at 100 GPM but in reality it applies to 110 GPM,
the actual flow rate at the impeller suction, of which 10 GPM is due to internal
Now assume that the impeller diameter is cut and, as a result, the internal
recirculation flow increases to 15 GPM. The NPSHR at 100 GPM discharge flow
will now appear to be higher because in reality it is the NPSHR for 115 GPM
The problem is that there is no practical way to measure internal flow recirculation
and hence there is no practical way to correct the NPSHR.
It is common practice in the pump industry to measure NPSHR based on a 3%
head loss. But 3% of what head? If a pump is tested at maximum impeller
diameter then it is 3% of the head at maximum impeller diameter; if tested at
minimum diameter, then it becomes 3% of the head at minimum diameter. In my
opinion this practice is inconsistent and needs to be corrected.
Here's why: Say that a pump, at 1000 GPM, has a head of 500 FT at maximum
impeller diameter. Under the present industry-wide practice a head drop of 15 FT
will determine its NPSHR. But if that same impeller is cut 20% of its diameter and
the head at 1000 GPM is reduced to 300 FT then its NPSHR will be based on a
head drop of only 9 FT.
See the inconsistency? The same pump, with same impeller and same suction
geometry, ends up with different NPSHR at same capacity because the NPSHR
are based on different absolute values of head drop. This seems to explain why
the NPSHR at impeller cut diameter may at times appear to be higher.
Specific speed (NS) and suction specific speed (NSS) are always calculated
based on data at maximum impeller diameter for data consistency. It makes
sense that NPSHR should also be based on 3% head drop at maximum impeller
diameter regardless of actual impeller cut diameter.
C: basics, operation
"Make it simple"