Section 2.1.1 - Centrifugal Pump Theory - Design Procedures

Establishing the Pump Configuration   The first step in designing a pump is to determine
the type and number of stages that are needed to meet the given set of operating
conditions, usually Q, p₁, p₂, available NPSH (= NPSHA) and specific gravity of the fluid.
If the pump must meet several such sets of operating conditions, one set has to be chosen
for the BEP or design point so all the others are satisfied, if possible. Making the proper
choice of this BEP may require some iteration: first making a trial choice, doing a preliminary
design, and determining the corresponding pump performance characteristic
curves, and then repeating these steps if necessary.

Pump rotative speed N (rpm) must be chosen in order to proceed. Selection of the highest
practical rpm is desirable because it yields the smallest size and therefore usually the
lowest cost and easiest containment of system pressure. If, for the chosen number of
stages, the stage specific speed is too low, Figure 10 indicates that efficiency is generally
improved with greater speed. The maximum possible rpm is that which yields a value for
the suction specific speed Ω_ss (Eq. 41) or N_ss (Eq. 42) that the first stage of the pump can
accommodate, where NPSH is found respectively from Eq. 41 or Eq. 42 as follows:

(45)

Typically, the Ω_ss-capability of an impeller does not exceed a value that is somewhere
in the range 3 to 4.5 (N_ss = 8,200 to 12,300), depending on minimum (off-design) flow rate
requirements, and it is typically 10 (27,000) or less for inducers. Generally, the supplier
furnishes a pump that has a value of NPSH-capability (or “NPSHR”) smaller than the
stated NPSHA. This difference or “NPSH-margin” is desirable if there is any uncertainty
as to the true value of NPSHA. It is essential for high values of Δp_stg; (see discussion of
“high-energy pumps” further on in this section). Often a higher speed can be employed
if a double-suction impeller (entry of fluid from both sides) can be used, as then only half
of the given Q enters each side of this impeller, and only that half can be used in the
Ω_ss-equation applying to that fluid and impeller type.

The double-suction configuration is popular for large, single-stage pumps because the
axial thrust is nominally zero. It is also found as the first stage in some multistage pumps,
in which case the arrangement of the remaining impellers can be “back-to-back” (half of
them facing in one direction and half opposite) to achieve axial thrust balance. On the
other hand, the interstage flow passages are simpler if these impellers are all facing in the
same direction—in which case the thrust is opposed by a balancing drum or disk as
described in Section 2.2.1. Forming the specific speed from N, Q, and ΔH (from Eq. 3) and
referring to Figure 9 for the kind of impeller (which then gives one an idea of the other
hydraulic components (inlet passage, diffuser and collector, and so on) and Figure 10 for
the expected efficiency, one can decide whether the pump would perform better and still
meet any installation and size restrictions with more than the one stage implied by this
first N_s-calculation.

If the user requires the performance characteristics of a positive displacement pump—
such as a wide range of pressure-rise at a nearly constant flow rate Q—selecting such a
pump may be possible if the specific speed is not too high. (See Figure 9.) The choice of
speed for positive displacement pumps is sometimes determined by mechanical considerations
rather than suction capability.

Establishing the Pump Configuration   The first step in designing a pump is to determine
the type and number of stages that are needed to meet the given set of operating
conditions, usually Q, p₁, p₂, available NPSH (= NPSHA) and specific gravity of the fluid.
If the pump must meet several such sets of operating conditions, one set has to be chosen
for the BEP or design point so all the others are satisfied, if possible. Making the proper
choice of this BEP may require some iteration: first making a trial choice, doing a preliminary
design, and determining the corresponding pump performance characteristic
curves, and then repeating these steps if necessary.

Pump rotative speed N (rpm) must be chosen in order to proceed. Selection of the highest
practical rpm is desirable because it yields the smallest size and therefore usually the
lowest cost and easiest containment of system pressure. If, for the chosen number of
stages, the stage specific speed is too low, Figure 10 indicates that efficiency is generally