Hydraulic Machines

MPE 437

1st Term 2024-2025

W8: Rotodynamic Pumps

Dr. Mohamed El-Ghandour

 Last Lecture
•Velocity triangles
•Some useful assumptions
•Effect of outlet blade angle
•Manometric (pump) head
•Losses in centrifugal pumps
•Pump performance
•Commence speed of a pump or minimum starting
speed
 Lecture outline
•Departures from Euler’s theory
•Slip Factor estimation
•Effect of the number of vanes
•Pump performance characteristics
•Pump operating range
•System characteristics
 DEPARTURES FROM EULER’S THEORY
The real flow in centrifugal pumps is very complicated and far
from the assumptions of the one-dimensional flow theory.
The reasons behind this reality are:

1- The velocities in the blade passages and at the impeller

outlet are not uniform due to the presence of blades and the
flow is 3 dimensional. This results in a reduction in the
velocity of whirl and, hence, reduces Euler’s head.

2- The second reason is that in a real impeller there are losses

of energy due to friction, separation and wakes associated
with the development of boundary layers. We will consider
these two effects separately.
Departures from one dimensional flow theory.
It results in a pressure difference between the two sides of the blade.
The fluid on this side of the blade is greater than that on the other
side, as indicated in the previous figure.
Hence, the velocity near the back side of the blade is greater than
that near the forward side.
This difference of velocity on the two sides of the blade gives rise to
the blade circulation.
But the non-uniform velocity distribution is responsible for the mean
direction of flow leaving the impeller, being 𝛽ሖ 2 = (β2 − φ) and not β2,
as assumed in the ideal flow situation.
This effect is responsible for deviation and results in the reduction of
the velocity of whirl from to the reduction being ∆Vw and called slip.
ሖ
𝑉𝑤2
Thus, the slip factor may be defined as 𝑆𝑓 = ൗ𝑉𝑤2
Sf can take values from 0 to 1
Practical values are 0.8 – 0.9
 Slip Factor estimation
Many authors attempted to predict the value of the slip factor. The
earliest was due to Stodola.
He assumed that ∆Vw2 = ωe
Where ω = u2/R2
𝜋𝑅2
e= sin 𝛽2
𝑍
e is the radius of a circle which may be inscribed between the two
adjacent blades at outlet, and Z is the number of blades
𝜋𝑅2
Hence, ∆Vw2 = (u2/R2)( sin 𝛽2 )
𝑍
𝜋𝑢2
= sin 𝛽2
𝑍

and 𝑉𝑤2 = 𝑢2 − 𝑉𝑓2 cot 𝛽2

Substitute these values in the slip factor equation
ሖ
𝑉𝑤2
ൗ𝑉𝑤2= (V w2−∆Vw2)
ൗV = 1- ∆Vw 2
𝑆𝑓 =
w2 Vw2
𝜋𝑢2 sin 𝛽2
= 1-
𝑍(𝑢2 − 𝑉𝑓2 cot 𝛽2 )
Stodol’s theory to predict the slip factor
 Effect of the number of vanes
The number of vanes must be sufficient to impart the energy to the
flow and at the same time it should not choke the flow. Further, it
should offer least amount of resistance to flow.

Too few no. of vane will increase secondary flow

Too many no. of vane will increase the friction losses.

A rule of thumb due to Stepanoff is Z = 𝛽2 / 3

𝛽2 in centrifugal pumps is from 20° to 40°.

Abo Elyamin et al. (2019) studied numerically the effect of number of vanes on the centrifugal
pump performance. They used 3 different impellers with 5, 7, and 9 blades at rotational speed of
2800 rpm.
They found that the head coefficient and the efficiency are higher for the case of the impeller with
7 blades than that for the two cases of 5 and 9 blades. The losses decrease by increasing the blades
number due to the reduction of the secondary flow for a certain limit. Then it increases again
because the friction losses in the impeller and the mixing losses after the impeller yield an
increment in the total losses in the region of the vaneless diffuser.
Yang et al., 2012 “Influence of Blade Number on the Performance and Pressure
Pulsations in a Pump Used as a Turbine
 Pump characteristics

• The head imposed upon the fluid (H) is the

energy given to it (E) minus any losses it has
incurred in travelling through the pump (him)
H = E - hlosses

H
 Expression for the head rise through a rotodynamic
pump
H = E - hlosses (1)

E = vw2u 2 / g (2)

hlosses = losses in impeller + losses in volute casing + kinetic (3)

energy at exit

The head loss in the impeller (him) depends upon the relative
velocity at exit h = k v 2 / 2g (4)
im im r 2

Similarly the loss in the volute hv = / 2g2

kvu2 (5)

The head rise across the pump is H and so including the dynamic
head at the delivery in the volute losses

H = vw2u2 / g − k im vr 2
2
/ 2g − kvu2
2
/ 2g (6)
• From the velocity triangles of whirless flow at
inlet
𝑤𝑒 𝑘𝑛𝑜𝑤 𝑡ℎ𝑎𝑡 𝑓𝑟𝑜𝑚 𝑡ℎ𝑒 𝑣𝑒𝑙𝑜𝑐𝑖𝑡𝑦 𝑡𝑟𝑖𝑎𝑛𝑔𝑙𝑒 𝑎𝑡 𝑒𝑥𝑖𝑡
𝑣𝑤2 = 𝑢2 − 𝑣𝑓2 cot 𝛽2 (7)
𝑎𝑛𝑑
2 + 𝑣2
𝑣22 = 𝑣𝑤2 (8)
𝑓2
𝑠𝑜 by substituting from 7 into 8 the value of 𝑣𝑤2 we get
2
𝑣22 = (𝑢2 − 𝑣𝑓2 cot 𝛽2)2 + 𝑣𝑓2 (9)
By extending the sq. bracket
2 cot2 𝛽 + 𝑣 2
𝑣22 = 𝑢22 − 2𝑢2𝑣𝑓2 cot 𝛽2 + 𝑣𝑓2 (10)
2 𝑓2
2 as a common factor, we get
Take 𝑣𝑓2
2 1 + cot2 𝛽
𝑣22 = 𝑢22 − 2𝑢2𝑣𝑓2 cot 𝛽2 + 𝑣𝑓2 (11)
2
as 1 + cot2 𝛽2 = cos 𝑒 𝑐2𝛽2
Then 𝑣 2 = 𝑢22 − 2𝑢2𝑣𝑓2 cot 𝛽2 + 𝑣2 cos 𝑒 𝑐2𝛽2 (12)
2 𝑓2

𝑎𝑙𝑠𝑜
2
𝑣𝑟2 2 cos 𝑒 𝑐2𝛽
= 𝑣𝑓2 2
 • By substituting in the pump head equation (H)
2
𝐻= 2 − 𝑘𝑣 𝑢22 − 2𝑢2 𝑣𝑓2 1 − 𝑘𝑣 cot 𝛽2 −𝑣𝑓2 𝑐𝑜𝑠𝑒𝑐 2 𝛽2 𝑘𝑖𝑚 − 𝑘𝑣 /2𝑔
now
u2 = ND2 / 60
and
v f 2 = Q /(D2b2 )
so
H = AN 2 − BNQ − CQ 2

𝐴 = 2 − 𝑘𝑣 (
𝜋𝐷2 2
) /2𝑔 • Where A, B, and C are
60
1 constants and B may be
𝐵= 1 − 𝑘𝑣 cot 𝛽2 positive or negative
30b2
1
2 depending upon the value
𝐶= 𝑐𝑜𝑠𝑒𝑐 2 𝛽2 𝑘𝑖𝑚 − 𝑘𝑣 /2𝑔 of Kv
D2b2
• By drawing the previous H-Q
equation for 𝑘𝑣 >1 and
𝑘𝑣 <1, two different curves
are appeared namely, rising
head curve and falling head
curve.

Kv > 1
Kv < 1

H-Q curve, (Munson et a 6th ed.)

• when the head curve continuously rises as the flowrate
decreases, and in this case the pump is said to have a rising
head curve. As shown by the next figure.

• pumps may also have curves that initially rise as Q is

decreased from the design value and then fall with a
continued decrease in Q. These pumps have a falling head
curve.
• The head developed by the pump at zero discharge is called
the shutoff head, and it represents the rise in pressure head
across the pump with the discharge valve closed.

• Although centrifugal pumps can be operated for short

periods of time with the discharge valve closed, damage will
occur due to overheating (power is dissipated as heat) and
large mechanical stress with any extended operation with
the valve closed.
• As can be seen from the figure , as the discharge is increased from zero
the brake horsepower increases, with a subsequent fall as the maximum
discharge is approached. As previously noted, the efficiency is a function
of the flowrate.
The curve reaches a maximum
value at some particula value of the
flowrate, commonly referred to as
the normal or design flowrate or
capacity for the pump.
• The points on the various curves
corresponding to the maximum
efficiency are denoted as the best
efficiency points (BEP).
• It is apparent that when selecting a
pump for a particular application, it
usually desirable to have the pump
operate near its maximum efficiency.
Typical performance characteristics for a
• The pump efficiency is around 75- centrifugal pump of a given size operating at a
80% (for turbine it is 85-95%)why?? constant impeller speed. (Munson et al., 6th ed.)
 What is the Preferred Operating Range (POR)?
The preferred operating range, or POR, is a region of the pump curve
defined in relation to the best efficiency point (BEP), and represents the
portion of the curve in which continuous operating conditions should be
located if possible.
The preferred operating range (POR) is defined by centrifugal pump
standards such as the Hydraulic Institute Standards (HI) or American
Petroleum Institute Standard 610 (API 610). This is in contrast to the
allowable operating region (AOR) which is a manufacturer-defined
region of the curve.
The POR is a narrow region within the broader AOR where the pump
exhibits the best efficiency and reliability. Operating beyond the AOR is
not recommended.
While the pump can still operate outside this range, it does so at the
expense of the reliability of the seals and bearings, as a number of
other conditions start to be a factor which will detrimentally impact
these items. Other parts of the pump, such as the impeller, volute and
shaft will also be subjected to the adverse operating conditions which
can contribute increased erosion and fatigue impact to these items.
https://www.introtopumps.com/pump-fundamentals/pump-terms/por-preferred-operating-range/
The POR, as defined by HI or API 610, varies depending on the type of
impeller, but in most cases extends from 70% to 120% of flow at the
best efficiency point (BEP).

In some high-energy pumps the POR is smaller, and may only extend
from 80% to 115% of flow at BEP.
To understand how the POR is determined, consider this example for a
standard centrifugal end-suction pump.

For a fictional end-suction pump, BEP falls at a flow of 1000 gallons

per minute (GPM).
The left side of the POR falls at 70% of BEP: 700 GPM in this example.
The right side of the POR falls at 120% of BEP: 1200 GPM in this
example.
The POR in this example extends from 700 GPM to 1200 GPM. Any
operating condition within those outer bounds falls within the POR.
Pump run-out
As every centrifugal pump does not operate reliably beyond a
certain flow rate, the published performance curve for that
pump is discontinued at that point.
This is referred to as the Run-Out condition.

Operation of the pump beyond that point (and often, even

approaching that point) will cause damage to the pump and
will also frequently overload the motor driver.

The damage caused at high flow rates will frequently be a

result of cavitation as the increase in flow rate through a
pump requires a much higher Net Positive Suction Head.
Further details on this matter will be discussed in later
section denoted “Cavitation in Rotodynamic pumps”.
Minimum flow point
On every pump curve, a number of
'minimum' flow points can be
identified, depending on the operating
requirements and equipment
reliability standards of the individual end
user.
The earliest minimum flow point used was that point on the curve at which the flow
was so low that it would result in a significant temperature increase of the pumpage.
When a centrifugal pump is operated very close to dead-head (also called “back” or
“to the left” of the curve), the discharge flow is throttled and reduced. With nowhere
else to go, this causes most of the fluid inside the pump to recirculate. This can
cause noise and erosion at the eye of the impeller and can also raise the
temperature of the fluid.
If you need to operate a centrifugal pump back on the curve, the minimum
suggested point is at 10% of BEP (Best Efficiency Point).
For example, a 620 pump with a full impeller has a BEP of 44%at 24 GPM.
Therefore, the minimum flow should be limited to 10% of 24 or 2.4 GPM.
 The pump is, of course, capable of pumping at lower flows;
however, excessive wear and shortened life may occur.
Suction Recirculation is a condition with similar symptoms to
Cavitation that will be discussed more fully in this Chapter.
It is a condition created by low flow operation and it frequently dictates the
low flow limit of stable operation discussed earlier above.

In some industries it is referred to as the 'Minimum Flow for Stable Operating Condition'.
Discharge Recirculation is another condition precipitated by low flow
operation that takes effect at a lower flow than Suction Recirculation,
and also displays similar symptoms.
Performance curves for a two-stage centrifugal pump operating
at 3500 rpm. Data given for three different impeller diameters.
(Munson et al., 6th ed.)
 System characteristics

2 4
Vd
HP = Hm = Hs +Hd +hfs +hsd +

Hd
2g 3
H P = H m = Hstatic +  h f
H sta tic =  z = z 2 − z1
 hf = f (Q 2 )
2

Hs
H = H static + kQ 2
1

(Munson et al., 6th ed.)

 Pump Operating points
To select a pump for a particular
application, it is necessary to
utilize both the system curve,
determine by the system
equation, and the pump
performance curve.
If both curves are plotted on the
same graph, as illustrated in the
figure.
their intersection (point A)
represents the operating point for
the system. That is, this point give
the head and flowrate that
satisfies both the system equation Utilization of the system curve and the pump
performance curve to obtain the operating
and the pump equation. point for the system. (Munson et al., 6th ed.)
• On the same graph the pump efficiency is shown.
Ideally, we want the operating point to be near
the best efficiency point (BEP) for the pump.
• For a given pump, it is clear that as the system
equation changes, the operating point will shift.
• For example, if the pipe friction increases due to
pipe wall fouling, the system curve changes,
resulting in the operating point A shifting to point
B in the previous figure with a reduction in
flowrate and efficiency.
• The following example shows how the system
and pump characteristics can be used to decide if
particular pump is suitable for a given pplication.
example

(Munson et al., 6th ed.)

• To solve
– Apply bernoulli’s equation
between 1 and 2
– Obtain expression for the
pump head as:
– H=Hs + kQ2
– Build a table for Q and H by
assuming the values of Q
and calculating the value of
H from the system
equation
– Draw the system
characteristic curve
– Find the intersection point
(operating point)