PUMPS

A rotodynamic pump is essentially a turbine ‘in reverse’; mechanical energy is transferred from the
rotor to the fluid. Like turbines, pumps are classified according to the direction of the fluid path
through them: there are thus

(i) Radial flow (or centrifugal)

(ii) Axial – flow, and
(iii) Mixed flow

CENTRIFUGAL PUMP

Centrifugal pumps consist of an impeller rotating within a spiral casing. The fluid enters the pump
axially through the suction pipe via the ‘eye’ of the impeller; It is discharged radially from the
impeller around the entire circumference either into a ring of stationary diffuser vanes or directly
into the casing. The casing collects the fluid, decelerates it – converting some of the kinetic energy
into pressure energy – and finally discharges the fluid through the delivery flange.

Centrifugal pumps are used for high head, low flow situations. There are several different types of
commonly used centrifugal pumps including: end suction, double suction, multiple stage, and split
case.

Centrifugal pumps are often physically located above the water source and water is permitted to
enter the pump through a suction pipe.

ANALYSIS OF CENTRIFUGAL PUMPS

Consider a centrifugal pump showing a small part of the impeller complete with vector triangles of
the flow pattern at entry and exit.

First consider the situation at entry, unlike turbines it is not usual for pumps to have guide vanes at
entry, therefore, the fluid enters the impeller with no whirl velocity, hence the absolute velocity V1 is
equal to the flow velocity Vf1.

Consideration of the inlet vector triangle leads to the following:

1
 (i) Since there is no whirl component at inlet velocity, V1 = Vf1, this means that the angle
between the vector V1 and u1 (the speed of the impeller tip at inlet) is a right angle.

(ii) In order to minimize energy loss the fluid should impinge on the blade tangentially, in
terms of the vector triangle this means that the angle between u1 and Vr1 should equal
the blade angle, θ.

When fluid enters the impeller tangentially, this is called shockless entry and is regarded as the
design condition. If, however, the fluid enters non-tangentially i.e. under non-design condition
the following would occur:

(a) Impact losses occur

(b) Boundary layer separation takes place
(c) Eddies arise which give rise to some backflow into the inlet pipe, this causes the incoming
flow to have some whirl velocity.

The result of non-tangential entry leads to a drop in the efficiency of the pump.

For a centrifugal pump under design conditions,

 V1  D1 N Q
  tan 1   ’ u1  , V1  V f 1 
 u1  60 D1b1

Where, D1 = diameter of the impeller at inlet

b1 = width of the impeller at inlet

 60Q 
   tan 1  
  D2
1
2
b1 N 

For the exit triangle, three cases have to be considered:

(i) Forward facing blades, β > 90o

2
 Vr2 Vf2

Thus, Vw2  u2  V f 2cot   

Or Vw2  u 2  V f 2 cot 

(ii) Radial blades, β = 90o

Vr2 = Vf2

(iii) Backward facing blades, β < 90o

3
Therefore,

Vw2  u 2  V f 2 cot 

Using the Euler equation for hydraulic machines, the work done per unit weight of fluid is

Ho 
1
Vw1u1  Vw2 u 2  i.e. for turbines
g

However, for pumps the sign in the equation changes, i.e.

Ho 
1
V2 u 2  Vw1u1 
g

But Vw1 is set to zero,

Vw 2 u 2
Ho  (theoretical head)
g

Vw 2 u 2
The head imposed on the fluid is the energy given to it less any losses, hi, in travelling
g
through the impeller. As the fluid leaves the impeller and enters the volute a relatively small amount
of the total energy is potential (i.e. pressure) energy much of it is kinetic; this has to be converted to
potential energy by the volute and diverging delivery pipe. However efficiently the volute converts
the kinetic energy to potential, there is still a head loss, hv.

The energy conservation equation can then be written in the following form:

Vw 2 u 2 V p2
 H  hi  hv 
g 2g

Where Vp is the velocity of flow in the outlet pipe, H is the net head.

2
Vw 2 u 2 Vr22 V22 V p
 H  ki  kv 
g 2g 2g 2g

Vp is very small compared to the other terms and can safely be ignored.

This expression can be written in terms of H-Q variation:

Since V22  Vw22  V f22

V22  u 2  V f 2 cot    V f22

2
Then

 
V22  u 22  2u 2V f 2 cot   V f22 1  cot 2  
4
Or V22  u 22  2u 2V f 2 cot   V f22 cos ec 2 

V f22
Also V 2
  V f22 cos ec 2 
sin 
r2 2

Therefore
2u 22  2u 2V f 2 cot   k iV f22 cos ec 2   k v u 22  2k v u 2V f 2 cot   k vV f22 cos ec 2 
H
2g

ND2 Q
But u2  and Vf 2 
60 D2 B2

Hence the equation can conveniently be re-written as

H  AN 2  BNQ  CQ 2

Where A, B and C are constants defined by the properties of the pump. This equation demonstrates
that the H-Q relationship for a pump is parabolic, and examples of the shape of the curves for
different blade angles are given in the figure below.

For water, backward facing blades are usually preferred because the absolute velocity, V2 is lower
therefore the velocity head in the volute is less resulting in a reduced head loss and a higher
efficiency. Though they give a smaller head for a particular size and flow rate, this disadvantage is
usually outweighed by the greater efficiency obtained.

5
MANOMETRIC EFFICIENCY

Manometric efficiency is the ratio of the manometric head, Hm, to the Euler head, Ho. The
manometric head is the difference in head that would be recorded on a manometer connected
between the suction and delivery flanges of the pump. Thus,

Hmg
 m an 
Vw 2 u 2

It represents the effectiveness of the pump in producing pressure from the energy given to the fluid
by the impeller.

The heads Hm and H are related by the expression

Hm  H 
V d
2

 Vs2
h
2g

Where Vd is the velocity of flow in the delivery pipe, Vs is the velocity of flow in the suction pipe and
h is the difference in elevation between the inlet and outlet flanges of the pump.

Vd Delivery
p
Vs

Suction

Manometric efficiency is also known as hydraulic efficiency.

VOLUMETRIC EFFICIENCY, ηv

Q
v 
Q  QL

Where Q = discharge actually delivered and QL = leakage of discharge

MECHANICAL EFFICIENCY, ηm

𝑏𝑟𝑎𝑘𝑒 𝑝𝑜𝑤𝑒𝑟−𝑝𝑜𝑤𝑒𝑟 𝑙𝑜𝑠𝑠 𝑑𝑢𝑒 𝑡𝑜 𝑓𝑟𝑖𝑐𝑡𝑖𝑜𝑛 𝐵𝑃−𝑃𝑓

ηm = =
𝑏𝑟𝑎𝑘𝑒 𝑝𝑜𝑤𝑒𝑟 𝐵𝑃

6
OVERALL EFFICIENCY, ηo

This is the ratio of the power given to the fluid by the shaft power.

gQH WP
o  
BP BP

Where BP = shaft power (brake power), WP = water power

This is appreciably lower than the manometric efficiency, because additional energy has to be
supplied by the shaft to overcome friction in the bearings, and in the fluid in the small clearances
surrounding the impeller.

Example

A centrifugal pump impeller has an outer diameter of 30 cm and an inner diameter of 15 cm. The
pump runs at 1200 rpm. The impeller vanes are set at a blade angle of 30o at the outlet. If the
velocity of flow is constant at 2.0 m/s, calculate

(i) The velocity and direction of water at outlet

(ii) The head developed, by assuming a manometric efficiency of 0.85
(iii) The blade angle at the inlet

PUMP AND SYSTEM CHARACTERISTICS

PUMP CHARACTERISTICS

Pump characteristics are the head (H), input power (BP) and efficiency (  o ) as functions of discharge
(Q). The most important is the H-Q relationship. Typical shapes of these characteristics are as
follows:

7
Ideally, one would like to operate the pump:

(a) As close as possible to the design point (Point of maximum efficiency);

(b) In a region where the H-Q relationship is steep; (Otherwise there are significant fluctuations
in discharge for small changes in head).

SYSTEM CHARACTERISTICS

In general the pump has to supply enough energy to:

(a) Lift water through a certain height – the static lift HSL
(b) Overcome losses dependent on the discharge, Q.

Thus, the system head is

H  H SL  hlosses

Typically, losses (whether frictional or due to pipe fittings) are proportional to Q2, so that the system
characteristic is often quadratic.

H  H SL  KQ 2

Where K is a constant

The static lift is often decomposed into the rise from the reservoir or sump to the level of the pump
(the suction head, Hs) and that between the pump and the delivery point, Hd.

The suction head is limited by the maximum suction height (approximately 10 m corresponding to 1
atmosphere).

FINDING THE DUTY POINT

The pump operates at a duty point where the head supplied by the pump precisely matches the
head requirements of the system at the same discharge i.e. where the pump and system H-Q curves

8
intersect. In practice, the duty point or operating point, shows the maximum discharge possible in
the system.

Example

A village requires a water supply; the most convenient source is a borehole which is 4,000 m away
and the water level in the borehole is 40 m below that at which the water needs to discharge. A
pipeline of 200 mm diameter can readily be constructed for which the Darcy f is 0.01. When the
pump is driven at 2,920 RPM the H - Q relationship can be written as

H  AN 2  BNQ  CQ 2

Where H head (in metres)

N speed of rotation (in RPM)
Q = flow in m3/s
A = 2.378 x 10-5
B = 0.069
C = 5000

Determine the quantity of water that the pump can deliver.

PUMP PERFORMANCE EQUATIONS, AFFINITY LAWS AND SPECIFIC SPEED

The pump performance equations for centrifugal pumps are:

 Q   Q 
(i)  3 
 3 
 ND 1  ND  2

 gH   gH 
(ii)  2 2  2 2
 N D 1  N D  2

9
  P   P 
(iii)    
5 

5 
 N D 1  N D  2
3 3

Example

A pump has an impeller diameter of 0.80 m and operates at 1200rpm. If the speed is increased to
1500 rpm, what impeller diameter would be needed to keep the power requirement the same? How
would the change in diameter affect the discharge and head produced by the pump?

AFFINITY LAWS

For the same pump (i.e. same diameter) operating at different speeds N1 and N2:

Q2 N 2
(i) 
Q1 N 1

2
H2  N2 
(ii)  
H 1  N 1 

3
P2  N 2 
(iii)  
P1  N 1 

For the same speed but variable diameter

3
Q2  D2 
(i)  
Q1  D1 

2
H 2  D2 
(ii)  
H 1  D1 

5
P2  D2 
(iii)  
P1  D1 

Apart from comparing the performance of two different pumps, the affinity laws are used to
investigate the performance of one pump under two different operating conditions.

SPECIFIC SPEED FOR PUMPS

10
The specific speed (or type number) is used as a basis for comparison of the performance of
different pumps and is defined as the theoretical speed at which a given pump would deliver 1 unit
of flow against 1 unit of head.

For a given pump, the hydraulic scaling laws give

Q Q1 / 3
 C1 or D
ND 3 C11 / 3 N 1 / 3

H H 1/ 2
And  C2 or D  1/ 2
N 2D2 C2 N

Eliminating D and multiplying all terms by the power 3/2 gives specific speed relationship.

3/ 2 3/ 2
 Q1 / 3   H 1/ 2 
 1/ 3 1/ 3    1/ 2 
 C1 N   C2 N 

NQ1 / 2
Therefore,  cons tan t  N s
H 3/ 4

The constant is the specific speed, Ns, when Q and H in specified units are numerically equal to 1.

NQ1 / 2
Ns 
H 3/ 4

With the commonest definition, N is in rpm, Q in m3/s and H in m. Because of the omission of g the
definition of Ns depends on the units of Q and H. A less common quantity is the dimensionless
specific speed Kn given by:

Q1 / 2 N
Kn 
gH 3 / 4
Approximate ranges of Ns are:

Type Ns

Radial (centrifugal) 10 – 70 Large head

Mixed flow 70 – 170

Axial flow > 110 Small head

Example

11
A pump is needed to operate at 3000 rpm with a head of 6 m and a discharge of 0.2 m 3/s. By
calculating the specific speed, determine what sort of pump is required. If a pump is needed at
another site where the head is 60 m, what sort of pump is required?

FINDING THE DUTY POINT AT A NEW PUMP SPEED

The following procedure should be followed:

(i) Draw the characteristic curves for the known operating conditions
(ii) Scale each (Q,H) pair on the original characteristic at speed N1 to get the new
characteristic at speed N2 i.e.

2
N  N 
Q2   2 Q1 ; H 2   2  H 1
 N1   N1 

Where this scaled characteristic intercepts the system curve gives the new duty point.

Point B is also called the corresponding point to A and the pump efficiency is the same for the two
points.

FINDING THE PUMP SPEED FOR A GIVEN DUTY POINT

To find the pump speed for a given discharge or head plot a hydraulic-scaling curve back from the
required duty point (Q2, H2) on the system curve, at unknown speed N2:
2
H Q 
 
H 2  Q2 

12
Note: The hydraulic scaling curve is not the same as the system curve.

Where the hydraulic scaling curve cuts the original pump characteristic gives a scaled duty point (Q1,
H1) and the ratio of pump speeds can be obtained from either the ratio of discharges or the ratio of
heads:
2
N 2 Q2  N2  H
 or    2
N 1 Q1  N1  H1

PUMPS IN SERIES AND IN PARALLEL

PUMPS IN SERIES

If one pump cannot produce sufficient head then two or more pumps may be used in series. To
connect two pumps in series means that the discharge from the first pump is piped into the inlet
side of the second pump. In this type of arrangement all the flow successively passes from one pump
to the next with each pump adding more energy to the water. For the great majority of pipelines this
would not be considered a good arrangement, it would be better to specify a larger pump. However,
such an arrangement is frequently used in deep boreholes.

13
Thus, 𝑄1 = 𝑄2 = 𝑄𝑡𝑜𝑡𝑎𝑙 i.e. same discharge

H  H1  H 2 i.e. add the heads

PUMPS IN PARALLEL

If a pump cannot produce sufficient flow then two or more pumps are used in parallel. This
arrangement is very common in water supply and sewerage pumping stations. In these designs it is
usual to have several pumps running in parallel, with the pumps cutting in sequentially as the
demand or flow increases.

14
Thus, H1  H 2  H i.e. same head

Q  Q1  Q2 i.e. add the discharges

Advantages of pumps in parallel are:

(i) High capacity: permits a large total discharge

(ii) Flexibility: pumps can be brought in and out of service if the required discharge varies.
(iii) Redundancy: pumping can continue if one is not operating due to failure or planned
maintenance.

Example

A rotodynamic pump, having the characteristics tabulated below, delivers water from a river at
elevation 112 m to a reservoir with a water level of 145 m, through a 300 mm diameter cast-iron
pipe. The frictional head loss in the pipeline is given by hf = 650Q2, where hf is the head loss in m and
Q is the discharge in m3/s. Minor head losses from valves and fittings amount to 60Q2 in the same
units.

Q (m3/s) 0 0.05 0.10 0.15 0.20

H (m) 60 58 52 41 25
Efficiency (%) --- 44 65 64 48

(a) Calculate the discharge and head in the pipeline (at the duty point).

If the discharge is to be increased by the installation of a second identical pump:

(b) determine the unregulated discharge and head produced by connecting the pump:
(i) in parallel;
(ii) in series;
(c) determine the power demand at the duty point in the case of parallel operation.

CAVITATION

When the suction pressure at the pump inlet is too low, vapour bubbles form in the fluid in a
manner similar to boiling. Thus, cavitation is the formation, growth and rapid collapse of vapour
bubbles in flowing liquids.

NET POSITIVE SUCTION HEAD REQUIRED

Pump manufacturers test each pump design to determine the level of suction pressure required to
avoid cavitation, reporting the results as the net positive suction head required, NPSHR, for the pump
at each operating condition of Q and H on the pump. It is the responsibility of the pump system
designer to ensure that the available net positive suction head, NPSHA, is significantly above NPSHR.

Thus, NPSHA > NPSHR

COMPUTING NPSHA

15
The value of NPSHA is dependent on the vapour pressure of the fluid being pumped, energy losses in
the suction piping, the location of the liquid reservoir, and the pressure applied to the fluid in the
reservoir. This can be expressed as:

Patm P
NPSH A   Zs  hf  v
g g

Where

Zs = elevation difference from the level of fluid in the reservoir to the centreline of the pump suction
inlet. If the pump is below the reservoir, Zs is positive, if above it is negative.

hf = headloss in the suction piping due to friction and minor losses

Pv = vapour pressure head of the liquid at the pumping temperature

Thoma cavitation parameter is defined by

𝑁𝑃𝑆𝐻𝐴
𝜎=
𝐻

Critical cavitation number is given by

𝑁𝑃𝑆𝐻𝑅
𝜎𝑐 =
𝐻

At cavitation conditions

 c

Example

A pump can deliver a discharge of 0.1 m3/s to a head of 30 m. The critical cavitation number  c for
the pump is found to be 0.12. The pump is to be installed at a location where the barometric
pressure is 96.0 kPa (abs) and the vapour pressure is 3.0 kPa (abs). Assuming an intake pipe friction
of 0.3 m, determine the minimum value of NPSH. What would be the maximum allowable elevation
above the sump water surface at which the pump can be located?

16