11

TURBINES AND PUMPS

11.1 Describe the difference between (a) a turbine and a pump; (b) an impulse turbine and a reaction
turbine; (c) a reciprocating pump and a rotodynamic pump.

(a) The obvious difference is that a turbine uses a flow of water (Q) with a head (H) to generate electricity,
whereas a pump (which may be powered by electricity) is used to raise a quantity of water (Q) over some height
(H). Thus a pump can be said to be a turbine operating in reverse.
(b) A pure impulse turbine uses the impact (or momentum) of a water jet to turn a runner. The force generated
and the speed of rotation of the runner depend entirely upon the mass flow rate of the jet (ρQ) and its velocity
(V). The impact takes place in the open atmosphere. The pressure of the water is unimportant.
On the other hand, a pure reaction turbine uses the pressure (P) of the water (or the pressure difference across a
runner) to turn the runner. In this case the velocity of the water would be unimportant.
Many turbines, such as the common Francis Turbine, use a combination of the water velocity and pressure to
drive the runner.
(c) Basically a reciprocating pump consists of a piston that moves back and forth in a cylinder, like a bicycle
pump. If the stroke of the piston is the same each time, then it displaces the same amount of fluid each time
regardless of the head pumped against. The output is not continuous, there being no discharge when the piston is
travelling backwards to reposition itself for the next stroke. Reciprocating pumps are rather limited in usefulness
and range of application.
The essential component of a rotodynamic pump is a rotating element (impeller) which is in continuous
motion, so providing a continuous discharge. The discharge from a rotodynamic pump decreases with increasing
lift (the head pumped against). Reciprocating pumps are available that can deliver a wide range of discharge
against a wide range of head, so making them very versatile and suitable for a wide range of applications.

11.2 Define (a) synchronous speed; (b) the overall efficiency of a turbine; (c) the overall efficiency of a
pump; (d) relative velocity.

(a) In the UK a turbine must produce electricity at a constant frequency (fre) of 50 Hz. If the turbine has a
number of poles, p, then the speed (N) at which the turbine must rotate to deliver electricity at the correct
frequency is called the synchronous speed and is given by:
N = (60 × fre)/p
The same equation can be used to calculate the speed (N) at which a motor with p poles will rotate when
supplied by electricity of a certain frequecy (fre).
(b) The overall efficiency of a turbine, εT, is the ratio of its output power (Pow) to the power of the stream of
water supplying it (ρgQH). Thus:
εT = Pow /ρgQH
where H is the total head available, that is the head difference between the inlet of the turbine and the tailwater
level.
(c) The overall efficiency of a pump εP, is the ratio of its fluid output power (ρgQH) to the mechanical power
input to the machine (Pow). Thus:
εP = ρgQH /Pow
where H is the actual total head difference between the inlet and the outlet of the pump.
(d) The relative velocity of two moving bodies can be obtained by considering one of the bodies to be stationary
and measuring the velocity of the other body relative to it. This is sometimes called the ‘bringing to rest
technique’. For example, if two cars are travelling in the same direction at 30 mph and 50 mph then the relative
velocity is 20 mph. This is the same as saying the 30 mph body is stationary and the other body is moving away
from it at 20 mph. However, if the two cars are travelling in opposite directions with velocities of 30 mph and 50
mph the relative velocity is 80 mph. This is the same as saying the body travelling at 30 mph is stationary and the
other is travelling at 80 mph. This is, of course, why head on collisions between vehicles travelling in opposite
directions are so much worse than bumping into a vehicle travelling in the same direction. Relative velocity is
explained more fully in Box 11.2 in the text.

11.3 A 15 mm diameter jet of water hits the centre of a stationary hemispherical cup (as in Fig 11.2d)
with a velocity of 6.0 m/s, divides, and flows smoothly over the cup without loss of velocity. The deflection

3rd ed, Understanding Hydraulics 73 © Les Hamill 2001, 2011

angle is 180 degrees. (a) What is the force exerted by the jet on the cup? (b) If the velocity of the water
leaving the cup (V2) is 0.97 of the initial velocity, what is the force exerted on the cup now?

(a) This is the situation to which equation 11.9 is applicable: FRX = 2ρQV1 where Q = AV1 so:
FRX = 2ρAV12
FRX = 2 × 1000 × (π × 0.0152/4) × 6.02
FRX = 12.72 N
(b) For this part of the question it is best to go back to equation 11.6:
−FRX = ρQ(V2X − V1X)
The jet initially travels along the x axis so V1X = V1 = 6.0 m/s and Q = AV1. The exit velocity V2 = 0.97V1 and
hence with a deflection angle of 180°, V2X = 0.97V1 cos 180°. Thus:
−FRX = ρAV1([0.97V1 cos 180°] − V1)
−FRX = ρAV12([0.97 × −1.00] −1)
−FRX = 1000 × (π × 0.0152/4) × 6.02 (−1.97)
FRX = 12.53 N

11.4 A jet of water flows tangentially onto a single stationary vane (as in Fig 11.2e) with a velocity V1 of
16.0 m/s. The jet is turned through 150 degrees and has an exit velocity V2 = 0.85 V1. The volumetric flow
rate of the jet is 0.04 m3/s. What is the magnitude and direction of the resultant force exerted on the vane?

With reference to equation 11.6, the component of the resultant force in the x direction is:
−FRX = ρQ(V2X − V1X)
Since the jet initially travels along the x axis, V1X = V1 = 16.0 m/s and V2X = 0.85V1 cos 150° so:
−FRX = ρQ (0.85V1 cos 150° − V1)
−FRX = 1000 × 0.04 ([0.85 × 16.0 × −0.866] − 16.0)
−FRX = 40 (−11.778 − 16.0)
FRX = 1111.12 N
With reference to equation 11.7, the component of the resultant force in the y direction is:
FRY = ρQ(V2Y − V1Y) where V1Y = 0 (since the jet initially travels along the x axis, it has no component in the y
direction) and V2Y = 0.85V1 sin 150°. Thus:
FRY = ρQ(0.85V1 sin 150° − 0)
FRY = 1000 × 0.04 ([0.85 × 16.0 × 0.50] − 0)
FRY = 40 (6.80)
FRY = 272.00 N
The magnitude of the resultant force can be calculated from the two component forces:
FR = (FRX2 + FRY2)1/2
FR = (1111.122 + 272.002)1/2
FR = 1144 N
The direction of the resultant force to the horizontal is:
φ = tan–1 (FRY/FRX)
φ = tan–1 (272.00 /1111.12)
= 13.8°

11.5 (a) A horizontal jet of water hits a flat plate angled at 40 degrees to the jet as in Fig 11.2a and is
deflected smoothly without loss of velocity. The diameter of the jet is 20 mm and V1 is 7.32 m/s. Calculate
the magnitude of the resultant force. (b) If everything is as in part (a) except that the deflection angle has
increased to 60 degrees, what is the magnitude of the resultant force now?

(a) −FRX = ρQ (V2X − V1X)

The jet initially travels along the x axis so V1X = V1 = 7.32 m/s and Q = AV1. The jet is deflected through
40 degrees by the plate but its velocity is undiminished so V1 = V2 and V2X = V1 cos 40° (see Fig 11.2b).
−FRX = ρAV1 (V2 cos 40° − V1)
−FRX = 1000 × (π × 0.022/4) × 7.32 (7.32 cos 40° − 7.32)
−FRX = 2.300 (5.607 − 7.320)
+FRX = 3.94 N
Now considering the y direction.
FRY = ρAV1 (V2Y − V1Y)
The initial jet along the x axis has a velocity V1 with no component in the y direction so V1Y = 0. With no loss of
velocity at impact V1 = V2 so V2Y = V1 sin 40°. Thus:

3rd ed, Understanding Hydraulics 74 © Les Hamill 2001, 2011

FRY = 1000 × (π × 0.022/4) × 7.32 (7.32 sin 40° − 0)
FRY = 2.300 (4.705)
FRY = 10.82 N
The resultant force is FR = (FRX2 + FRY2)1/2
FR = (3.942 + 10.822)1/2
FR = 11.52 N
(b) The calculations are as above but with a deflection angle of 60 degrees, hence:
−FRX = ρAV1 (V2 cos 60° − V1)
−FRX = 2.300 (7.32 cos 60° − 7.32)
−FRX = 2.300 (3.660 − 7.320)
+FRX = 8.42 N
FRY = ρAV1 (V2 sin 60° − V1Y)
FRY = 2.300 (7.32 sin 60° − 0)
FRY = 2.300 (6.339)
FRY = 14.58 N
FR = (8.422 + 14.582)1/2
FR = 16.84 N

11.6 A jet of water with a velocity of 25.0 m/s and a diameter of 200 mm slides tangentially onto a
stationary curved vane as in Fig 11.6a and is turned through an angle of 165 degrees. The velocity of the
water leaving the vane is 90 percent of the original jet velocity. (a) Calculate the magnitude and direction
of the resultant force. (b) If everything is as above except that the vane is moving at 12.0 m/s away from
the jet and the relative velocity at exit is 90 percent of the initial relative velocity, what is the new
magnitude and direction of the resultant? (c) If the conditions are as in part (b) except that the vane is
now mounted on a runner, what is the magnitude and direction of the resultant?

(a) If the vane configuration is as in Fig 11.6a but the vane is stationary.
−FRX = ρQ (V2X − V1X)
The jet enters along the x axis so V1X = V1 = 25.0 m/s and Q = AV1. Now V2 = 0.90V1 so with a deflection angle
of 165°, V2X = 0.90V1 cos 165°.
−FRX = ρAV1 (0.90V1 cos 165° − V1)
−FRX = 1000 × (π × 0.22/4) × 25.0 (0.90 × 25.0 × cos 165° − 25.0)
−FRX = 785.398 (−21.733 − 25.0)
+FRX = 36 704 N
Now considering the y direction.
FRY = ρQ (V2Y − V1Y)
The water enters along the x axis so V1Y = 0. Again V2 = 0.90V1 and V2Y = 0.90V1 sin 165°.
FRY = ρAV1 (0.90V1 sin 165° − 0)
FRY = 785.398 (0.90 × 25.0 × sin 165°)
FRY = 785.398 (5.823)
FRY = 4573 N
FR = (FRX2 + F RY2)1/2 = (36 7042 + 45732)1/2 = 36 988 N
φ = tan–1 (FRY/FRX) = tan–1 (4573/36 704) = 7.1° to the horizontal
(b) Starting this time with the general equation in Box 11.5 for a single vane not mounted on a runner:
−FRX = ρA(V1 − U)[η(V1 − U) cos θ − (V1 − U)]
Now θ = 165°, η = 0.90, V1 = 25.0 m/s and U = 12.0 m/s.
−FRX = 1000 × (π × 0.22/4) × (25.0 − 12.0) [0.90 (25.0 − 12.0) cos 165° − (25.0 − 12.0)]
−FRX = 408.407 [−11.301 − 13.0]
+FRX = 9925 N
FRY = ρA(V1 − U)[η(V1 − U) sin θ − 0]
Again θ = 165°, η = 0.90, V1 = 25.0 m/s and U = 12.0 m/s so (V1 − U) = 13.0 m/s
FRY = 408.407 [0.90 × 13.0 × sin 165°]
FRY = 1237 N
FR = (99252 + 12372)1/2 = 10 002 N
φ = tan–1 (1237/9925) = 7.1° to the horizontal
(c) Now starting with the general equation in Box 11.5 for a vane that is mounted on a runner:
−FRX = ρAV1 [η(V1 − U) cos θ − (V1 − U)]
Most of the values are exactly as in part (b) so:
−FRX = 1000 × (π × 0.22/4) × 25.0 [−11.301 − 13.0]

3rd ed, Understanding Hydraulics 75 © Les Hamill 2001, 2011

+FRX = 785.398 [24.301]
+FRX = 19 086 N
FRY = ρAV1 [η(V1 − U) sin θ − 0]
FRY = 785.398 [0.90 × 13.0 × sin 165°]
FRY = 2378 N
FR = (10 0862 + 23782)1/2 = 19 234 N
φ = tan–1 (2378/19 086) = 7.1° to the horizontal

11.7 Describe and illustrate what the following look like, showing clearly the flow path of the water
through the machine (a) a Pelton wheel; (b) a Francis turbine; (c) a centrifugal pump; (d) an axial flow
pump. (e) For the above, describe the most important performance parameters and illustrate how they
vary for a particular type of machine.

(a) Basically a Pelton wheel is a wheel or runner that has a series of vanes or buckets attached to its perimeter
(Figs 11.3 and 11.6b). A high velocity jet of water from a nozzle (Fig11.7), or a number of nozzles, hits the
buckets causing the wheel to rotate. The buckets are usually designed so that the jet hits a central splitter fin,
divides and flows around the two ‘sides’ of the bucket before being discharged to the tailrace (Fig 11.3). The
deflection angle of the buckets is usually around 165 degrees. After impact the discharged water may be directed
to the side and backwards relative to the moving bucket (Fig 11.4). The Pelton wheel is rather unique compared
to other turbines in that the impact takes place in the atmosphere, or at least in a chamber which contains air at
atmospheric pressure (Fig 11.8). Most other turbines have a watertight casing within which the water is confined
at high pressure. A diagrammatic representation of the water flow path is given in Fig Q11.7.
(b) A Francis turbine is shown in Fig 11.10. It has a watertight, pressurised volute chamber shaped like a snail’s
shell. This is designed to increase the velocity of the water. The Francis turbine uses both the velocity and the
pressure of the water to turn the central runner. When water enters horizontally from a supply pipe, it spirals
around the volute casing in a circle of decreasing radius, flows through adjustable guide blades, and then passes
through slots in the runner (exerting a force on the vanes of the runner as it does so causing it to rotate) before
dropping vertically down the draft tube. It then turns through 90° before being discharged to the tailwater (Fig
Q11.7).
(c) The flow path through a centrifugal pump is basically the reverse of that through a Francis turbine. The water
may typically be sucked vertically up a suction pipe to the pump. The water enters the pump at the centre of the
rotating impeller, which is located inside the watertight, pressurised snail-shell like volute chamber (Fig 11.15).
The water then turns through 90° and flows horizontally over the impeller, which is driven by a motor. The
raised vanes of the impeller then throw the water outwards against the walls of the volute chamber, increasing its
pressure and generating the ‘lift’ (Fig Q11.7). The water is then piped away, initially horizontally (but the
discharge pipe can subsequently bend in any required direction). Of course, the pump can be mounted
differently, the main point being that the inlet and outlet direction are always at 90° to each other. For example,
if the impeller is in the vertical plane, the water may enter horizontally and leave either horizontally along a
diffrent axis or vertically.

Fig Q11.7 Diagrammatic representation of flow paths.

3rd ed, Understanding Hydraulics 76 © Les Hamill 2001, 2011

(d) An axial flow pump has the simplest flow path, being nearly a straight line. If the pump is mounted with the
impeller horizontal as in Fig 11.16, then the flow path is essentially straight up. The fins and guide blades before
and after the impeller are designed to prevent rotation of the liquid within the casing, and ensure that the
direction of flow is uniformly in an upward direction without too much of a spiralling motion (Fig Q11.7).
(e) At the design stage the diameter of the runner may also be an important variable, but for a particular turbine
of fixed diameter the most important performance parameters are: inlet head and flow rate, power output and
efficiency (see equations 11.14 to 11.16). These give rise to the power-speed curves, efficiency- speed curves
and efficiency-part load curves of Fig 11.14. See also Question 11.8(a).
For a pump the diameter of the impeller may be important initially, but once this is fixed the most important
parameters are the power requirement, lift (head) achieved, output discharge and efficiency (see equations 11.14
and 11.15 and 11.17 to 11.21). These four parameters are usually combined on a single graph as in Figs 11.17
and 11.18. See also Question 11.8(b).

11.8 Define what is meant by the specific speed of (a) a turbine and (b) a pump. (c) A turbine is required
to generate 5.5 MW of electricity from the regulating releases from Kielder Water reservoir. The nominal
head of water available is 47.35 m. If the turbine has a rotational speed of 500 rpm, by calculating the
specific speed determine what sort of turbine is required. (d) A second turbine is required at Kielder to
utilise the compensation flow to generate 500 kW from the 47.35 m head when running at 1000 rpm.
What sort of turbine should this be? (e) If the turbine in part (c) has a water requirement of 14.1 m3/s,
what would be the required flow rate if the turbine was run at 1000 rpm, and what would be the new
power output?

(a) The specific speed of a turbine (NS) is the speed (N, in rpm) needed to develop 1 kW of power (Pow) when
operating with a head (H) of 1 m. Under these conditions NS = N in equation 11.16 below:
NS = N Pow1/2/H 5/4 (11.16)
The specific speed can be used to select the type of turbine for a particular location, see Chapter 11.4 and below.
(b) The specific speed of a pump (NS) is the speed (N, in rpm) needed to discharge 1 m3/s (Q) against a head (H)
of 1 m. Under these conditions NS = N in equation 11.21 below:
NS = N Q1/2/H 3/4 (11.21)
As with a turbine, the specific speed can be used to select the type of pump for a particular location, see Chapter
11.6.2.
(c) From equation 11.16 above, NS = N Pow1/2/H 5/4 where N = 500 rpm, Pow = 5500 kW, and H = 47.35 m.
NS = 500 × 55001/2/47.355/4
NS = 37 081/124.2
NS = 299
From Chapter 11.4 this corresponds to a Kaplan turbine (low head, large discharge).
(d) Applying equation 11.16 again with N = 1000 rpm, Pow = 500 kW and H = 47.35 m.
NS = 1000 × 5001/2 /47.355/4
NS = 22 361/124.2
NS = 180
From Chapter 11.4 this corresponds to a Francis turbine (moderate head, moderate discharge).
(e) For a turbine of constant diameter the appropriate affinity law (on page 403 of the text) is:
(QA/NA) = (QB/NB) where A represents the initial condition and B the changed condition. Thus:
QB = QANB /NA where QA = 14.1 m3/s, NA = 500 rpm and NB = 1000 rpm.
QB = 14.1 × 1000/500
QB = 28.2 m3/s
The new power output can be obtained using the appropriate affinity equation on page 403 of the text.
(PowA/NA3) = (PowB/NB3) or PowB = PowA NB3/NA3
PowB = 5.5 × 106 × 10003/125 × 106
PowB = 44 × 106 W or 44 MW

11.9 Water is to be pumped for several hours per day from a sump into a rising main of either 150 mm or
200 mm diameter. The static lift is 6.0 m and the pipe friction factor (λ) for both pipes is 0.02. The
effective length of the pipe is 28 m. Two types of pump, A and B are under consideration. Both operate at
maximum efficiency in the range under consideration. The head-discharge characterisic of each pump is
as follows:

Pump A
Hm 9.80 8.68 8.00 7.48 6.88 6.10 4.87 3.45 1.40
Q m3/s 0 0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08

3rd ed, Understanding Hydraulics 77 © Les Hamill 2001, 2011

Pump B
Hm ---- ---- 9.65 7.63 5.88 4.00 1.75
Q m3/s ---- ---- 0.02 0.03 0.04 0.05 0.06

(a) Plot the rising main system curve for pipes of 150 mm and 200 mm diameter from Q = 0 to
Q = 0.10 m3/s.
(b) Superimpose the two pump curves on the two system curves. What is the discharge obtained from
each pump when connected to each of the rising mains?
(c) A discharge of about 0.05 m3/s is required. Which of pumps A and B acting either alone or with two
identical pumps in parallel or series can deliver this quantity of water with the 6.0 m static lift and either
of the rising main diameters?
(d) Select the rising main diameter and the pump or combination of pumps you think most suitable.
(d) For whichever combination of pumps and rising main you have selected, calculate the velocity in the
delivery pipe. Is this satisfactory?
(e) Is surge likely to be a serious problem with this installation?
(f) Sketch a possible layout for the pump(s) and the sump.

(a) Equation 11.23 is the Darcy equation for friction loss: hF = λLQ 2/12.1D5 where λ = 0.02 and L = 28 m.
hF = 0.02 × 28 × Q 2/12.1D5
hF = 0.04628 Q 2/12.1D5
With D = 0.150 m, hF = 609.462 Q 2
With D = 0.200 m, hF = 144.625 Q 2
These two equation can be used to calculate the head loss, hF , in the two alternative rising mains. The values
obtained are shown below.

---------------------------------------------------------------------------------------------------------------------------------
Q (m3/s) 0 0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08 0.10
---------------------------------------------------------------------------------------------------------------------------------
150 mm, hF (m) 0 0.061 0.244 0.549 0.975 1.524 2.194 2.986 3.901 6.095
200 mm, hF (m) 0 0.014 0.058 0.130 0.231 0.362 0.521 0.709 0.926 1.446
---------------------------------------------------------------------------------------------------------------------------------
The static lift of 6.0 m must be added to the head losses above. The resulting values are plotted as the system
curves in Fig Q11.9.

(b) Superimposing the curves for pumps A and B over the two system curves and finding the points of
intersection gives:
Pump A with 150 mm rising main, Q = 0.040 m3/s Pump A with 200 mm rising main, Q = 0.048 m3/s
3
Pump B with 150 mm rising main, Q = 0.035 m /s Pump B with 200 mm rising main, Q = 0.038 m3/s
(c) For two identical pumps in parallel the discharge (Q) should be doubled for the same head (H), thus:

A//A
H (m) 9.80 8.68 8.00 7.48 6.88 6.10
2Q (m3/s) 0 0.02 0.04 0.06 0.08 0.10

B//B
H (m) ----- ----- 9.65 7.63 5.88 4.00
2Q (m3/s) ----- ----- 0.04 0.06 0.08 0.10

For two identical pumps in series the head (H) is doubled for the same discharge (Q), thus:

A+A
Q (m3/s) 0 0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08
2H (m) 19.60 17.36 16.00 14.96 13.76 12.20 9.74 6.90 2.80

B+B
Q (m3/s) 0 0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08
2H (m) ----- ----- 19.30 15.26 11.76 8.00 3.50 ----- -----

These four curves (A//A, B//B, A+A and B+B) can now be plotted over the two system curves and the points of
intersection obtained. This gives:
A//A with 150 mm rising main, Q = 0.053 m3/s A//A with 200 mm rising main, Q = 0.081 m3/s
3
B//B with 150 mm rising main, Q = 0.057 m /s B//B with 200 mm rising main, Q = 0.070 m3/s

3rd ed, Understanding Hydraulics 78 © Les Hamill 2001, 2011

A+A with 150 mm rising main, Q = 0.065 m3/s A+A with 200 mm rising main, Q = 0.071 m3/s
B+B with 150 mm rising main, Q = 0.051 m3/s B+B with 200 mm rising main, Q = 0.053 m3/s

Fig Q11.9 Pump and system curves

As to which pump acting alone or in combination with an identical pump can deliver 0.05 m3/s against a static
head (lift) of 6.0 m, from the values above it is apparent that:
Pump A by itself with a 200 mm rising main gives 0.048 m3/s, which is very close to 0.050 m3/s.
A//A with a 150 mm rising main gives 0.053 m3/s.
B//B with 150 mm rising main gives 0.057 m3/s.
B+B with 150 mm rising main gives 0.051 m3/s.
B+B with 200 mm rising main gives 0.053 m3/s.
The other alternatives deliver either a discharge which is too small or considerably larger than that required.
(d) Water is to be pumped for several hours per day through a pipeline only 28 m long. The pipeline is short, so
the capital cost of using a 200 mm rising main will probably not be much greater than that of a 150 mm main,
but the head loss in the 200 mm pipe will be smaller. This means that the running cost (the cost of pumping
against the friction head loss) will be less, and the resultant saving may be significant when pumping for several
hours per day, every day, over a long period of time. Therefore, it is logical to use a 200 mm rising main. With
regard to which pump to connect to a 200 mm rising main, the cheapest alternative is to use pump A by itself.
Why buy two pumps when one will do? Thus the best, most economical option is pump A by itself delivering

3rd ed, Understanding Hydraulics 79 © Les Hamill 2001, 2011

through a 200 mm rising main. Some consideration could be given to a standbye pump, depending upon whether
or not a pump failure would have significant consequences.
(e) VD = Q /A where VD is the mean velocity in the delivery pipe (rising main), Q is the discharge from the pump
and A is the cross sectional area of the rising main. If a 200 mm rising main is used with pump A which delivers
0.048 m3/s, then VD = (0.048 × 4 / π × 0.2002) = 1.53 m/s. Generally VD is acceptable if it is between about 1.2
and 3.0 m/s, so this is OK.
(f) A very crude test for surge is to calculate the value of 2L/c where L is the pipeline length (28 m) and c the
acoustic velocity, say about 1000 m/s in this case. Thus 2L/c = (2 × 28/1000) = 0.056 s. Since this is less than 2 s
there should not be a significant problem, the pipeline being very short.
(g) See Fig 11.22 in the text.

11.10 (a) Describe what is meant by cavitation, net positive suction head, surge and waterhammer.
(b) What are the symptoms of cavitation, surge and waterhammer? If you wanted to find out if these
phenomena were present in a hydraulic system, what would you look for? (c) If these three phenomena
are causing problems, what steps can be taken to minimise their effect?

(a) Cavitation is the phenomenon where small bubbles of vapour implode violently. The bubbles form when a
reduction in absolute pressure occurs, followed by an increase in pressure.
The net positive suction head (NPSH) is the head (or pressure) required to prevent cavitation as a liquid enters
the impeller of a pump. This is defined as:
NPSH (m) = HATM − HVAP − HS − hFS (11.24)
where HATM is atmospheric pressure (about 10 m of water), HVAP is the vapour pressure of the liquid, HS is the
static suction lift from the water level in the sump to the centreline of the pump (Fig 11.19) and hFS is the friction
loss in the suction pipe. It is called the ‘net positive’ suction head because for equation 11.24 to yield a positive
answer HATM must be greater than (HVAP + HS + hFS ). If it is not, then a negative suction head exists (that is a
pressure below atmospheric). A manufacturer will quote the limiting value of NPSH for a particular pump, after
which cavitation may be expected.
Surge occurs in a pressurised pipeline when the velocity of flow is changed, or when valves are opened or
closed. For example, if a valve is closed the momentum of the water downstream of the valve will cause it to
continue flowing through the pipe. This causes an empty space after the valve, which will have a pressure below
the static head of the liquid in the pipeline. Consequently liquid from the high pressure part of the pipeline surges
into this space causing a pressure in excess of the static pressure. The pressure then fluctuates around the static
pressure until it stabilises.
Waterhammer is the pressure caused by closing a valve. The moving liquid upstream of the valve has
momentum, so if the valve is suddenly shut the water will collide with it, generating quite large forces in the
process (in this case Newton’s Second Law could be stated as force = mass/second x deceleration). The pressure
of the liquid in the pipe increases as it collides with the closed valve. In some cases the pressure may be
sufficient to burst the pipe. The pressure tends to fluctuate sinusoidally and gradually attenuate.
(b) The symptoms of cavitation are described in Box 11.6, and are basically the formation of bubbles, pitting of
a solid surface, instability of the flow, noise and vibration. The symptoms of surge and waterhammer are a
rapidly fluctuating liquid pressure about the mean, perhaps accompanied by clonking noises and vibration or
jerking of the pipe.
(c) Cavitation of a pump can be avoided to some extent by designing the sump and suction pipe to minimise
losses, such as by keeping the suction pipe straight as in Fig 11.22, using a generously sized suction pipe (low
velocity, high pressure), and ensuring the strainer (filter) at the foot of the suction pipe does not become blocked.
The inlet to the impeller can be made large (again, to keep the pressure high) and the impeller can be made from
something like chrome vanadium steel which is relatively resistant to pitting.
Surge can be avoided by using some form of damping, such as pnuematic loading by means of a compression
tank containing air or a non-condensable gas under pressure, injecting liquid from another source such as a surge
shaft and by prolonging the period over which a pump stops or starts or a valve opens and closes. The same
methods can be used to control waterhammer, namely shutting valves slowly or (as with the hydro-electric
scheme in Fig 11.1) using a surge shaft or chamber.
It is obviously better to design initially so as to try to avoid these problems than it is to try to remedy them by
altering the completed works. However, if something has been completed and suffers from cavitation, surge or
waterhammer some of the above measures are still practicable, and it may be possible to add them at a later date.

11.11 In a proposed pipeline a steady flow of water (K = 2100 × 106 N/m2) will exist at a velocity of
2.5 m/s. The intention is to use a 0.3 m internal diameter PVC pipe (E = 26 × 108 N/m2, wall thickness
d = 20 mm). (a) Calculate the wave velocity cP in the pipe assuming c = 1450 m/s. (b) Calculate the
additional surge head ∆H that will be experienced by an instantaneous valve closure. (c) If the maximum

3rd ed, Understanding Hydraulics 80 © Les Hamill 2001, 2011

static head on the pipe is 9 m and the the maximum working pressure of PVC pipes is 150 m of water,
determine whether or not PVC is suitable for this application. (d) If the pipeline is to be 100 m long, draw
the head-time graph at the valve.

(a) cP = c / [1 + (KD/Ed)]1/2 = 1450 / [1 + (2100 × 106 × 0.3/26 × 108 × 0.02)]1/2 = 400 m/s
(b) ∆H = cPV0/g = 400 × 2.5/9.81 = 101.9 m
(c) Adding the static head gives a maximum head of 101.9 + 9 = 110.9 m < 150 m.
A PVC pipe is satisfactory.
(d) 2L/cP = 2 × 100/400 = 0.5 s. This is the time interval used on the graph in Fig 11.27.

3rd ed, Understanding Hydraulics 81 © Les Hamill 2001, 2011