10th International Conference on The Role Of Engineering Towards A Better Environment

Intelligent Environmental Engineering: From Vision to Action

15-17 December, 2014 Alexandria, Egypt
Keyword Index ‫ ׀‬Author Index ‫ ׀‬Title Index ‫ ׀‬About RETBE’14

ELUCIDATION OF DIFFERENT PROTECTION METHODS AGAINST

WATER HAMMER IN PIPES NETWORK OF ASSIUT CITY
M. S. DARWEESH1, N. A. ALI2 AND G ABOZEID3.
1
Assistant Professor, Civil Eng. Dept., Assuit University, 71516 Assuit, Egypt.
2, 3
Professors of hydraulics & water resources, Civil Eng. Dept., Assuit Univ., 71516 Assuit.

ABSTRACT: This paper presents the results of using different protection devices on water
hammer in water supply systems. Flow of Assiut city network is studied under steady case,
without protection, and with different protection device(s) such as non-return valve (NRV), open
surge tank, air chamber and pressure relief valve (PRV). The results are performed for different
scenarios of pumps shutting off. Water Hammer and Mass Oscillation (WHAMO) software is used
in the analysis which uses the implicit finite difference scheme for solving the momentum and
continuity equations at unsteady case. The results showed that using surge tank or air chamber
with NRV protects the pipe network effectively from the damage of water hammer. Also, using
PRV with NRV saves the network from extreme pressures. Although using NRV only doesn’t
have a great effect on the maximum pressure head but, protects the network from the more low
pressures than without protection one. All protection methods safeguard the network from the
extremes of water hammer, but increase its wave period.

INTRODUCTION

A hydraulic transient is a flow condition where the velocity and pressure change rapidly with time.
The occurrence of transients can introduce large pressure forces and rapid fluid accelerations into a
water distribution system. Many researchers studied the Water hammer phenomenon along the last
decades with different viewpoints [1, 4-9, 14, 15 and 19]. Lohrasbi and Attarnejad15 described the physical
phenomenon of water hammer and the mathematical model which provides the basis for design
computations using the method of characteristics and effect of valve opening and closure. It has
been shown that the more rapid of the closure of the valve, the more rapid is the change in
momentum. Abreu et al.3 showed that the transient behavior is governed by the fluid/pipe
characteristics, the inherent boundary characteristic, and one associated characteristic time. Fouzi
and Ali12 studied water hammer in gravity piping system due to sudden closure of valves, using
WHAMO program and the method of characteristics with software AFT Impulse. They showed
that the pressure fluctuations are very dangerous especially in the case of pipes has variable
characteristics (section changes with a divergence, a convergence or a bifurcation). Mohamed17
indicated that the pipe friction factor and the time of valve closing have a significant effect in
pressure transient reduction and the elastic pipe such as PVC are better than the rigid pipes in
pressure damping. Abozeid et al.2 investigated the effect of sudden contractions and enlargement
on transient flow in pipeline system due to pump shut down. They found that the pump shut down
accompanied by increasing of pipe contraction or enlargement ratios increases the fluctuation of
piezometric heads and flow rates than the shut down of pump only. Choon et al.10 investigated the
prevention method by installing bypass pipe with non-return valve from water hammer effect in
single pipeline. From their experiments, they found that, this prevention method is successfully to
reduce the water hammer effect in the pipeline, where the mean pressure is reducing by 33.33%,
and this method is most useful in the household usage as the non-return valve. Friedman et al.13
studied the control of pressure transients by air-vacuum valves. They found that, the installation of
air-vacuum valves throughout the system offers some protection against negative pressures but
they are not effective as surge tanks. Larger pumps have more inertia because they have more

1
Moustafa Samir Darweesh, [email protected], 00201005475466, Fax: 0020882332553
2
Nahat Abdel-lah Ali, [email protected], 00201006201785, Fax: 0020882332553
3
Gamal Abozeid Abdel Raheem, [email protected], 00201063572412, Fax: 0020882332553
rotating mass. Pumps with higher inertias can help to control transients because they continue to
move water through the pump for a longer time as they slowly decelerate (Magzoub and
Kwame16). Niţescu et al.18 studied asymmetrical hydraulic resistance devices, and stated that, these
devices efficiency is proved not only by the harmless pressure in the installation during the water
hammer, but also by the reduced water change flow rate between the chamber and the discharge
duct. Air vessels generally alleviate negative pressures more effectively than other forms of water
hammer protection, and they can maintain a positive pressure in the line at all stages following
pump trip (Stephenson21).
According to the aforementioned studies, water hammer in pipes networks has little attention
from the investigators. However, every water supply network has its own special characteristics
which makes it different from the other networks. Also, due to a lack of field measurements which
are costly, it becomes important to use numerical models to gain an indication about the behavior
of network under transient effect. Present study is performed to investigate the effect of sudden
shut down of pumps on transient pressure heads and flow rates with different protection cases for
Assiut city water supply network.

THEORETICAL CONSIDERATIONS

Governing Equations for Unsteady Flow in Pipelines

The unsteady flow inside the pipeline is described in terms of unsteady mass balance (continuity)
equation and unsteady momentum equation, which define the state of variables of V (velocity) and
P (pressure) given as (Simpson and Wu20);
∂ρ ∂ρ ∂V ρ dA (1)
+V +ρ + =0
∂t ∂x ∂x A dt
∂V ∂V 1 ∂P f V V (2)
+V + − g sin α + = 0
∂t ∂x ρ ∂x 2D
Where x = distance along the pipeline; t = time; V = velocity; P = hydraulic pressure in the pipe; g
= acceleration due to gravity; f = Darcy-Weisbach friction factor; ρ = fluid density; D = pipe
diameter; α = pipe slope angle, and A = cross sectional area of the pipe.
Equation (1) is the continuity equation and takes into account the compressibility of water and the
flexibility of pipe material. Equation (2) is the equation of motion. In Eq. (1), the terms
1  ∂ρ ∂ρ  are replaced by equivalent 1 d ρ where V = dx , d ρ = ρ dP , hence
 + V
ρ  ∂t ∂x  ρ dt dt dt K dt
1 ∂ρ 1 dP , and K is the bulk modulus of the fluid. Also, the fourth term in Eq. (1) can be
=
ρ ∂t K dt
expressed as (1 − ν 2 )
ρ D dP , where ν is the Poison’s ratio of the pipe, e is the pipe wall
eE dt
thickness and E is the Young's modulus of elasticity of the pipe. Substitution by these
abbreviations in Eq. (1), it can be reduced to the following formula;
dP  1 1 −ν 2  D  ∂V (3)
 +    + = 0
dt K  E  e  ∂x
Wave speed can be defined as the time taken by the pressure wave generated by instantaneous
change in velocity to propagate from one point to another in a closed conduit. Wave speed (c) can
be expressed as;
   2  
1 2 1 2 (4)
 1 1−ν  D     Kc 1 D  
c =  (1 ρ )  +  = (K / ρ ) 1 + 
   e    
 K  E    Ee

2
Where: c 1 = (1 − υ 2 ) , and we have  ∂ P ∂P  dP . Substitution by Eq. (4) into Eq.
+ V  =
 ∂ t ∂x  dt
(3) and dividing the result by γ yields;
 ∂H ∂H  c 2 ∂V (5)
 ∂ t + ∂ x V  + g ∂x
= 0

Where H is the piezometric head, i.e. pressure head plus the elevation head. The term ∂H / ∂x is
small compared to ∂H / ∂t and it is often neglected. Thus the simplified form of the continuity
equation in terms of discharge, and the substitution by V = Q A , Eq. (5) becomes;
2
∂H ∂Q c (6)
+ = 0
∂t ∂x gA
By the same way, the momentum equation, i.e. Eq. (2) can be simplified and written in terms
of discharge and piezometric head as follows;
∂H 1 ∂Q fQ Q (7)
+ + = 0
∂x gA ∂ t 2 gDA 2
Various methods of analysis were developed for the problem of transient flow in pipes. They
range from approximate analytical approaches to numerical solutions. These methods such as:
Graphical Method, Method of Characteristics, Wave-Plan Method, Implicit Method and others.

Implicit finite difference solution method

The computer program WHAMO uses the implicit finite-difference technique but converts its
equations to a linear form before it solves the set of equations (Fitzgerald and Van Blaricum11).
The solution space is discretized into the x-t plane, so that at any point on the grid (x, t) there is a
certain H and Q for that point, H (x, t) and Q (x, t) as shown in Fig. (1).

Figure 1. The finite difference grid

The momentum equation and the continuity equation can be represented in a short form by
introducing the following coefficients for the known values in a system;
2 ∆ tc 2j θ , (1 − θ ) (8)
α j = β j = ( H n, j +1 + H n, j ) + α j (Q n , j − Q n , j + 1 )
gA j ∆ x j θ

∆x j , δ = (1 − θ ) ( H (9)
γ j = j n , j − H n , j +1 ) + γ j (Q n , j + Q n , j +1 )
2 gθA j∆ t θ

3
Where θ is a weighing factor included for numerical stability. All parameters for the coefficients
should be known from the properties of the pipe or the values of head and flow at the previous
time step. With the coefficients, the momentum and continuity equations of the jth segment of the
pipe become as given by Batterton7 as follows;
Momentum: −H +H
n, j +1 + γ (Q +Q
n+1, j +1
n+1, j ) =δ
n+1, j +1 j j (10)

Continuity: H n, j +1 + H n+1, j +1 + α j (Qn+1, j +1 − Qn+1, j ) = β j (11)

Now, with equations for the all links and nodes in the system, the initial and boundary
conditions, a matrix of the linear system of equations can be set up to solve for head and flow
everywhere, simultaneously, for the first time step. The process is repeated for the next time step,
and again for the next step until the specified end of the simulation.

CASE STUDY

The analysis of transient flow was performed for Assiut city water supply network. The water is
pumped into the network through nodes 27 and 28 (constant head reservoirs) (Fig. 2). The network
is composed of 29.6 Km of different diameter pipelines with lengths of P1 through P35 and 26
junctions, the different proprieties for each pipeline and the average base demand for the different
nodes are shown in Table (1). Elevations of all the network junctions are assumed to be zero level.
All pipes are High Density Polyethylene (HDPE) and the head loss in each pipe is computed using
Darcy-Weisbach formula. The results are performed for three scenarios; the first, is the normal
operation of pump 36 and failure of pump 37, the second, is the normal operation of pump 37 and
failure of pump 36, and finally failure of pumps 36 and 37 together. For the previous three
scenarios, the pipe network is studied with steady case (pumps 36 and 37 working normally),
without any protection against water hammer, and is studied with an transient protection device(s)
such as non-return valve, open surge tank, air chamber and pressure relief valve.
P27 J20 P26 J 15 P19 J14 P18 J6 P8 J5
J21
P20 P9
P33
P7
P25 J16 J4
P17
J19 P10 ElHelaly plant
P13
P21 J7
J3
R
J 26 P11 PU36 28
P24 P35 P22 J 13 P6
P28
J24 P12
J 17 J2
P16 P5
J8
J 18 P23
J12 P1
P30
P4
P31
P15
Nazlet Abdellah plant
P14 J 25 J1
J 11 J9
P29 P3
J 22 P2 P34 R27
PU37
P32 J 10
J23
Figure 2. Pipes, nodes numbering, pumps, sources and flow directions for normal operation
condition of Assiut pipe network for pumps 36 and 37

4
 Table 1. Properties of the different pipes and junction demands
Average base demands for the different junction nodes
Node number 1 2 3 4 5 6 7 8 9 10 11 12 13
Base demand (Lit./s) 0.0 69 0.0 72 41 45 72 64 32 32 64 98 86
Node number 14 15 16 17 18 19 20 21 22 23 24 25 26
Base demand (Lit./s) 45 19 36 75 110 69 53 90 90 128 0.0 45 0.0
Lengths and diameters of the different pipes
Pipe number Length (m) Diameter (m) Pipe number Length (m) Diameter (m)
P1 1600 800 P19 300 600
P2 300 1000 P20 600 400
P3 600 1000 P21 300 500
P4 900 500 P22 600 400
P5 200 500 P23 600 400
P6 300 500 P24 950 400
P7 1400 500 P25 950 300
P8 1100 800 P26 1200 600
P9 500 800 P27 400 600
P10 800 800 P28 2650 600
P11 150 800 P29 2100 600
P12 850 500 P30 1500 400
P13 1100 500 P31 1600 400
P14 500 1000 P32 1500 800
P15 750 500 P33 700 400
P16 850 500 P34 500 1200
P17 1000 500 P35 150 500
P18 100 800

RESULTS AND DISCUSSIONS

To investigate the water hammer effect on the transient pressure heads at points representing
different places on the network, point J17 represents the middle point of the network and point J2
represents the nearer point from the pumps while point J21 represents the further point from the
pumps. At each point, the transient change of piezometric pressure heads is investigated. The
investigations are based on the comparison between the use of different protection devices and no
protection case with the three studied scenarios of pumps shut down. Steady normal operations are
included for the comparison. Figure 3 shows the changes of pressure head with the time at node
J17 for the three scenarios and the network has no protection devices. It is seen that the shutting
down of pump 36 or 37 only lowering the pressure from +57 m of water to be approximately +5
and +11 m, respectively while it reaches to -35 m of water for shutting down of pumps 36 and 37
together. Whereas shutting down of pumps 36 and 37 together lowers the pressure to be negative
at this point for unprotected case.
The use of different protection devices with these scenarios is studied. Shown in Fig. 4 is a
comparison between the effect of the use of these devices on the pressure head. It is seen that the
pressure is lowering from +57 m (steady case) to -12 m of water for using NRV only or using
NRV with PRV together. For using NRV with open surge tank, the pressure head decreases
gradually without fluctuations to +41 m of water and to +31 m of water for using NRV with air
chamber after 300 seconds. For the nearer points from the pumps, Fig. 5 is drawn to show the
variations of pressure head with the time at node J2 for the studied scenarios and without any

5
protection case. It is shown from the figure that, the shutting down of pump 37 only decreases the
head from +59 m to +31 m, and for the shutting down of pump 36 alone the reduction in the head
reached to +14 m, while for the shutting down of pumps 36 and 37 together the pressure head
reduces to be -6 m of water. For protection cases, the pressure reaches to -3 m of water for using
NRV or NRV with PRV together. For using NRV with open surge tank, the pressure head
regularly decreases without fluctuations to +43 m and to +34 m of water for NRV with air
chamber after 300 seconds as shown in Fig. 6. To show the water hammer effect at the further
point from the pumps, the changes of pressure head with the time at node J21 for the previously
scenarios and the network has no protection devices are shown drawn as in Fig. 7. It is noticeable
that the shutting down of pump 36 or 37 only lowering the pressure head from +57 m of water to
be approximately +7 and +4 m, respectively while shutting down the both pumps together lowers
the pressure head to -43 m. Shutting down of pumps 36 and 37 together lowers the pressure to
negative at this point from +59 m of water to be approximately -43 m of water for unprotected
case, and to -17 m of water when using NRV or with PRV together. By using NRV with open
surge tank, the pressure head gradually decreases without fluctuations to +31 m of water and to
+16 m of water for using NRV with air chamber after 300 seconds as shown in Fig. 8.
Finally, from the results shown in Figs. (3 to 8), it is evident that, the appropriate protection
method for the pipe network is the using of air chamber or open surge tank with non-return valve.
This because, they absorb the water hammer and decrease the pressure heads gradually to reach
the final stable pressure head without any waves and the maximum pressure heads act as the
steady case. In addition, they have minimum percent of deviation in pressure heads than the
normal steady case, compared to other protection methods. Also, the figures show that the sudden
shut down of the pumps 36 and 37 together has a larger effect on the transient pressure heads than
the shut down of a single pump of them.

60

50 Close Pump 37
40 Close Pump 36
Close pumps 36, 37
30
Steady
Values of H (m)

20

10

0
0 50 100 150 200 250 300
-10

-20

-30

-40
Values of t (sec.)

Figure 3. Changes of pressure head with time due to different scenarios at node J17 for
unprotected case

6
 60

50

40

30
Values of H (m)

20

10

0
0 50 100 150 200 250 300
-10
Without protection
-20 NRV only
NRV with PRV
-30
NRV with surge tank
-40 NRV with air chamber
Values of t (sec.) Steady

Figure 4. Changes of pressure head with time at node J17 due to close pumps 36 and 37 together
and using different protection methods
60
Close Pump 37
50 Close Pump 36
Close pumps 36, 37
40 Steady
Values of H (m)

30

20

10

0
0 50 100 150 200 250 300
-10
Values of t (sec.)

Figure 5. Changes of pressure head with time due to different scenarios at node J2 for unprotected
case
60

50

40
Values of H (m)

30 Without protection
NRV only
NRV with PRV
20
NRV with surge tank
NRV with air chamber
10 Steady

0
0 50 100 150 200 250 300
-10
Values of t (sec.)

Figure 6. Changes of pressure head with time at node J2 due to close pumps 36 and 37 together
and using different protection methods

7
 60

50 Close Pump 37
Close Pump 36
40
Close pumps 36, 37
30
Steady
Values of H (m)

20

10

0
0 50 100 150 200 250 300
-10

-20

-30

-40

-50
Values of t (sec.)

Figure 7. Changes of pressure head with time due to different scenarios at node J21 for
unprotected case
60

50

40

30
Values of H (m)

20

10

0
0 50 100 150 200 250 300
-10

-20 Without protection

-30 NRV only
NRV with PRV
-40
NRV with surge tank
-50 NRV with air chamber
Values of t (sec.) Steady

Figure 8. Changes of pressure head with time at node J21 due to close pumps 36 and 37 together
and using different protection methods
To investigate the influence of water hammer on the transient flows through pipelines
representing different places on the network, pipeline P22 represents the middle pipe of the
network and pipeline P12 represents the nearer pipe from the pumps while pipeline P28 represents
the further pipe from the pumps. At each element, the transient change of discharges is
investigated. The investigations are based on the comparison between the use of different
protection devices and no protection case with the three studied scenarios of pumps shut down.
Steady normal operations are included for the comparison. Negative sign means that the discharge
is in the reverse direction in comparison with that of normal operation conditions. Figure 9 shows
the changes of flow rates with the time through pipe P22 for the three scenarios and the network
has no protection devices. It is seen that the shutting down of pump 36 or 37 only or both together
changes the discharge directions and values by different percents, the higher values are from +82
L/sec. to approximately -25, -77 and -166 L/sec. (minimum), respectively. Also, the first and third
shutting down scenarios, increase the discharge to be +170 L/sec. (maximum). For the nearer
pipelines from the pumps, Fig. 10 is drawn to show the variations of discharge with the time
through pipe P12 for the studied scenarios of pumps shutting down and without any protection
case. It is shown from the figure that, the shutting down of pump 36 only or the two pumps

8
together, changes the discharge directions and values by different percents, the higher values are
from +126 L/sec. to approximately -357 and -296 L/sec. (minimum), respectively but for the
shutting down of pump 37 only the transient discharge increases to +531 L/sec. (maximum). To
investigate the water hammer effect at the further pipelines from the pumps, the changes of flow
rate with the time through pipeline P28 for the previously mentioned scenarios of pumps shutting
down and the network has no protection devices are shown drawn as in Fig. 11. The figure shows
that the shutting down of pump 37 only or the two pumps together, changes the discharge
directions and values by different percents, the higher value is from +46 to nearly -202 L/sec.
(minimum). Also, the shutting down of pump 36 only or the both pumps together, increase the
discharge to +489 and +606 L/sec. (maximum), respectively.
Figures (12, 13 and 14) show the effect of using different protection devices on the transient
flow through the pipes of the studied network. The figures are drawn for the case of shutting down
of both pumps. It is seen from the figures that for all the selected pipes and for protecting the pipes
by using non-return valve only or NRV with pressure relief valve, the behavior of flow
fluctuations is same. Also, the cases of using non-return valve with open surge tank or with air
chamber, the discharge values and directions remain close to the values of steady operation with
very small fluctuations compared to the other protection methods.
In general, from Figs. (9 to 14) it is evident that, the best protection device on flow through the
pipe network from the water hammer harm is the using of air chamber or open surge tank with
non-return valve together. This due to the deviation in the transient discharge from the normal
case, decreases regularly to reach the final steady discharge without any waves or disturbances.
Also, the figures show that the final stable discharges after 300 seconds from hammering through
the pipelines, for all protection devices and unprotected case are equals and are the same of the
normal operation discharges. It is noticeable from the previous discussions that, the most affected
points and pipes by shutting the pumps are the closer points or pipes from the pump where they
have high fluctuations in both of pressures and discharges. Also, the time taken by the water
hammer wave to dissipate and reach stable state for the closer points or pipes to the shut pump is
longer than it in case of the further ones. The discussions show that, probably the middle points or
pipes have large effect than the end points or pipes. This may due to the water takes longer paths
to reach middle part of the network than the end part, consequently takes longer time which leads
to have small fluctuation and less water hammer wave time than the far points or pipes.

200

150

100
Values of Q (L/sec.)

50

0
0 50 100 150 200 250 300 350 400
-50

Close Pump 37
-100
Close Pump 36
-150 Close pumps 36, 37
Steady
-200
Values of t (sec.)

Figure 9. Changes of discharge with time due to different scenarios through pipe P22 for
unprotected case

9
 600 Close Pump 37

500 Close Pump 36

Close pumps 36, 37
400
Steady
300
Values of Q (L/sec.)

200

100

0
0 50 100 150 200 250 300 350 400
-100

-200

-300

-400
Values of t (sec.)

Figure 10. Changes of discharge with time due to different scenarios through pipe P12 for
unprotected case
650 Close Pump 37
Close Pump 36
550
Close pumps 36, 37
450 Steady
Values of Q (L/sec.)

350

250

150

50

-50 0 50 100 150 200 250 300 350 400

-150

-250
Values of t (sec.)

Figure 11. Changes of discharge with time due to different scenarios through pipe P28 for
unprotected case
200

150
Values of Q (L/sec.)

100

50

0
0 50 100 150 200 250 300 350 400
-50

Without protection
-100
NRV only
-150 NRV with PRV
NRV with surge tank
-200 NRV with air chamber
Values of t (sec.) Steady

Figure 12. Changes of discharge through pipe P22 with time due close pumps 36 and 37 together
and using different protection methods

10
 400

300
Values of Q (L/sec.)

200

100

0
0 50 100 150 200 250 300 350 400
-100 Without protection
NRV only
-200 NRV with PRV
NRV with surge tank
-300 NRV with air chamber
Values of t (sec.) Steady

Figure 13. Changes of discharge through pipe P12 with time due close pumps 36 and 37 together
and using different protection methods
700 Without protection
NRV only
600
NRV with PRV
500 NRV with surge tank
Values of Q (L/sec.)

400 NRV with air chamber

Steady
300

200

100

0
0 50 100 150 200 250 300 350 400
-100

-200

-300
Values of t (sec.)
Figure 14. Changes of discharge through pipe P28 with time due close pumps 36 and 37 together
and using different protection methods

CONCLUSIONS

1. Using non-return valve with open surge tank or with air chamber have the same effect and
protect the pipe network from negative pressures as well as high pressures and high flow
fluctuations.
2. Although using pressure relief valve with non-return valve protect the pipe network from
extreme pressures, it has high fluctuations for along time.
3. Using non-return valve only doesn’t have a great effect on the maximum pressure head than
the normal case and protects the network from the more low pressures resulting from the water
hammer than without the protection case.
4. Using non-return valve only or with pressure relief valve has the same flow fluctuations which
decreased more than the unprotected case.
5. The all protection methods used in this study save the pipe network from the water hammer but
increase its wave period.

11
REFERENCES

[1] Abd El-Gawad, S. M., “Water hammer analysis for the pipeline Ahmed Hamdi tunnel, Abu-
Radis”, Journal of Eng. Res., Helwan University, Vol. 6, 1994, pp. 40-54.
[2] Abozeid, G., Mohamed, H. I., and Hassan, I. M., “Studying of water hammer phenomenon
caused by sudden variation of water demand at water supply pipes network”, Journal of Eng.
Science, Assiut University, Vol. 40, Iss. 2, 2012, pp. 353-366.
[3] Abreu, J., Cabrera, E., Izquierdo, J., and Garcia-Serra J., “Flow modeling in pressurized
systems revisited”, Journal of Hyd. Eng.,Vol. 125, Iss. 11, 1999, pp. 1154-1169.
[4] Achouyab, E. H., and Bahrar, B., “Modeling of transient flow in plastic pipes”, Journal of
Contemporary Engineering Sciences, Vol. 6, Iss. 1, 2013, pp. 35-47.
[5] Ali, N.A., Mohamed, H.I., El-Darder, M.E., and Mohamed, A.A., “Analysis of transient flow
phenomenon in pressurized pipes system and methods of protection”, Journal of Eng. Science,
Assiut University, Vol. 38, Iss. 2, 2010, pp. 323-342.
[6] Al-Khomairi, A. M., “Use of the steady-state orifice equation in the computation of transient
flow through pipe leaks”, The Arabian J. for sci. and Eng., Vol. 30, Iss. IB, 2005, pp. 33-45.
[7] Batterton, S., “Water hammer: an analysis of plumbing systems, intrusion, and pump
operation”, M.Sc. Thesis, Virginia Polytechnic Institute and State University, Civil Eng. 2006.
[8] Chamani, M. R., Pourshahabi, S., and Sheikholesalm, F., “Fuzzy genetic algorithm approach
for optimization of surge tanks”, J. of Scientia Iranica A, Vol. 20, Iss. 2, 2013, pp. 278-285.
[9] Chaudhry, H. M., “Applied hydraulic transients”, 2nd ed., Van Nostrand, New York, 1987.
[10] Choon, T., W., Aik, L., K, Aik, L., E., and Hin, T., T., “Investigation of water hammer effect
through pipeline system”, Journal of Advanced Science Eng. Information Technology, Vol. 2,
Iss. 3, 2012, pp. 48-53.
[11] Fitzgerald, R., and Van Blaricum, V. L., “Water hammer and mass oscillation”, (WHAMO)
3.0 user's manual, U.S. Army Corp of Engineers Construction Engineering Research
Laboratories ADP Report 98/129, 1998.
[12] Fouzi, A., and Ali, F., “Comparative study of the phenomenon of propagation of elastic waves in
conduits”, Proc. of the World Cong. on Eng. (WCE), July 6-8, Vol. III, London, U.K., 2001.
[13] Friedman, M., Radder, L., Harrison, S., Howie, D., and Britton, M., “Verification and control
of pressure transients and intrusion in distribution systems”, AwwaRF, 2004.
[14] Kim, S. H., “Impulse response method for pipeline systems equipped with water hammer
protection devices”, Journal of Hyd. Eng., Vol. 134, Iss. 7, 2008, pp. 961-969.
[15] Lohrasbi, A. R., and Attarnejad, R., “Water hammer analysis by characteristic method”,
Journal of Engineering and Applied Sciences, Vol. 1, Iss. 4, 2008, pp. 287-294.
[16] Magzoub, M. A. E., and Kwame, S. O. A., “Hydraulic transient in a pipeline using computer
model to calculate and simulate transient”, Master’s thesis, Lund Univ., Dept. of Building and
Environmental Technology, Sweden, 2007.
[17] Mohamed, H. I., “Parametric study for the water hammer phenomenon in pipelines”, 1st Int.
Conf. of civil Eng. Science, ICCES1, 7-8 Oct., Vol. II, Assiut, Egypt, 2003.
[18] Nitescu, C. S., Constantin, A., and Stănescu, M., “Hydraulic study on pumping stations
equipped with air chamber mounted next to the pump”, Journal of Mathematical Models and
Methods in Applied Sciences, Vol. 5, Iss. 8, 2011, pp. 1318-1325.
[19] Ramos, H., Covas, D., Borga, A., and Loureiro, A., “Surge damping analysis in pipe systems:
modeling and experiments”, Journal of Hyd. Res., Vol. 42, Iss. 4, 2004, pp. 413-425.
[20] Simpson, A. R., and Wu, Z. Y., “Computer modelling of hydraulic transient in pipe networks
and associated design criteria”, MODSIM97, International Cong. on Modelling and
Simulation, Modelling and Simu. Society of Australia, Hobart, Tasmania, Australia, 1997.
[21] Stephenson, D., “Simple guide for design of air vessels for water hammer protection of
pumping lines”, Journal of Hyd. Eng., Vol. 128, Iss. 8, 2002, pp. 792-797.

12