Sustainable Energy Technologies and Assessments 44 (2021) 100990

Contents lists available at ScienceDirect

Sustainable Energy Technologies and Assessments

journal homepage: www.elsevier.com/locate/seta

Investigation of the influence of dimensions and material of the pipes on the

water hammer effect in microbial fuel cells wastewater treatment plants
Khalideh Al Bkoor Alrawashdeh a, *, La’aly A Al-Samrraie b, Abeer Al-Bsoul c, AlEssa Abdullah a,
Amani Lymoon b, Eid Gul d, Hewen Zhou e, Qing Yang e, f, Pietro Bartocci g, *, Francesco Fantozzi g
a
Mechanical Engineering Department, Al-Huson University College, Al- Balqa’ Applied University, Al-Huson- Irbid, P. O. Box 50, 21510, Jordan
b
Water and Environmental Engineering Department, Huson University College, Al- Balqa’ Applied University, Jordan
c
Al-Balqa Applied University, Al-Huson University College, Department of Chemical Engineering, Jordan
d
Biomass Research Centre, University of Perugia, Via G. Duranti - Strada S. Lucia Canetola s.n.c, Perugia, Italy
e
China-EU Institute for Clean and Renewable Energy, Huazhong University of Science and Technology, Wuhan 430074, PR China
f
State Key Laboratory of Coal Combustion, Huazhong University of Science and Technology, Wuhan 430074, PR China
g
Department of Engineering, University of Perugia, Via G. Duranti 67, Perugia 06125, Italy

A R T I C L E I N F O A B S T R A C T

Keywords: Water hammer is a transient flow phenomenon generated in a pipeline, when there is a sudden change in ve­
Waste water treatment plants locity, due to the closing of valves. This is particularly important for plants based on complicated hydraulic
Water hammer circuits and piping, to avoid malfunctioning in the urban sewer network. In this study, the effect of a set of
Surge tank
parameters: length, diameter and material of the pipe with and without surge tank on transient pressure is
Transient pressure
Linear programming optimization
investigated by using the software HAMMER V8i. The results show that the maximum Transient Pressure is
strongly influenced by an increase of 41.24% of the pipe diameter when the Surge Tank is not present. When the
surge tank is inserted the transient pressure increases only of about 12.2%. So the surge tank can have a positive
influence on the entire network. Moreover, the increase in the length will lead to a decrease in the Total Pressure
of about 10% without Surge Tank and of 8.62% with Surge Tank. While the change in the roughness coefficient
of the material used for the pipes will cause a decrease in the Total Pressure of about 42.6%, without Surge tank,
and of 12.2% with Surge Tank.

(TF) is the shift from one steady-state to another in a fluid stream flow
system where the pressure and velocity change quickly [6]. The Tran­
Introduction sient Flow happens in all the fluid streams, confined and unconfined. In
a confined system, such as a pipeline of water, a water hammer (WH) is
Waste water treatment facilities consists often of quite developed an immediate change in the flow that produces large transient pressure
hydraulic networks with many pipes. We can think to the sewer fluctuations (TPF) [7]. Water hammer can be investigated in two
network, see [1]. Microbial fuel cells can also be connected with hy­ different ways: by the theory of rigid water column, which ignores the
draulic networks [2] or can be built up as decentralized systems [3,4]. fluid compressibility and by the pipe walls elasticity, integrating
Also in the case of decentralized systems when the plant is scaled up a investigation of pipes elasticity [8]. The pressure wave velocity may
small hydraulic circuit has to be realized as proposed in [5]. surpass 1000 m/s and the values of TPF shifts from very high to very low
Water hammer (or surging, or shock wave) is a commonly used term values, which causes serious damages in the water pipelines and high
in water treatment facilities; which is defined as the momentary increase noise, different cavitation forms and other serious problems. The causes
in pressure inside a pipe, which corresponds to pressure (acoustic) of WH can be many. The main four probable causes usually produce
waves. The phenomenon is based on a sudden change in the flow in large changes in pressure: sudden power failure of the pumps, starting or
pipes which can be due to three causes: pumps and valves; vapor pocket stopping of pumps, fast changes in the valve setting and unstable pump
collapse and impact of water following the rapid flow of air out of a vent. characteristic curve [9,10]. Hydraulic systems must be designed to
The design of the pumping station has to be carefully based on models to support both design and off-design operations and be safeguarded to
avoid this negative effect. This is the topic of this work. Transient flow

* Corresponding authors.
E-mail addresses: [email protected] (K. Al Bkoor Alrawashdeh), [email protected] (P. Bartocci).

https://doi.org/10.1016/j.seta.2020.100990
Received 6 July 2020; Received in revised form 15 November 2020; Accepted 28 December 2020
2213-1388/© 2020 Elsevier Ltd. All rights reserved.
K. Al Bkoor Alrawashdeh et al. Sustainable Energy Technologies and Assessments 44 (2021) 100990

Nomenclature Q Volumetric flow /Discharge (m3/s)

R R reservoir
a Speed of pressure wave (m/s) S Cross area (m2)
C Roughness coefficient ST Surge tank
CD Complex device TPF transient pressure fluctuation (kPa)
D Pipe diameter (m) v Flow velocity (m/s)
Ep Compression elasticity modules–pipe material WH Water hammer
Ew Compression elasticity modules–water ΔH Change in head loss
g Acceleration of gravity (m/s2) ΔP Change n pressure (kPa)
h Pressure head (m) Δv change of velocity (m/s)
HPP Hydroelectric power plant π Constant
J Junction ρ Density (kg/m3)
L Length of pipe (m) τ The thickness of pipe (m)
PW Wave velocity (m/s)

overcome all the above causes [11]. airflow in the ventilation tube of an ST. Riasiand Tazraei (2017)
In water supply systems, the Water Hammer shock can be enlarged acknowledged the expenses of a ST construction and suggested the idea
along the pressure pipe as a result of inertia and flexibility. However, in of replacing a ST with a surge relief valve [39].
complex operations of long water supply systems [6–8,12,13] equip­ Skulovich and Bent (2015) [39] optimized the position and di­
ment like surge tanks (ST) [14–16] chambers and valves of air [16,17] mensions of a ST with the purpose of improving transient control using
are commonly implemented in different studies to guarantee security. an algorithm method. This technique is useful for solving transient
Other important approaches are shown in [18–24]. boundaries. Furthermore, they considered a smooth relationship be­
By interrupting the continuity of the pressure pipes, the energy tween the tank volume and the maximum pressure and suggested that
carried by the water can be quickly released increasing and decreasing the protection devices selection is susceptible to transient event condi­
the level inside the Surge Tank [15]. Recently, with the development of tions. Feng et al. (2017) [40] introduced the idea of an improved multi-
computer science, numerical simulation became the predominant way objective bacterial-foraging chemotaxis gravitational search algorithm,
to transient analysis [25]. Different methods have been introduced and which controlled the WH pressure and the rate of the unit rotational
implemented for transient simulations, such as the finite different velocity in water supply system with two surge tanks at two sides. The
method and the finite element method [26]. self-adaptive auxiliary control (SAC) system was proposed by Wuyi et al.
Particularly, to reproduce the transient process in engineering pro­ (2019) [16] to improve the operation performance and make the surge
jects the method of characteristics (MOC) is the most used technique. tank more efficient.
However, the Mac Cormack technique is also used to solve hyperbolic The aim of this study is to investigate the effects of various param­
partial differential equations of the free surface flow in dynamic fluids eters on the WH, including a set of variables among which: Diameter (D),
[27]. It’s still quite difficult to simulate the device’s boundary condi­ Length (L) and Roughness coefficient of the pipes material (C). A project
tions. In the free-surface flow framework a shock capturing method is implemented in the TP HAMMER V8i Software, in order to investigate
should be investigated, to allow the convergence of the solution (see the effect of change in D, L, and C in two scenarios. The first scenario
[28–31]). analyses D, L, and C changes, without surge tank. While the second
Other studies reported how complex devices (CD) can become scenario analyses the changes in D, L, and C with surge tank. Finally,
simpler ones using lumped elements, maintaining the original charac­ linear programming optimization in MATLAB is applied to obtain stable
teristics [32]. Notably, conditions of simple pipe network can be rep­ conditions also without the ST. This is done by defining the optimum C,
resented by a system of equations, which expresses the CD’s physical L and D ranges which lead to minimize the WH effect and to reduce the
characteristics. Fang and Chen (2008) created a simulation system for a impact of the pressure fluctuation.
hydroelectric power plant (HPP) that contains upstream and down­
stream STs; that has proved to be effective [33]. Zhou and Xu (2018) Methodology
[34] demonstrated that hybrid simulation of 3D SP-VOF (Simplified
Piecewise Linear Interface Calculation and Volume of Fluid) can be an Any interruption in the system of pipes implies that the flow will start
efficient model, by modifying a simulation of a power-off process in producing pressure waves (PW). As the PW are produced, they create
pumped-storage of hydropower. TPF and alter flow conditions. Probably these transient conditions can
The ST is used in water supply systems to control TPF [35,36]. The have negative consequences which can be avoided by fitting analysis,
ST’s location and size can greatly affect the pressure control perfor­ construction and design considerations.
mance. Normally, to improve the pressure control, the ST should be An analysis is proposed to define the effects of different parameters,
located near to the transient generator, like pumps and valves. Should to minimize the WH problems. In the present work, HAMMER V8i
the size of the ST be insufficient, the TPF cannot be reduced effectively. software is applied with different factors, to reduce the impacts of WH,
However, a common ST is sufficient to meet most applications. Ver­ where the critical factors are: the diameter and the length of the pipes,
eideand Svingen (2017) [17] showed the effects of a ST throttling on their roughness coefficient and the surge tank position.
hydraulic transients in an HPP. The outcome confirms that throttling can In first stage of the study, the network is designed without surge tank,
cause some small improvement in TP control. while in the second stage of the study it is designed with a surge tank,
A linear mathematical model of irregular flow for the tailrace system which is implemented after the pump. In both stages the values of
with an open pipe based on the state-space technique was suggested by pressure are modelled by changing variable parameters. In each stage
Yang and Wang (2016) [37]. Also, by inspecting the water level vacil­ the pipe diameter after the pump, the length of the pipe after the pump
lation in tailrace ST, they discovered that the open tailrace pipe has a and the roughness coefficient of the material are changed, to examine
strong impact on the hydraulic transient process. Moreover, Yang and their effect on the WH.
Wang (2016) [37] and Riasi and Tazraei (2017) [38] examined the

2
K. Al Bkoor Alrawashdeh et al. Sustainable Energy Technologies and Assessments 44 (2021) 100990

Water hammer modeling ( )

S = πD2 /4 (7)
The newest mathematical models to evaluate the WH, based on the
Where S is the cross-sectional area of the pipe, and the D is the diameter
one-dimensional wave equation have been implemented by [10,17].
of the pipe (m).
The maximum line pressures and shock generation times in a water
Head losses in a network pipe system are due to a set of various
distribution system were correctly prognosticated by Joukowsky equa­
factors. The main cause of the head loss is normally the internal friction
tion, when immediate valve closures happened. To investigate a WH
between liquid particles moving at different speeds. Secondary causes of
case when a pump stops, hydraulic flow equations have been used.
head loss are turbulence and interruption of the streamlines, such as
Calculations to estimate the change of pressure and velocity can be made
interruptions produced from valves and other fittings in the pipe pres­
using the Joukowsky equation:
sure, or disruptions from a changing section area in a network. The head
ΔP = ρaΔV (1) loss may be divided into two parts, the steady-state head loss (Hstatic),
and the transient head loss (water hammer, h). The head loss (m) over
Where ΔP is the pressure rise, resulting from WH (N/m2), ρ is the density the length of the pipe, due to friction, is expressed by the Hazen Williams
of water (kg/m3), and a is the velocity of the impulse wave (m/s), while formula, that is based on an empirical relationship which relates the
ΔV is the velocity change of water in the pipeline (m/s). flow of water in a pipe with the physical properties of the pipe and the
ΔH = aΔV/g (2) pressure drop caused by friction.

In Eq. (2) ΔH is the pressure increase, due to WH, in terms of water hf =

10.67*L*Q1.852
(8)
column (m) and g is the gravitational acceleration (m/s2), which is equal C1.852 D4.8704
to 9.81 m/s2.
Hmax = Hstatic + h(hammer) + hf (9)
The pressure wave speed is a function of pipe diameter, liquid elas­
ticity module, thickness of wall, specific weight, elasticity module of the where the hf is the head loss (m), C is the pipe roughness coefficient, L is
pipe material and distance between the support points. The pressure the length of pipe (m), Q is the volumetric flow rate (m3/s), d is the
wave equation is expressed as [10,17,41]: inside pipe diameter (m), Hazen-Williams coefficient depends on the
√̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅ variation of the flow field and usually is considered as a fixed value for
√ K
√
a=√ ( ) (3) specific materials. C value is related to pipe material type, as shown in
√
ρ 1 + e*E DK
ψ Table 1.

Case study
Where K is the volumetric compressibility modulus of the liquid, ρ is the
liquid density, e is the pipe wall thickness, E is the pipe elasticity
The network runs from one reservoir to another. The network con­
modulus (Young); ψ is a factor related with the pipe supporting condi­
tains a pump, valve and a set of pipes (P1–P7) and junctions (J1–J6). The
tion (i.e., cross-section dimensions, pipe axial constraints etc.).
lowest water level of the reservoir − 1 (Res-1) is 383 m, and the
To check the inlet flow discharge, a valve was fixed up in the water
elevation difference between the reservoir − 1 and reservoir-2 (Res-2) is
supply pipe. It is placed after the pump and before the inlet of the
73 m. The total pipe length is 1.857 km. The total water volumetric flow
reservoir 2, where the pump is installed between reservoir 1 and the
is 0.468 m3/s. The study data inventory is presented in Table 2 and the
valve. Consider a pipe which is at one end joined to a bond (RES2) with
main characteristics of the network are reported in Table 4.
water and on the other end, a valve is fitted. In steady-state conditions
The pipeline elevation at the pipe inlet and outlet are shown in
water is flowing and at time equal to zero (t = 0) the valve is closed
Table 3. Here the main characteristics of the pump, the valve and the
abruptly. This makes the velocity equal to zero in front of the valve and
surge type are proposed.
consequently the water pressure increases, which drives to an expansion
Pump, main characteristics:
of the pipe diameter at the same time. The pressure wave reaches the
bond at the time [25]:
T = L/a (4)
Table 1
In the bond, the pressure wave is reflected at the water surface. In the Hazen-Williams coefficient, C for different material of pipe [42].
hydraulic concept, a water surface is regarded as a body of constant Pipe Material CH.W.
inertia. That means for a pressure wave a water surface acts as a solid Brass 130–140
means. When the wave is reflected in an unstable situation at the tran­ Cast iron (common in older water line)
sition region between the pipe cross-sectional area and the bond New, unlined 130
10-year-old 107–113
transpires.
20-year-old 89–100
After the region of transition (in the bond), the pressure increases 30-year-old 75–90
and the wave velocity decreases, after that the water starts to flow from 40-year-old 64–83
pipe toward the bond (flow from high pressure to lower pressure). That Concrete or Concrete lined
leads to collision between the water body and the wave. As a conse­ Smooth 140
Average 120
quence, the energy converts from pressure energy into kinetic energy. In
Rough 100
this modality water again flows into the pipe and enters the valve at Copper 130–140
vibration time, which is expressed as: Ductile iron (Cement mortar lined) 140
Glass 140
T = 2L/a (5) High-density polyethylene (HDPE) 150
Plastic 130–150
where L is the pipe length (m). Polyvinyl chloride (PVC) 150
The water velocity can be obtained using the following equation: Steel
Commercial 140–150
V = Q/S (6) Riveted 90–110
Welded (seamless) 100
Vitrified clay 110

3
K. Al Bkoor Alrawashdeh et al. Sustainable Energy Technologies and Assessments 44 (2021) 100990

Table 2 junction J1, to supply flow into the pipeline, upon the power failure,
Parameters of network (Diameter of all the pipes is 0.6 m and roughness coef­ keeping the upstream water column moving and minimizing the size of
ficient - Hazen-Williams C = 130 Ductile Iron). the vapor pocket at the high point (or even preventing it from forming).
Pipe Length Total length Junction Elevation Initial The Hydropneumatic Tank element properties in the Properties ed­
(m) from reservoir (m) Pressure itor are set as follows:
(m) (kpa)
PS1 50 50 Res-1 383
a. Elevation (Base) and Elevation are set to 408.000 m respect to J1;
–
PMP1S 40 90 PJ1 363 –
PMP1D 10 100 PJ2 363 977.18 b. The Operating Range Type is set to Elevation.
P-1 20 120 J1 408 536.7 c. The initial HGL (Initial) is set to 465 m.
P-2 380 500 J2 395 650 d. The (Initial) Liquid Volume is set to 14,200 L.
P-3 300 800 J3 395 638
e. The Minor Loss Coefficient (Outflow) is set to 1.0.
P-4 250 1050 J4 386 716
P-5 400 1450 J5 380 759 f. the Tank Calculation Model is set to Gas Law Model.
P-6 250 1700 J6 420 359 g. the Volume (Tank) is set to 20,000 L.
P-7 175 1875 Res-2 456 29 h. the treat as Junction field command in the software is set to True.
This means that the hydropneumatic tank is not included in the
calculations of initial conditions. Instead, the HGL in the hydro­
Table 3 pneumatic tank is assumed to be the same as if there was a junction at
Pipeline elevation at the pipe inlet and outlet. the tank location.
Pipe Elevation inlet (m) Elevation Outlet (m) i. The diameter (Tank inlet Orifice) is set to 450 mm.
Res-1 WSE*(383) Invert (380) j. the Ratio of Losses is set to 2.5.
PS1 380 363
PMP1S 363 363 First scenario
PMP1D 363 363
P-1 363 408
P-2 408 395 The network doesn’t contain a surge tank, as shown in Fig. 1. The
P-3 395 395 network of pipes will be investigated in the terms of WH by controlling
P-4 395 386 the PT. The TP is calculated with variable parameters, such as:
P-5 386 380
P-6 380 420
P-7 420 453 - variable diameter;
Res-2 Invert (453) WSE (456) - variable length
- variable roughness coefficient of the pipes.

Table 4 Only one parameter is changed per analysis, while the other remain
Design criteria of pipe line network. constant; so, the effect of each parameter is evaluated independently.
Design criteria of pipe line network
Second scenario
H 81.3 m Pump head
H0 73.0 m Static head (difference between Res-2 and Res-1)
In this scenario, the network of pipes includes a surge tank, which is
L 1875 m Pipeline length
D 600 mm Pipeline diameter = 0.600 M installed after the pump instead of junction 1 as shown in Fig. 2. The
t 8.0 mm Pipe wall thickness = 0.0080 M surge tank can significantly decrease the fluctuation of the TP.
Qp 1684.0 m3/ h Total flow (pump delivery capacity) = 0.4680 m3/ sec Furthermore, when the pressure rises in the pipe, the surge tank regu­
vw Wave velocity (m/s) larly absorbs water from the pipe to limit the rise of the TP and when the
ρ= 1000 Kg/ Water density
pressure decreases in the pipe, the surge tank will discharge water to the
m3 principal pipe to stop the decrease of the TP.
EW = 2.1E+09 N/m2 Compression elasticity modulus (Young’s modulus)
for water (Pa). Result and discussion
EP = 2.1E+11 N/m2 Tension elasticity modulus (Young’s modulus) for
pipe material (Pa)
g= 9.81 m/s2 Acceleration of gravity The behavior of WH on a network including a pipe with variable
diameter and length and roughness coefficient is estimated. For this
purpose, the HAMMER V8i software has been used to analyze the in­
a. Inertia (Pump and motor): This is the combined pump, shaft, and fluence of different variables on the TP value.
motor inertia: 17.2 kg*m2.
b. Speed (Full): 1760 rpm. First scenario evaluation
c. Specific Speed: SI = 25, US = 1280.
The equations of the characteristic design of the pipe system have
Valve, main characteristics: been determined. The TP is calculated for 3 changing parameters:
diameter, length and pipe roughness coefficient.
- Pump Valve Type: set to default (Check Valve). The power failure is
assumed to be instantaneous and the check valve is allowed to close Effect of the pipe diameter without surge tank
without any delay (zero) to protect the pump from damage. When In this study, the considered pipes diameters are in the range of
the valve closes the pump so it is already stopped. 500–1000 mm. All the characteristics are implemented as reported in
- The check valve is built into the pump, so the distance between the Tables 2–4, the length of the pipe is 20 m, the roughness factor 130 and
pump and the downstream check valve is negligible. the time of valve closure is 5 sec.
The results of the calculations and the behavior of PT with changing
Surge, main characteristics: diameter for all the pipes network (P1 - P7) are shown in Fig. 3 and
It is clear that high pressures are caused by the sudden collapse of a Table 5. The results show that as the pipe diameter increases the TP will
vapor pocket at node J1. So we can install a Hydropneumatic Tank at also decrease, however, the decreasing value is very small. The law

4
K. Al Bkoor Alrawashdeh et al. Sustainable Energy Technologies and Assessments 44 (2021) 100990

Fig. 1. Pipeline network without surge tank.

Fig. 2. Pipeline network with surge tank.

Fig. 3. The effect of change in the diameter of pipe on the value of TP.

The percentage of decrease in TP is about 41.2% with fitting R2 =

Table 5
0.95, as shown in the Fig. 3, and the change in diameter is linear with the
Maximum pressure with change in diameter at junction 1.
change in TP.
Diameter (mm) 500 600 700 800 900 1000 It could be inferred that with the increase of the diameter the TP
Maximum Pressure (kPa) 4151 3862 3258 2963 2964 2439
range decreased and as a consequence, the energy distribution is faster.
Moreover, with the increase of pipe diameter, the flow transient
describing the TP value based on the change in pipes diameter is behavior lasts less time. It is obvious that the designer must analyze the
expressed in the following equation: costs of increasing the diameter and must provide an optimized layout
for decreasing PT and decreasing the costs of purchasing and installing a
TP = − 3.299D + 5747 (10) new pipeline. Furthermore, when the diameter increases, the number of

5
K. Al Bkoor Alrawashdeh et al. Sustainable Energy Technologies and Assessments 44 (2021) 100990

required pumps increase. Table 6

Maximum pressure with change in length of pipe 1.
Effect of the pipe length without surge tank Length (m) 40 60 80 100 150 200
In this section a changing length of pipe P1 is considered, while the Maximum Pressure (kpa) 3497 3581 3207 3177 3248 3149
other parameters are constant. The values of pipe length are varied in
the range of 40–200 m. The network characteristics are constant as re­
ported in Tables 2–4, and the diameter of the pipe is 600 mm, the Table 7
roughness factor is 130 and the time of valve closure is 5 sec. Maximum pressure with change in roughness coefficient of pipe 1.
As it is shown in Fig. 4 and Table 6, when using the pipe with length
Roughness Coefficient (C) 100 110 120 130 140
200 m instead of 40 m, the maximum TP head decreased by 10% (Eq. Maximum Pressure (kpa) 3075 3586 3511 3862 4385
(12)). The TP behavior depending on the change of pipe length is
expressed in the following equation (Table 7):
According to Eq. (13), TP increases with the increase in the rough­
TP = − 2.183L + 3539 (11) ness coefficient of Hazen-William, because the roughness coefficient C
The mathematical correlation which expresses the change of TP has an inverse relation with the head loss hf, this implies an increase in C
depending on the change in pipe length isn’t linear and the linear cor­ and a decrease in hf and an increase in Hmax of water hammer, as it is
relation has a low coefficient (R2 = 0.514). To obtain a better correla­ shown in Eqs. (8) and (9).
tion the equation must be of fourth-degree (R2 = 0.888) as per the
following: Pressure profile without surge
From the Fig. 11, an increase in the pressure value is observed at J1,
TP = − 3*10− 5 L4 + 0.011L3 − 1.8L2 + 104.3L + 1537 (12) so the decision was to replace this point with a surge tank to reduce the
According to Eq. (12), the effect of the change of pipe length on the resulting pressure value (Figs. 6 and 7).
TP value is very little (≈ 10%). This same results were observed by The three different curves describe the relation of each curve against
Gómez (2018) [43] he investigated WH with different times of valve the distance plotted. The red curve presented that the pressure is
closure, and analyzed the pressure behavior along the pipeline during 0 maximum if the distance is small, while with the increase of distance the
≤ t ≤ 14 sec. His results showed that when the transient started, the pressure decreases. Mean while the green curve shows a variable
pressure gradually dropped (at 0 ≤ t ≤ 2); at 2 ≤ t ≤ 7 the pressure behavior with the distance. The third orange curve shows a steady state
continued to decrease (transient behavior propagates through the pipe linear behavior as the pressure remain constant with increasing of the
length) and at 7 ≤ t ≤ 14 the pressure increased in a small fraction (t = 7; distance.
transient back to pump & after that the transient travels through the
pipe), finally at t = 14 the transient reaches the pump and the pressure Pressure time history without surge
started decreasing again. The above graph presented a transient analysis of pressure against
the time. At the initial release time the pressure reaches a maximum
Effect of the roughness coefficient of the pipe material, without surge tank point and after the pressure drops with steady oscillations and reaches
In this section, the roughness coefficient of the pipe materials is its nominal range.
changed and other parameters are maintained constant. The TP value is
modelled with different roughness coefficients (only for P1). The Effect of surge tank
roughness coefficient (C) is varied from 100− 140, assuming that the In this part, the effect of the surge tank is detected, using the pa­
pipe is made of different materials. The TP increases by 42.6% with the rameters reported in Tables 2–4. The effect of variable parameter on WH
increase of the roughness coefficient. The relationship between the TP in terms of TP is investigated by varying three parameters: effect of
and the change in the roughness coefficient is presented in Eq. (13), the change in pipe diameter, change in pipe length and change in roughness
relationship is characterized by a linear increase, with R2 = 0.898 as coefficient with the use of a surge tank. The results show that the TP
shown in Fig. 5. decreases to 1462.21 KPa instead of 4151 KPa, this was already obtained
in the previous scenario without surge tank.
TP = 28.96C + 208.6 (13)

Fig. 4. The effect of change in the length of pipe on the value of TP.

6
K. Al Bkoor Alrawashdeh et al. Sustainable Energy Technologies and Assessments 44 (2021) 100990

Fig. 5. The effect of change in the roughness coefficient of the pipe on the value of TP.

Effect of pipe diameter with surge tank

The effect of the change in pipe diameter in the range of 500–1000
mm is analysed in this section. The TP values obtained are reported in
Table 8 and Table 9. As it is shown in Fig. 8, the change of pipe diameter
led to the decrease in the TP, according to the following equation:
TP = − 0.322D + 1693 (14)
The equation relating the change in pipe diameter with the change of
pressure has R2 = 0.618, this means that the linear correlation is poor;
the phenomenon is better described by a second-order equation (Eq.
(15)).

TP = − 0.001D2 + 2.05D + 849.9 (15)

2
R becomes 0.936 and it decreases of 9%. The result shows that the
surge tank decreases the effect of the change in diameter on the TP, by
absorbing the large portion of TP. However, the change in TP is more
sensitive to the change in pipe diameter than in the first scenario without
Fig. 6. Pressure profile without surge. surge tank.
Our results are confirmed by [42] in which the collateral damage in
the pipe network depends on pipe thickness.
Also, in agreement with the results reported by [44] the pressure
wave amplitude decrease with the increase in conduit diameter.

Effect of pipe length with surge tank

The effect of the change in pipe length within the range of 40–200 m
is examined in presence of a surge tank, and the effect of it on TP is
expressed as per the following Eq. (16):
TP = 0.753L + 1495 (16)
The effect of change in length here becomes better and more sensi­
tive, as the pipe length increases the TP increases as well (see Table 8),
with a percentage of 8.62%. Also, the relationship becomes linear as
shown in Fig. 9.
This result is confirmed by Choon and Aik (2012) results [8] who
investigated experimentally the effect of the L on the WH, they used two
Fig. 7. Time history without surge.
lengths: 1 m and 1.5 m as the length of the tested pipe. Their results
showed that the long pipe imply a larger pressure, compared with the
shorter pipe.

Table 8 Effect of the roughness coefficient of pipe material with surge tank
Maximum pressure with change diameter at junction 1. The roughness coefficient has been changed within the range of
100–140. The effect of the change in the roughness coefficient with
Diameter 500 600 700 800 900 1000
(mm) surge tank on TP is expressed by the following equation:
Maximum 1462.21 1538.37 1518.38 1462.69 1397.16 1332.6
TP = 4.28C + 979.9 (17)
Pressure
(kpa) As it is shown in Fig. 10, the effect of change in the roughness

7
K. Al Bkoor Alrawashdeh et al. Sustainable Energy Technologies and Assessments 44 (2021) 100990

Table 9
Maximum pressure with change in length of pipe 1.
Length (m) 40 60 80 100 150 200
Maximum Pressure (kpa) 1524.64 1531.78 1531.87 1563.61 1614.93 1656.01

Fig. 8. The effect of change in the diameter of pipe with surge tank on the value of TP.

Fig. 9. The effect of change in the length of pipe with surge tank on the value of TP.

Fig. 10. Effect of the change in roughness coefficient of pipe on TP.

8
K. Al Bkoor Alrawashdeh et al. Sustainable Energy Technologies and Assessments 44 (2021) 100990

Table 10
Maximum pressure with change in roughness coefficient of pipe 1.
Roughness Coefficient 100 110 120 130 140
(C)
Maximum Pressure 1403.82 1452.87 1497.75 1538.37 1575.08
(kpa)

Fig. 11. Pressure profile with surge.

coefficient is linear with TP and R2 = 0.996. The change in the pipe

material with a surge tank becomes quite effective on TP (see Table 10)
with an increase of about 12.2%.
The regression equation (Fig. 10) shows that the pipe material has
not a significant influence on the pressure and on the flow.
Fig. 12. Time history with surge.
This result is confirmed by Bombardieri (2014) [45] they deduced by
using EcosimProR software in conjunction with ESPASS library that the
friction value must be increased up to 3 times to take into account the
unsteady flow condition. It is also confirmed by the results reported by
Choon and Aik (2012) [46–53] that show that the WH effect in PVC is
greater than steel [8].

Time history with surge

Fig. 12 presents the transient analysis where the pressure rapidly
increases at the initial release but with the time the pressure drops and
gives a more linear behavior.

Reduction of water hammer effect, without surge tank, through optimization

Actually, installing a surge tank in a pipe network requires high
capital costs, to avoid these costs we suggest to use optimal parameters
(at least in the designing phase), such as pipe diameter, length and
material roughness. By using linear programming optimization imple­
mented in MATLAB for all the linear curves which have been obtained
through fitting in the 1st scenario for the 3 parameters taken into ac­
count, we define the optimum solution, to avoid using a surge tank, as it
is shown in Fig. 13.
Fig. 13 shows the effect of diameter and length of pipe and pipe Fig. 13. Optimum solution, to avoid installing a surge tank.
material on the TP. The shaded area in Fig. 13 is the zone that represents
the best values for the variables (D, L, C) without the surge tank. The tank (linear relation). Moreover, the TP will be decreased if the D in­
optimal point is that with pipe diameter of 500 mm, length of P1 of 20 m, creases by 41.2% without surge tank (linear relation) and 9% with the
roughness coefficient of 130. surge tank (non-linear relation). While the increase in L implies a
decrease in TP by 10% without surge tank (non-linear relation) and
Conclusions 8.62% with the surge tank (linear relation).
The linear programming optimization using MATLAB obtained an
This study on the water hammer (HW) phenomena in the pipeline is optimum combination of pipe material, diameter and length, in order to
very important for household and industrial wastewater treatment avoid to use the surge tank. This can be achieved by determining the
plants. This leads to an increase in the transient pressure TP and as a optimum range of D, L and C.
result of wave velocity. HAMMER V8i software was used to investigate
the effect of some parameters with surge tank and without surge tank on Declaration of Competing Interest
the Total Pressure. These parameters are the diameter (D), the pipe
length (L) and the pipe material. The authors declare that they have no competing financial interests
The results of the study show that the previously mentioned pa­ or personal relationships that could have appeared to influence the work
rameters influence the water hammer phenomenon. The pipe material reported in this paper.
effect is modelled through the use of a roughness coefficient (C). We
concluded that the HW effect in the steel pipe is lower than the HW References
effect in PVC pipe, when the C increases the TP will increase as well by
42.6% without the surge tank (linear relation) and 12.2% with the surge [1] Ridgway K, Kumpula GJ. Surge modeling in sewers using alternative hydraulic
software programs. JWMM 2008. https://doi.org/10.14796/JWMM.R228-10.

9
K. Al Bkoor Alrawashdeh et al. Sustainable Energy Technologies and Assessments 44 (2021) 100990

[2] Winfield J, Ieropoulos I, Greenman J. Investigating a cascade of seven [29] Triki Ali. Resonance of free-surface waves provoked by floodgate maneuvers.
hydraulically connected microbial fuel cells. Bioresour Technol 2012;110:245–50. J Hydrol Eng 2014;19(6):1124–30.
[3] Arkatkar A, Mungray AK, Sharma P. Bioelectrochemical behaviour of a [30] Triki A. Multiple-grid finite element solution of the shallow water equations: water
sequentially added biocatalytic coculture in a microbial fuel cell. J Basic Microbiol hammer phenomenon. Comput Fluids 2014;90:65–71.
2020;60(7):562–73. [31] Triki A. A finite element solution of the unidimensional shallow-water equation.
[4] Pushkar P, Mungray AK. Real textile and domestic wastewater treatment by novel ASME J Appl Mech March 2013;80(2):021001.
cross-linked microbial fuel cell (CMFC) reactor. Desalin Water Treatment 2016;57 [32] Karney BW, McInnis D. Efficient calculation of transient flow in simple pipe
(15):6747–60. networks. J Hydraul Eng ASCE 1992;118:1014–30.
[5] Dannys E, Green T, Wettlaufer A, Madhurnathakam CMR, Elkamel A. Wastewater [33] Fang HQ, Chen L, Dlakavu N, Shen ZY. Basic modeling and simulation tool for
treatment with microbial fuel cells: a design and feasibility study for scale-up in analysis of hydraulic transients in hydroelectric power plants. IEEE Trans Energy
microbreweries. J Bioprocess Biotech 2016;6(1). https://doi.org/10.4172/2155- Convers 2008;23:834–41.
9821.1000267. [34] Zhou J, Xu Y, Zheng Y, Zhang Y. Optimization of guide vane closing schemes of
[6] Elbashi M, Amoah S. Hydraulic transient in a pipeline (using computer model to pumped storage hydro unit using an enhanced multi-objective gravitational search
calculate and simulate transient). M.Sc. Thesis, Department of Building and algorithm. Energies 2017;10(911).
Environmental Technology, Lund University; 2007. [35] Vournas CD, Papaioannou G. Modeling and stability of hydro plant with 2 surge
[7] Batterton S. Water hammer: an analysis of plumbing systems, intrusion, and pump tanks. IEEE Trans Energy Convers 1995;10:368–75.
operation. M.Sc. Thesis, Virginia Polytechnic Institute and State University; 2006. [36] Demello FP, Koessler RJ, Agee J, Anderson PM, Doudna JH, Fish JH, et al.
[8] Choon T, Aik L, Hin T. Investigation of water hammer effect through pipeline Hydraulic-turbine and turbine control-models for system dynamic studies. IEEE
system. Int J Adv Sci Engand Inform Technol 2012;2(3):2088–5334. Trans Power Syst 1992;7:167–79.
[9] Su CK, Camara C. Cavitation luminescence in a water hammer: upscaling [37] Yang JD, Wang MJ, Wang C, Guo WC. Linear modeling and regulation quality
sonoluminescence. J Phys Fluids 2003;15:1457–61. analysis for hydro-turbine governing system with an open tailrace channel.
[10] Wood DJ. Water hammer analysis – essential and easy. J Environ Eng 2005;131: Energies 2015;8:11702–17.
1123–31. [38] Riasi Alireza, Tazraei Pedram. Numerical analysis of the hydraulic transient
[11] Jalut QH, Ikheneifer AA. Mathematical simulation for transient flow in pipe under response in the presence of surge tanks and relief valves. Renew Energy 2017;107:
potential water hammer. Diyala J Eng Sci 2010;22(23):222–36. 138–46.
[12] Yu X, Zhang J, Miao Di. Innovative closure law for pump-turbines and field test [39] Skulovich O, Bent R, Judi D, Perelman LS, Ostfeld A. Piece-wise mixed integer
verification. J Hydraul Eng 2015;141(3):05014010. https://doi.org/10.1061/ programming for optimal sizing of surge control devices in water distribution
(ASCE)HY.1943-7900.0000976. systems. Water Resour Res 2015;51:4391–408.
[13] Wylie EB, Streeter VL, Suo L. Fluid transients in systems prentice hall. Englewood [40] Feng ZK, Niu WJ, Zhou JZ, Cheng CT. Multiobjective operation optimization of a
Cliffs, NJ, USA; 1993. cascaded hydropower system. J Water Resour Plann Manage 2017;143:10.
[14] Wan W, Zhang B. Investigation of water hammer protection in water supply [41] David C. Hazen-Williams C-factor Assessment in an Operational Irrigation Pipeline.
pipeline systems using an intelligent self-controlled surge tank. Energies 2018;11 Montreal, august: Department of Agricultural and Biomass Engineering McGill
(6):1450. University; 2001.
[15] Vereide K, Svingen B, Nielsen TK, Lia L. The effect of surge tank throttling on [42] Roy Joyanta Kumar, Basak Pijush. Investigating water hammer problem in piped
governor stability, power control, and hydraulic transients in hydropower plants. water distribution system. IOSRJAP 2014;6(2):51–8.
IEEE Trans Energy Convers 2017;32:91–8. [43] Gómez AM. Physical aspects of air in pipe systems, including its effect on pipeline
[16] Wan W, Zhang B, Chen X, Lian J. Water hammer control analysis of an intelligent flow capacity and surge pressure (water hammer). M.Sc. Thesis, Department of
surge tank with spring self-adaptive auxiliary control system. Energies 2019;12 Civil Engineering, Aalborg University; 2018.
(13):2527. [44] Gabriel PP. Influence of pipe-diameter on water hammer phenomenon. JMEA
[17] Bergant A, Kruisbrink A, Arregu F. Dynamic behaviour of air valvesina large-scale 2015;5(6). https://doi.org/10.17265/2159-527510.17265/2159-5275/
pipeline apparatus. Strojniškivestnik J Mech Eng 2012;58:225–37. 2015.0610.17265/2159-5275/2015.06.006.
[18] Trabelsi M, Triki A. Exploring the performances of the dual technique-based water [45] Bombardieri C, Traudt T, Manfletti C. Experimental and numerical analysis of
hammer redesign strategy in water supply systems. J Water Suppl Res Technol- water hammer during the filling process of pipeline. In: Space Propulsion. https://
Aqua 2020;69(1):6–17. www.ecosimpro.com/wp content/uploads/2015/02/SP2014_2968541_
[19] Trabelsi M, Triki A. Dual control technique for mitigating water-hammer Bombardieri.pdf; 2014.
phenomenon in pressurized steel-piping systems. Int J Press Vessels Pip 2019;172: [46] Michael J, Morgan MJ. Pressure transient characterization test for star-2
397–413. propulsion system fuel manifold. In: Paper presented at the 40th AIAA/ASME/
[20] Triki Ali, Chaker Mohamed Amir. Compound technique-based inline design SAE/ASEE Joint Propulsion Conference and Exhibit, 11–14 July 2004. Fort
strategy for water-hammer control in steel pressurized-piping systems. Int J Press Lauderadale Florida.
Vessels Pip 2019;169:188–203. [47] Hearn H. Development and application of a priming surge analysis methodology.
[21] Triki Ali. Dual-technique-based inline design strategy for water-hammer control in In: Paper presented at the 41st AIAA/ASME/SAE/ASEE Joint Propulsion
pressurized pipe flow. Acta Mech 2018;229(5):2019–39. Conference & Exhibit, 10–13 July 2005. Tucson, Arizona.
[22] Fersi M, Triki A. Investigation on redesigning strategies for water-hammer control [48] Scroggins AR. A streamlined approach to venturi sizing. In: Paper presented at the
in pressurized-piping systems. ASME J Pressure Vessel Technol 2019. 48th AIAA/ASME/SAE/ASEE Joint Propulsion Conference & Exhibit, Vol. 30
[23] Triki Ali. Water-hammer control in pressurized-pipe flow using a branched July–1 August 2012, 2012-4028. Atlanta, Georgia.
polymeric penstock. J Pipeline Syst Eng Pract 2017;8(4):04017024. https://doi. [49] Lecourt Renaud, Steelant Johan. Experimental investigation of waterhammer in
org/10.1061/(ASCE)PS.1949-1204.0000277. simplified feed lines of satellite propulsion systems. J Propul Power 2007;23(6):
[24] Triki Ali. Water-hammer control in pressurized-pipe flow using an in-line 1214–24.
polymeric short-section. Acta Mech 2016;227(3):777–93. [50] Prickett RP, Mayer E, Hermel J. Water hammer in a spacecraft propellant feed
[25] Ghidaoui MS, Zhao M, McInnis DA, Axworthy DH. A review of water hammer system. J Propul Power 1992;8:592–7.
theory and practice. Appl Mech Rev 2005;58:49–57. [51] Lin TY, Baker TD. Analysis and testing of propellant feed system priming process.
[26] Chaudhry M, Hussaini M. Second-order accurate explicit finite-difference schemes J Propul Power 1992;11:505–12.
for water hammer analysis. J Fluids Eng 1985;107:523–9. [52] Yang JD, Wang H, Guo WC, Yang WJ, Zeng W. Simulation of wind speed in the
[27] Triki A. Further investigation on the resonance of free-surface waves provoked by ventilation tunnel for surge tanks in transient processes. Energies 2016;9(95).
floodgate maneuvers: negative surge waves. J Ocean Eng 2017;2017(133):133–41. [53] Zhou D, Chen H, Zhang L. Investigation of pumped storage hydropower power-off
[28] Triki A. Further investigation on the resonance of free-surface waves provoked by transient process using 3D numerical simulation based on SP-VOF hybrid model.
floodgate maneuvers: negative surge waves. Ocean Eng 2017;133:133–41. Energies 2018;11:1020.

10