Renewable and Sustainable Energy Reviews 45 (2015) 135–144

Contents lists available at ScienceDirect

Renewable and Sustainable Energy Reviews

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

State of the art in numerical modelling of Pelton turbines

Audrius Židonis, George A. Aggidis n
Lancaster University Renewable Energy Group and Fluid Machinery Group, Engineering Department, Lancaster LA1 4YR, United Kingdom

art ic l e i nf o a b s t r a c t

Article history: Pelton turbine (or Pelton wheel) is among the most efficient impulse turbines and has retained its
Received 1 October 2014 existence in hydropower for well over a century. However unlike in the development of the reaction
Received in revised form turbines, where Computational Fluid Dynamics (CFD) have been successfully applied for more than 20
19 November 2014
years now, up until recently it was not feasible to perform CFD analysis of Pelton turbines due to the
Accepted 6 January 2015
nature of the flow which is much more complex than in the reaction turbines. The recent developments
in CFD models and tools together with the continuous increase in computational resource are bringing
Keywords: the CFD modelling up to a level suitable for industrial applications in development of Pelton turbines.
Renewable energy Current published research in the field worldwide can be divided into two distinct branches of CFD
Hydropower
models: the Eulerian specification of flow field, which tends to be more accurate, but also more
Pelton turbine
computationally expensive, and the Lagrangian specification which is known to be less computationally
Numerical modelling
Computational fluid dynamics demanding, however to date it cannot compete with Eulerian specification in terms of accuracy. This
Optimisation review paper is aiming at establishing the state of the art in numerical modelling of Pelton Turbines and
would serve as guidance when choosing the optimum CFD modelling methodology and software
available.
& 2015 Published by Elsevier Ltd.

Contents

1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 136
2. Background . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 136
2.1. Hydropower . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 136
2.2. Reaction and impulse turbines . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 136
2.3. Pelton turbine . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 136
2.4. Difficulties in Pelton turbine modelling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 137
3. Pre-CFD development of Pelton turbines . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 138
3.1. History . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 138
3.2. Available design guidelines . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 138
4. Computational fluid dynamics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 139
4.1. Governing equations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 139
4.2. Eulerian and Lagrangian methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 139
5. Application of CFD on Pelton turbines . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 139
5.1. ANSYS Products (CFX and Fluent) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 139
5.1.1. Injector and jet simulations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 140
5.1.2. Stationary bucket simulations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 140
5.1.3. Rotating bucket simulations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 140
5.2. OpenFOAM . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 142
5.3. Other Eulerian solvers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 142
5.4. Smoothed particle hydrodynamics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 142
5.5. Fast Lagrangian solver . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 142

n
Corresponding author. Tel.: þ 44 (0) 1524 593052.
E-mail address: g.aggidis@lancaster.ac.uk (G.A. Aggidis).

http://dx.doi.org/10.1016/j.rser.2015.01.037
1364-0321/& 2015 Published by Elsevier Ltd.
136 A. Židonis, G.A. Aggidis / Renewable and Sustainable Energy Reviews 45 (2015) 135–144
5.6. Other Lagrangian solvers. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
6. Discussion and suggestions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
7. Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Acknowledgments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
1. Introduction
Energy resources nowadays known as renewable resources
were the very first ones to be harnessed by humankind. The first
ships that were not powered by manpower were sailing-ships, the
first automated mills were powered by wind or hydro power and
the first heat source was firewood. It is only later when fossil-fuel
and nuclear technologies were discovered and boosted the indus-
try. The rapid development of the modern world happened
because of the relatively high controllability and flexibility of fossil
and nuclear power technologies at the time, leaving the renewable
resources aside. However, the intensive use of conventional energy
resources has gradually produced many problems such as pollu-
tion, contaminated waste management, depletion of resources and
in many cases strong dependence on imported unreliable supplies.
It was because of the reasons like these that the world started
reconsidering the sustainable and renewable power resources as
with modern technology more and more renewables are becom-
ing competitive with conventional power [1] which is getting
more expensive as the supply is getting shorter with time. Not to
mention the world's concern about the impact made on our planet
expressed by various international agreements like a binding
target of all the EU members to produce 20% of final energy
consumption from renewable power resources which was stated
in Renewable Energy Sources (RES) Directive which entered into
force in June 2009 [2] or Kyoto Protocol aiming to reduce green-
house gasses worldwide which entered into force in 2005 [3]. It is
a clear message that renewable energy is the way to proceed.
2. Background
2.1. Hydropower
There are plenty of different sustainable energy resources that
could be used but usually there is no single resource that could
provide enough energy on its own to completely satisfy the
demand. That is why each and every resource is important and
they all have to be used collectively. However, some countries are
richer with one type of resources and some with another. For
example in the United Kingdom, the amount of solar energy that
could be taken from direct sunlight would not be such a big player
as hydropower. The average UK's demand for electricity in 2013
was around 35 GW [4], whereas there are studies that show a
capacity of 1.5 GW (more than 4% of the average demand) of
untapped hydro power available in the UK [5].
Same as the wind is a transformed energy from the sun which
heats the air and causes convection, hydro power is stimulated by
radiant energy supplied by the sun. As the radiant energy heats
the surface of the Earth most of which, 70% [6], is covered by water
(mostly the oceans) the evaporation is caused. This is how water
molecules are lifted. When hot air and water mixture masses reach
high altitudes and cool down, the condensation takes place, which
results in water droplets emerging and then forming streams and
rivers that go back to the oceans. The whole phenomenon is
known as the hydrologic cycle [7]. It could be seen from the
potential energy formula (Eq. 1) that after water vapour condenses
142
142
143
143
143
back into water droplets it has potential energy because of their
altitude, h.
Ep ¼ mgh ð1Þ
Where Ep is potential energy in J, m is mass in kg, g is acceleration
due to gravity in m/s2 and h is altitude in m.
According to Mosonyi [8] under normal run of river all that
potential energy is dissipated in a form of heat loss when over-
coming friction and creating eddies and swirls. Hence, Mosonyi
states that ‘the fundamental principle of water power development is
to reduce the amount of energy dissipated as heat, without paralysing
the flow of water’. These are the main reasons why hydropower is a
very attractive field for further development. There is a huge
variety of hydro turbines depending on specific requirements and
on available river conditions. Based on the working principles they
are separated into two distinct classes: reaction and impulse
turbines and will be described in the following section.
2.2. Reaction and impulse turbines
In general reaction turbines produce power by combining
kinetic energy of the moving water and potential energy available
from the pressure difference. Popular examples of reaction tur-
bines are Kaplan, Francis or Archimedes Screw. Usually reaction
turbines are used for lower head and higher flow applications than
impulse turbines. Typically reaction turbines are completely sub-
merged in the water. Fig. 1 presents performance envelopes of
these turbines and gives graphical comparison to impulse turbines
that are described in the following paragraph.
Impulse turbines generate power by converting potential
energy available from the pressure difference, i.e. difference in
water levels upstream and downstream. It converts this potential
energy into movement by discharging water into atmospheric
pressure to act on the runner. The water stream hits each blade or
bucket, therefore there are two phases around the runner: water
and air. Examples of impulse turbines are Pelton, Turgo or Cross-
flow turbines. Typically impulse turbines are used for higher head
and lower flow rate applications than reaction turbines as shown
in Fig. 1.
2.3. Pelton turbine
Pelton turbine (or Pelton wheel) is among the most efficient
impulse turbines and has retained its existence in hydropower for
well over a century since it was invented by Lester A. Pelton [11] in
1880. The turbine produces power by utilising water momentum
impinging on buckets mounted on the periphery [12,13]. Despite
its age, the design of Pelton turbine keeps improving [14] and this
development is driven by a tough commercial competition
between turbine manufacturers and availability of new tools for
analysis and optimisation. The guidance for designing of Pelton
turbine available in the public domain is based on existing know-
how. This means that any design improvements were mainly
conducted after extensive experimental testing by the trial-and-
error approach. However, experimental testing is a very complex
task itself [15]. Not to mention the high costs and very long time
scales of manufacturing that would be inevitable part of prototype
 A. Židonis, G.A. Aggidis / Renewable and Sustainable Energy Reviews 45 (2015) 135–144
Fig. 1. Performance envelopes of hydro turbines [9,10].
Fig. 2. Illustrations of the Pelton wheel from its patent [11].
testing. In the recent years significant effort has been directed
towards a better understating of the details of the complex
unsteady flow in the runner with the aid of modern numerical
modelling known as Computational Fluid Dynamics (CFD).
2.4. Difficulties in Pelton turbine modelling
Computational Fluid Dynamics (CFD) has been successfully
used as numerical modelling tool for developing of reaction
turbines for more than 20 years now [16]. However, modelling of
impulse turbines and in this case Pelton is still a challenging task.
The main difficulties in simulating the performance of Pelton
turbines are pressure losses, secondary flows, jets, film flow, free
137
Fig. 3. Sketch of a bucket that has a cutout published in 1937 [25].
surfaces, spray formation, ventilation losses, unsteadiness and
complex interaction between components [17]. In addition to that,
quasi-steady state approximations do not work for flows in Pelton
buckets, requiring completely unsteady time dependent simula-
tions with rotor-stator interaction to be used, hence significantly
adding to the already high computational demands. Fortunately,
despite the difficulties explained above, the technology is being
pushed forward making simulated results to agree with the
experimental data satisfactorily and allowing the computed meth-
ods to be used for development of new designs or optimisation of
the existing ones.
There are four major sections of interest for efficiency analysis:
distributor, nozzle, bucket and casing. The distributor and nozzle
play an important role for jet quality which is very important for
both efficiency of the whole system [18] and the operational life of
the runner as for high head application the nozzle dispersion can
induce damages on buckets [17,19,20] On the other hand, the
efficiency of a turbine is affected by the bucket shape, which has
already improved significantly since the original Pelton design.
However, an accurate technique for modelling the flow in a
rotating Pelton runner is required for further improvements and
design validations as it is done for such turbines like Francis or
Kaplan at the moment. Last but not least is the casing design,
which is very important for Pelton turbines as it might cause flow
energy losses due to disturbance of incoming jets interfering with
water sheets that have not evacuated [21]. Successful simulation of
138 A. Židonis, G.A. Aggidis / Renewable and Sustainable Energy Reviews 45 (2015) 135–144

the whole system from the branchpipe to the casing is important

but maybe not feasible at the moment because of such limiting
factors like timescale or very high computational costs. That is
why compromises are introduced to achieve an optimum effect
within economically reasonable costs.

3. Pre-CFD development of Pelton turbines

3.1. History

The Pelton turbine or Pelton Wheel was invented by Lester A.

Pelton [11] in 1880 and followed by few modifications published in
the same century [22,23]. Impulse turbines that were available
before that time were extremely inefficient. The first Pelton
turbine shown in Fig. 2 [11] consisted of rectangular shape buckets
that had a splitter in the middle to symmetrically divide the jet
into two streams and deflect the flow back almost through 1801.
However, first buckets of Pelton turbine had no cutout which is
always present in modern designs of Pelton and the injector
design was very simplistic.
In the first half of the XX century Pelton turbines have evolved
into something more similar to what they look nowadays. There
are publications available [24,25] showing a bucket shape that has
a cutout (Fig. 3), injector design that includes spear valve (Fig. 4) Fig. 5. Sketch of a multi-jet vertical shaft Pelton published in 1937 [25].
and multi-jet arrangement of the turbine for vertical shaft opera-
tion (Fig. 5).
The Layout of a typical modern Pelton turbine is provided in
Fig. 6 [26]. It can be seen that the bucket shape has evolved into
much smoother round shape to reduce the flow losses inside of it.
Moreover, the buckets have a cutout that ensures better transition
as the jet goes from one bucket to another. In addition to that, the
injector contains nozzle and a spear valve to control the flow rate
and maintain good quality of the jet.
Before CFD was applied on impulse turbines a graphical
method [27,28] was used to analyse and develop Pelton turbines.
By the end of the XX century, first numerical results of Pelton jet
simulation were published [29–32]. These were followed by more
publications on numerical modelling of the jet [19,33].
Few years after presenting the first free jet simulations, pub- Fig. 6. Typical layout of modern Pelton turbine [26].
lication on numerically modelled jet and bucket interaction
appeared [34–36]. The jet and bucket interaction is the key part
of Pelton turbine as this is where the kinetic energy of the free jet In 2012, Solemslie and Dahlhaug published a paper [37]
is converted into momentum on the runner. Even though these regarding this absence of publically available numerically based
results looked promising and suggested that CFD is finally getting and experimentally validated Pelton optimisation results. The
to the level where it could be used for analysis and development of intensions of the on-going PhD research at Norwegian University
Pelton turbines, issues with accuracy or computational cost have of Science and Technology expressed in that paper were to fill this
caused the lack of publications regarding the design optimisation gap in the public knowledge. However the outcome of this project
based on numerical results. is still unknown.

3.2. Available design guidelines

As discussed in the previous section, huge amount of research

Fig. 4. Sketch of a nozzle controlled by a spear valve published in 1911 [24].
was performed on Pelton turbine design throughout the years
providing the hydro community with some guidelines or best
practice recommendations for Pelton turbines. This section
reviews the main design guidelines known up to date and the
most know authors in the area.
Probably the best known book containing design guidelines for
Pelton turbines is Hydraulic Turbines: Their Design and Equipment
by M. Nechleba [12]. However, looking at modern commercial
turbines it is obvious that some of the designing trends have
evolved since then. Nevertheless, theoretical calculations or equa-
tions for analysis of Pelton turbine performance provided in that
book are still applicable. More recent textbooks that include
design guidelines for Pelton turbines are Water Power Development
 A. Židonis, G.A. Aggidis / Renewable and Sustainable Energy Reviews 45 (2015) 135–144
by E. Mosonyi [13], MHPG Series: Harnessing Water Power on a
Small Scale. Volume 9: Micro Pelton Turbines by M. Eisenring [38] or
The Micro-Hydro Pelton Turbine Manual by J. Thake [39].
In addition to these textbooks, there are some publications that
include design guidelines for Pelton turbines [40,41]. However,
usually it is not known what kind of research these guidelines are
based on as most of the experimental data is not available to the
public and is kept as commercial secret by the turbine manufac-
turers [37]. Nevertheless, some of the guidelines are based on
theoretical calculations and assumptions and are usually in agree-
ment with the commercial product designs of the leading turbine
producers.
4. Computational fluid dynamics
Computational Fluid Dynamics is a part of larger subject called
fluid mechanics and is a very powerful tool for detailed flow
analysis. CFD uses algorithms and numerical techniques to solve
and investigate problems that include fluid flow. Computers are
used to numerically solve the governing equations that cannot be
solved analytically. CFD allows a simulation of fluids, gasses and
solids interaction within a domain enclosed by boundary condi-
tions. The main limitation when using CFD is the computational
cost usually requiring a compromise between the accuracy and the
timescale.
There are two main ways to mathematically specify the flow
field: Eulerian and Lagrangian [42–44]. In CFD, the Eulerian
methods solve the governing equations at the fixed positions in
the domain, hence might require very fine meshes at regions
where variables have high gradients, whereas Lagrangian methods
follow the moving fluid particles. The two following sections will
introduce the reader to Eulerian and Lagrangian codes that are
available and then a review of application of these codes on Pelton
turbines will follow.
4.1. Governing equations
CFD is based on governing equations that describe three
fundamental principles of physics:
1) Mass Conservation
2) Energy Equation or 1st Law of Thermodynamics
3) Newton's Second Law
These equations were developed many years ago and are valid
everywhere in the flow field of the flow continuum. However, no
generalised analytical solution has been developed to solve these
equations. Solutions are available for only a limited number of
simplified flow geometries. Therefore CFD, a numerical technique
that approximates the partial differential equations by replacing
them with discretised algebraic linear equations, is used. These
linear equations are then numerically solved at the discrete points
in space and/or time.
4.2. Eulerian and Lagrangian methods
In the Eulerian specification of the flow field the discretised
space is called the grid or the mesh. The values at the non-mesh
points are acquired by using interpolation schemes. From physical
point of view, the flow is observed at fixed positions and calcula-
tions are performed at these positions in discrete timesteps. Being
a discrete technique, the accuracy of CFD results highly depend on
the level of space and time discretisation or in other words the
mesh density and the timestep size. On the other hand, computa-
tional cost is also dependent on this level of discretisation. There
139
are plenty of different commercial or open source CFD codes
developed but the most popular Eulerian CFD codes used in turbo
machinery are ANSYS CFX, ANSYS Fluent and OpenFOAM. Work
performed on Pelton turbines using these codes will be reviewed
in the following section.
Lagrangian specification of the flow field is an alternative way
of solving the governing equations. The flow is observed by
following of individual particles and tracking their trajectories.
This method is especially attractive when modelling impulse
turbines as it does not require the usage of a mesh which typically
has to be very fine to capture the free surface jets and predict their
interaction with the runner. Lagrangian methods are usually much
faster but they are less developed therefore less accurate than
Eulerian CFD codes. The most popular Lagrangian techniques
applied in modelling of impulse turbines are SPH, FLS and MPS.
The application of these codes on modelling of Pelton turbines will
be reviewed later in this article.
5. Application of CFD on Pelton turbines
5.1. ANSYS Products (CFX and Fluent)
By far the largest amount of publications on modelling of
Pelton turbines use commercial codes developed by ANSYS: CFX
and Fluent. The capability of solving complex impulse turbine
related problems that include multiphase with free surfaces has
been demonstrated by number of previous studies [34,45–48].
There are some fundamental differences between these two
codes:
1) CFX uses cell vertex numerics (finite elements) whereas Fluent
uses cell-centred numerics (finite volumes) to discretise the
domain.
2) CFX focuses on coupled algebraic multi grid approach to solve
the governing equations of motion, Fluent offers more flex-
ibility in choosing one of the three methods: density based,
segregated pressure based and coupled pressure based.
3) CFX has two multiphase models applicable for free surface
simulations known as homogeneous and inhomogeneous indi-
cating whether the velocity field of different phases is shared or
separate respectively. Fluent uses Volume of Fluid (VOF)
method for problems with free surface.
However, even if these differences are known, it is still quite
difficult to say which code is more suitable for modelling of Pelton
turbines. There is a study [49] that compares numerical results
acquired with Fluent and CFX to experimental data on a fixed
Pelton bucket case. The study shows good consistency in terms of
pressure distribution and the location of air-water interface.
However, the comparison is not direct as the author had to use
fairly different settings to get the convergence. Moreover, no
transient rotating runner simulations were performed therefore
there is no information on the compared ability of Fluent or CFX to
assess the hydraulic efficiency. Therefore, due to the lack of
comparison between these two codes and their similarity they
are reviewed together. The studies performed can be grouped
depending on the complexity as the ones that model only the
bifurcation, injector and/or the jet (5.1.1 Injector and Jet Simula-
tions), ones that model jet interaction with a stationary bucket
(5.1.2 Stationary Bucket Simulations) and finally the most complex
ones that model jet interaction with full or part of the runner in
rotation (5.1.3 Rotating Bucket Simulations).
140 A. Židonis, G.A. Aggidis / Renewable and Sustainable Energy Reviews 45 (2015) 135–144
5.1.1. Injector and jet simulations
A numerical analysis of free surface jet flow was carried out by
Sulzer Ltd. together with VA Tech Escher Wyss Ltd. and published
by Muggli et al. [30]. Simulations were carried out using CFX-4 and
Flow 3D (another commercial code that is not commonly used for
impulse turbines in more recent publications). The velocity pro-
files were compared to measurements acquired using LDA. Velo-
city profiles modelled with CFX corresponded very well to the
measured ones whereas the profiles from Flow 3D simulations did
not completely match the measurements. Turbulence model used
for CFX simulations was k-ε and the multiphase model was
Homogeneous.
ČKD Blansko Engineering has published a case study [50] using
Fluent to improve the injector design and increase the overall
efficiency. Turbulence was modelled using k-ε RNG method and
multiphase was modelled using Volume of Fluid method. The
overall efficiency was successfully increased, however the exact
contribution of the new nozzle design is not clear as the runner
was also modified.
A publication from Vienna University of Technology [51] pre-
sents work on the free jet impinging a flat plate and then proceeds
to modelling of the free jet impinging on a runner made of flat
plates. The code used was Fluent 6.1.18, chosen turbulence model
was k-ε and the multiphase was modelled using Volume of Fluid
method.
Hochschule Lucerne and the Ecole Centrale de Lyon in colla-
boration with ANDRITZ Hydro performed a research on jets of
Pelton turbines [48]. The flow in a branchpipe, nozzles and a jet
was modelled with CFX using the k-ω Shear Stress Transport (SST)
turbulence model. A multiphase free-surface boundary was simu-
lated using Homogeneous model. The magnitudes of secondary
(tangential) flows observed were up to 3% of the axial flow velocity
magnitude. A clear relationship between the efficiency of the
runner and the jet dispersion caused by upstream bends
was found.
5.1.2. Stationary bucket simulations
Results of modelling the flow inside a stationary bucket
performed with Fluent were published by Zoppé et al. [52]. The
VOF multiphase model was used for simulating the free surface
and the k-ε model was used for turbulence. The work consisted of
different variations of head, jet incidence and flow rate and
reading pressure and torque as well as visualising the flow.
Approximately 300,000 cells used for half-bucket geometry have
provided stable results. The numerical results have an excellent
agreement with experimental results in most cases except for the
flow rate loss through the cutout. The numerical results under-
estimate this loss.
A master thesis from the Norwegian University of Science and
Technology [49] analyse the interaction between a jet and a
stationary bucket. Both CFX and Fluent are used and the numerical
results are compared with experimental data in terms of water
film thickness, visual observation, static pressure at the wall and
total pressure sweeps. The numerical results showed excellent
agreement between the codes even if the problem was solved in
steady state with CFX and transient with Fluent. However, there
were some discrepancies between experimental and numerical
results. Author suggests that this disagreement might be caused by
false assumption that the jet is axisymmetric.
5.1.3. Rotating bucket simulations
A research on flow in a rotating Pelton turbine bucket was
performed by Perrig et al. [53] which was a collaboration between
Ecole Polytechnique Fédérale de Lausanne (EPFL) and Vatech
Hydro Ltd. The aim of this research was to model an ideal jet
striking a rotating runner and compare the results to the experi-
mental data. The simulation was performed using CFX-5 code and
validated by pressure readings taken on one bucket which was
instrumented with five pressure taps. Turbulence was modelled
using the k-ε model and the multiphase flow was modelled using
the Homogeneous model. The main findings of this research said
that k-ε model had difficulties when modelling flow near the cut-
out region. Moreover, the simulations showed a presence of the
Coanda Effect [54], i.e. back of the bucket contributing to the
torque as the passing jet is attracted to the surface. However, this
observation could not be validated experimentally as the pressure
was not measured on the backside of the bucket. Nevertheless, the
error of the simulation was said to be around 5%, whereas the
modelled backside contribution was around 7.5%.
Results of collaboration between VATECH HYDRO Ltd. and
ANSYS are presented by Parkinson et al. [36]. It is not specified
what code or settings of the simulation were used. However,
capabilities of modelling of full runner with multi jet operation
using supercomputer were presented. In addition to that, results of
Finite Element Analysis (FEA) of stress coupled with CFD are also
present. Finally, experimental and numerical curves of local
pressure readings on the bucket indicate the outside of the bucket
contributing to the torque for some period of the duty cycle as
observed by Perrig et al. [53] and discussed in the previous
paragraph.
PhD thesis from the EPFL [55] presents an extensive study on
modelling of the Pelton turbine and comparing the results to
experimental data. Symmetry is assumed to reduce the computa-
tional domain and only seven buckets are modelled to reach
periodic behaviour. The jet was assumed to be ideal. The simula-
tions were performed with the CFX-10 commercial code using k-ω
SST model for turbulence and the 2-Phase Homogeneous model
for multiphase. The numerical results were compared with experi-
mental data in terms of pressure distribution, water film thickness
and flow patterns in the buckets. The Coanda effect on the
backside of the bucket was identified experimentally. This effect
is causing a depression as the jet separation occurs at the
beginning of the duty cycle. CFD results were able to capture this
effect as well; however there were issues with the accuracy in
terms of amplitude.
Another research simulating a performance of a Pelton runner
with a real jet was performed in University of Padova, Italy [47].
The software used for simulations was CFX-11 code. The chosen
turbulence model was the k-ω SST. After the initial comparison of
homogeneous and inhomogeneous models for multi-phase flow
the Inhomogeneous model was chosen despite the slightly higher
computation demands and publications of Janetzky et al. [45],
Kvicinsky [34] and Zoppé et al. [52] highlighting a satisfactorily
agreement between numerical and experimental results with
Homogeneous model. The decision was justified by the fact that
Inhomogeneous model showed higher stability. The results high-
light an importance of accurate modelling of the real jet, a
possibility of poor bucket shape designs inducing breaking torque
and presence of cutout leakage. Numerical and experimental
results were compared with numerically modelled results over
predicting the efficiency by 6%.
A research project performed by Turboinštitut, Slovenia [56,57]
produced results that were acquired using CFX-12.1. A two-jet
horizontal axis Pelton turbine was modelled at several operating
points. The turbulence was modelled using the k-ω SST model and
the free surface was modelled using the two-phase Homogeneous
model. Half of the runner geometry was meshed and the mesh
consisted of 25.7 million elements and 11 million nodes. The time
required for calculating performance of the turbine at one point
was five days on a supercomputer with 2048 processor cores.
Hence the main limitation of the simulation was the
Table 1
Comparison of CFD methods and software used for Pelton turbine modelling.

Code description Problems analysed (Number of publications) Capabilities Usage in the field Comment

Flow field Name of Commercial Injector and Stationary Rotating Reliable Computational References Most recent
specification code Jet bucket bucket accuracya cost publication

A. Židonis, G.A. Aggidis / Renewable and Sustainable Energy Reviews 45 (2015) 135–144
Eulerian CFX Yes 2 1 7 Yes High [30,47–49,53,55–59] 2014  Good accuracya
(mesh based)  Suitable for optimisation
 Most widely used for
rotating
bucket (runner)
simulations
 Good accuracy
Fluent Yes 2 2 1 Yes High [49–52,59] 2014  Suitable for optimisation
a

 Open source
OpenFOAM No – – 1 No High [61] 2013  No successful attempts
to model
the Pelton turbine

 No recent publications
Flow-3D Yes 1 – – No High [30] 2000 on Pelton simulations

 Open source
Lagrangian SPH No – – 3 No Moderate [62–64] 2013  Useful for initial
(mesh free) design stage

 Useful for initial

FLS No – – 3 No Low [59,67,69] 2014 design stage

 No recent publications
MPS No – 1 – No Low [70] 2009 on Pelton simulations

a
this assessment of accuracy is concentrating on modelling of Pelton turbines and is based on the findings of available publications.

141
142 A. Židonis, G.A. Aggidis / Renewable and Sustainable Energy Reviews 45 (2015) 135–144
computational cost. The results of the research show that accurate
simulation of the jet and the mesh density were the main factors
affecting the accuracy. Results were compared to the experimental
data and the CFD simulation showed 4% lower power output than
the measured one.
A master thesis [58] from the Norwegian University of Science
and Technology (NTNU) presents the analysis of Pelton turbine
using CFX 13.0. Author models three consecutive half buckets to
represent the torque. The turbulence model used was k-ω SST,
multiphase model was Homogeneous. The comparison between
numerical and experimental results showed that CFD over pre-
dicted the efficiency by approximately 1.5%.
A collaborative study [59] between the National Technical
University of Athens (NTUA) and the Lancaster University provided
two case studies where different Pelton runners were parametri-
cally optimised by performing a numerical study using Fluent 14.0
and CFX 14.0. The turbulence model used was k-ω SST. In Fluent
VOF multiphase model was used and in CFX Homogeneous model
was chosen for multiphase. The study looked at 11 design para-
meters and compared their influence on efficiency for two Pelton
runners of different specific speed. The numerical models were
validated against initial experimental data of the original designs.
Experimental testing of the optimised designs is reported as the
next step.
5.2. OpenFOAM
There are attempts to model Pelton turbines using open source
CFD codes as an alternative to commercial CFD packages such as
ANSYS CFX or ANSYS Fluent. OpenFOAM [60] is one of the most
popular mesh based (Eulerian) open source codes used by the CFD
community. In some cases OpenFOAM is able to compete quite
well with the commercial packages. However, the publications on
modelling of the Pelton turbine using OpenFOAM suggest that it
needs to be further developed to become an alternative to ANSYS
products. The published work [61] shows that OpenFOAM was
over predicting the torque on the bucket as compared to the
results acquired experimentally or numerically using CFX. More-
over the torque curve contained instabilities and did not coincide
with the curve modelled in CFX. Finally, the required simulation
time reported was almost 30 times larger than what was required
by CFX.
5.3. Other Eulerian solvers
In addition to Eulerian codes such as CFX, Fluent or OpenFOAM
there were only few attempts to model Pelton turbines using
alternative Eulerian solvers [30]. There are more attempts to create
alternative numerical methods that could be satisfactorily applied
on Pelton turbines using Lagrangian flow specification as this
would reduce the simulation time drastically and this area is less
researched. The development of these Lagrangian techniques is
presented in the following sections.
5.4. Smoothed particle hydrodynamics
Smoothed Particle Hydrodynamics (SPH) is a mesh free
(Lagrangian) solver represented by the SPH European Research
Interest Community (SPHERIC). Being a Lagrangian technique SPH
is a promising tool for free surface, multiphase flow modelling as it
does not require any mesh. However, even the latest publications
on modelling of a Pelton runner using SPH express concerns on
the accuracy of the code [62,63] or suggest that the method needs
some further validation and development [64]. There is a publica-
tion on modelling of Turgo, another impulse turbine, using SPH
claiming that it has produced similar results to Fluent in much less
time and therefore is a good alternative to Eulerian methods [65].
However, looking at the quality of the results in the comparison of
the torque curves one can clearly see that SPH produces unwanted
oscillations. In addition to that, authors do not provide any
experimental validation of these results. Finally, the blade geome-
try is represented as a surface rather than a solid geometry
suggesting that this method was ignoring the negative pressure
on the outside of the blade as it was entering the jet. This
phenomenon is quite significant in Pelton turbines as discussed
before in more detail in Section 5.1 ANSYS Products (CFX and
Fluent).
5.5. Fast Lagrangian solver
Fast Lagrangian Solver (FLS) is another very interesting and
promising mesh free modelling technique which was developed at
the National Technical University of Athens to numerically develop
and optimise impulse turbines at minimal computer cost [66]. The
main advantage of this technique is its speed making the time
duration of the performance simulation almost negligible [67].
Therefore this Lagrangian approach can be a very powerful tool for
parametric optimisation if coupled with some type of stochastic
optimisation software [68]. However, the disadvantages of this
technique are its accuracy and the need for some constants in the
particle motion equation to be tuned depending on the applica-
tion. Therefore, it cannot be used as a standalone CFD code
without having experimental or numerical results from more
accurate CFD solvers [69]. In addition to that, the main limitations
of this approach is similar to the ones of SPH where only the inside
surface of the bucket is modelled and no interaction between the
buckets or negative pulling pressure on the outside of the bucket
is taken into account. Therefore, according to the study done in
collaboration with the developers of this code, FLS is very useful
for early design stage when designing a turbine from scratch but
when further developing or optimising a turbine of an already
good performance, more accurate CFD methods should be used
[59].
5.6. Other Lagrangian solvers
Moving Particle Semi-implicit (MPS) method [70] is another
Lagrangian method that could be used to model Pelton turbine
performance. However only stationary bucket is modelled in this
publication making it difficult to judge how accurate and fast it is
when modelling a rotating runner.
6. Discussion and suggestions
Numerical modelling of Pelton turbine performance is a chal-
lenging task. That is why there is no single best software, code or
methodology for it. The current worldwide research is developing
in few directions as can be seen from the previous section. Table 1
provides a detailed comparison of CFD codes that are applied on
Pelton turbine modelling together with the references to publica-
tions of each code being used to simulate flow in different parts of
the turbine at different level of complexity. This table is a
summary and is intended to be the guidance for the researchers
or engineers to decide which code is the most suitable for their
needs and what has been achieved so far.
Based on the number of publications stating successful simula-
tion of the most complex case, the rotating bucket simulation, CFX
is the code for the most accurate and stable results. However CFX
as well as another commercial software Fluent requires high
computational resources and is not open source. Therefore
researchers with limited budgets or those interested in code
 A. Židonis, G.A. Aggidis / Renewable and Sustainable Energy Reviews 45 (2015) 135–144
development might try alternative codes such as OpenFOAM, SPH
or FLS. OpenFOAM is the closest option to CFX or Fluent as it is also
using the Eulerian (mesh based) flow field specification. However
it is not quite there yet in terms of modelling of Pelton turbine and
acquiring stable, accurate results within reasonable timescales.
SPH or FLS, on the other hand, have their specific niche in the
world of CFD and especially in CFD application on modelling of
Pelton turbines. Due to Lagrangian (mesh free) specification of the
flow field, SPH and FLS are much faster but limited in accuracy up
to date. These Lagrangian methods are very attractive in the early
design stage when large amount of design options have to be
evaluated. Usually the early design stage requires less accuracy. For
the final evaluation of the best design candidates it is suggested to
use CFX based on the success of various independent researches.
The most recent publications that show successful simulation
of Pelton turbine where the results were stable and accurate used
k-ω SST turbulence model and Homogeneous of Volume of Fluid
multiphase model. Gravity and surface tension were usually
assumed as negligible and simplifications of the runner geometry
were made to reduce the computational cost. These simplifications
included the introduction of the symmetry plane and modelling of
only few buckets of the runner.
7. Conclusions
The development of Pelton turbines since its invention up to
modern days was reviewed in this article. Difficulties in applying
of Computational Fluid Dynamics (CFD) to analyse and improve
the performance of Pelton turbine were presented and the state of
the art in this area was established. To summarise, from the
number of available publications it is clear that nowadays ANSYS
CFX code is the most widely used CFD tool for accurately simulat-
ing the jet interaction with rotating runner. Lagrangian particle
tracking methods are promising and attractive due to much lower
computational cost but need to be further developed to be able to
compete with commercial Eulerian codes like CFX. Finally, the
modelling techniques and computational resources have reached
the state where numerical design optimisation of Pelton turbine is
becoming possible.
Acknowledgments
The authors would like to thank Lancaster University Renew-
able Energy Group and Fluid Machinery Group.
References
[1] European Renewable Energy Council. Renewable energy in Europe: markets,
trends, and technologies. 2nd ed. United Kingdom: Earthscan; 2010.
[2] European Commission. Directive 2009/28/EC of the European parliament and
of the council of 23 April 2009 on the promotion of the use of energy from
renewable sources and amending and subsequently repealing Directives 2001/
77/EC and 2003/30. Off J Eur Union Belg 2009.
[3] United Nations. Framework convention on climate change. Fact sheet: the
Kyoto protocol, 〈http://unfccc.int/files/press/backgrounders/application/pdf/
fact_sheet_the_kyoto_protocol.pdf〉; 2011 [accessed 19.09.14].
[4] National Grid. Data explorer, 〈http://www2.nationalgrid.com/uk/Industry-in
formation/Electricity-transmission-operational-data/Data-Explorer/〉; 2014
[accessed 14.08.14].
[5] British Hydropower Association. Hydropower: opportunities, challenges &
sharing best practice. The British hydropower association perspective, 〈http://
www.ciwem.org/media/142437/The%20British%20Hydropower%20Associa
tion%20Perspective.pdf〉; 2010 [accessed 18.09.14].
[6] Turner DS. Applied earth science. Dubuque, Iowa: W.C. Brown Co.; 1969.
[7] Pidwirny M. The hydrologic cycle. Fundam Phys Geogr 2006 (2nd ed).
[8] Mosonyi E. Water power development: low-head power plants. 3rd ed. .
Budapest: Akademiai Kiado; 1987.
143
[9] Aggidis GA, Luchinskaya E, Rothschild R, Howard D. The costs of small-scale
hydro power production: Impact on the development of existing potential.
Renew Energy 2010;35:2632–8.
[10] Aggidis G. Performance Envelopes of Hydro Turbines, 〈http://www.engineer
ing.lancs.ac.uk/lureg/nwhrm/engineering/〉; 2010 [accessed 01.10.14].
[11] Pelton LA. US1880.
[12] Nechleba M. Hydraulic turbines: their design and equipment. Prague: Artia;
1957.
[13] Mosonyi E. Water power development: high-head power plants. 3rd ed.
Budapest: Akademiai Kiado; 1991.
[14] Patel K, Patel B, Yadav M, Foggia T. Development of Pelton turbine using
numerical simulation. Twenty fifth IAHR symposium on hydraulic machinery
and systems: IOP Conference. Ser: Earth Environ Sci 12 (2010) 012048; 2010.
[15] Aggidis GA, Židonis A. Hydro turbine prototype testing and generation of
performance curves: fully automated approach. Renew Energy.71:433-41.
[16] Keck H, Sick M. Thirty years of numerical flow simulation in hydraulic
turbomachines. Acta Mech 2008;201:211–29.
[17] Sick M, Keck H, Grunder R, Parkinson E, Sallaberger M. Recent examples of
modern CFD - Methods and their application in practical turbine designs.
Twenty third CADFEM Users' meeting 2005 international congress on FEM
technology with ANSYS CFX & ICEM CFD Conference2005.
[18] Staubli TAA, Weibel P, Bissel C, Parkinson E, Leduc J, Leboeuf F. Jet quality and
Pelton efficiency. Hydro 2009. Lyon2009.
[19] Peron M, Parkinson E, Geppert L, Staubli T. Importance of jet quality on Pelton
efficiency and cavitation. IGHEM2008. Italy: Milan; 2008.
[20] Marongiu JC, Maruzewski P, Parkinson E. Modelling the flow in a Pelton
turbine. Manchester, UK: Spheric newsletter; 2005. p. 8–9.
[21] Staubli T, Weibel P, Bissel C, Karakolcu A, Bleiker U. Efficiency increase by jet
quality improvement and reduction of splashing water in the casing of Pelton
turbines. Sixteenth international seminar on hydropower plants 2010.
[22] Pelton Water Wheel Company. The Pelton water wheel… Embracing in its
variations of construction and application the Pelton system of power. San
Francisco 1898.
[23] Davidson SC,. Impact-wheel or turbine. In: Office USPaT, editor. Washington,
DC: Google Patents; 1900.
[24] Prášil F. Results of experiments with Francis turbines and tangential (Pelton)
turbines. Proc Inst Mech Eng 1911;81:647–79.
[25] Fulton A. Present tendencies in water turbine machinery. Proc Inst Mech Eng
1937;135:387–444.
[26] Ecopolis. Pelton turbine HHP - H. 〈http://www.ecopolisla.com/en/hydro-
power/products/turbines/detail-pelton-turbine-hhp—h-50/〉; 2010 [accessed
22.09.14].
[27] Brekke H. A general study on the design of vertical Pelton turbines. Ljubljana:
Turboinstitut; 1984.
[28] Hana M. Numerical analysis of non-stationary free surface flow in a Pelton
bucket. Trondheim: Norwegian University of Science and Technology; 1999.
[29] Avellan F, Dupont P, Kvicinsky S, Chapuis L, Parkinson E, Vullioud G., Flow
calculations in Pelton turbines,-Part 2: free surface flows. Proceedings of
theninteenth IAHR symposium, Singapore, Republic of Singapore1998.
p. 294–305.
[30] Muggli F, Zhang Z, Schärer C, Geppert L. Numerical and experimental analysis
of Pelton turbine flow. Part 2: the free surface jet flow. Tweentyth IAHR
symposium 2000.
[31] Sick M, Keck H, Parkinson E, Vullioud G. New challenges in pelton research.
Hydro 2000 Conference. Bern 2000.
[32] Sick M, Schindler M, Drtina P, Schärer C, Keck H. Numerical and experimental
analysis of Pelton turbine flow. Part 1: distributor and injector. XX IAHR
symposium, Charlotte 2000.
[33] Parkinson E, Garcin H, Vullioud G, Muggli F, Zhang Z, Casartelli E. Experi-
mental and numerical investigations of the free jet flow at a model nozzle of a
Pelton turbine. Proceedings of the XXI IAHR symposium on hydraulic
machinery and systems. Lausanne 2002.
[34] Kvicinsky S, Kueny J-L, Avellan Fo, Parkinson E. Experimental and numerical
analysis or free surface flows in a rotating bucket. The proceeding of the
twenty first IAHR symposium on hydraulic machinery and systems 2002. p.
359-64.
[35] Parkinson E, Vullioud G, Geppert L, Keck H. Analysis of Pelton turbine flow
patterns for improved runner-component interaction. Int J Hydropower Dams
2002;9:100–3.
[36] Parkinson E, Neury C, Garcin H, Vullioud G, Weiss T. Unsteady analysis of a
Pelton runner with flow and mechanical simulations. Int J Hydropower Dams
2006;13:101–5.
[37] Solemslie B, Dahlhaug O. A reference Pelton turbine design. IOP conference
series: earth and environmental science. Beijing, China: IOP Publishing; 2012.
p. 032005.
[38] Eisenring M. Micro pelton turbines. St. Gallen: Swiss Center for Appropriate
Technology; 1991.
[39] Thake J. The micro-hydro Pelton turbine manual: design, manufacture and
installation for small-scale hydro-power. London: ITDG publishing; 2000.
[40] Atthanayake IU. Analytical study on flow through a Pelton turbine bucket
using boundary layer theory. Int J Eng Technol 2009;9:241–5.
[41] Nasir BA. Design of high efficiency Pelton turbine for micro-hydropower plant.
Int J Electr Eng Technol (IJEET) 2013;4:171–83.
[42] Batchelor GK. An introduction to fluid dynamics. Cambridge University Press;
1973.
[43] Lamb H. Hydrodynamics. Cambridge, UK: Cambridge University Press; 1994.
144 A. Židonis, G.A. Aggidis / Renewable and Sustainable Energy Reviews 45 (2015) 135–144
[44] Munson BR, Young DF, Okiishi TH. Fundamentals of fluid mechanics. 5th ed.
John Willey & Sons, Inc.; 2005.
[45] Janetzky B, Göde E, Ruprecht A, Keck H, Schärer C. Numerical simulation of
the flow in a Pelton bucket. Proceedings of ninteenth IAHR Symposium 1998.
p. 276-83.
[46] Favre J, Garcin H, Parkinson E. Computational analysis in Pelton hydraulic
turbines. Twenty third CADFEM users meeting 2005.
[47] Santolin A, Cavazzini G, Ardizzon G, Pavesi G. Numerical investigation of the
interaction between jet and bucket in a Pelton turbine. Proc Inst Mech Eng
Part A: J Power Energy 2009;223:721–8.
[48] Staubli T, Abgottspon A, Weibel P, Bissel C, Parkinson E, Leduc J, et al. Jet
quality and Pelton efficiency. Proc Hydro 2009 Prog-Potential-Plans 2009.
[49] Klemensten LA. An experimental and numerical study of the free surface
Pelton bucket flow. Trondheim: Norwegian University of Science and Tech-
nology; 2010.
[50] Veselý JV, M. A case study of upgrading 62.5 MW Pelton turbine. Proceedings
of international conference: IAHR 2001. Prague 2001.
[51] Matthias HB, Promper O. Numerical simulation of the free surface flow in
Pelton turbines. The Sixth international conference on hydraulic machinery
and hydrodynamics. Timisoara 2004.
[52] Zoppe B, Pellone C, MaÃstre T, Leroy P. Flow analysis inside a Pelton turbine
bucket. J Turbomach 2006;128:500–11.
[53] Perrig A, Avellan F, Kueny J-L, Farhat M, Parkinson E. Flow in a Pelton turbine
bucket: numerical and experimental investigations. J Fluids Eng 2006;128:
350–8.
[54] Taylor MJH. Jane's encyclopedia of aviation. Grolier Publishing, US: Grolier
Educational Corp.; 1980.
[55] Perrig A. Hydrodynamics of the free surface low in Pelton turbine buckets:
École polytechnique fédérale de Lausanne: Lausanne, France; 2007.
[56] Jošt D, Lipej A, Meznar P. Numerical prediction of efficiency, cavitation and
unsteady phenomena in water turbines. ASME 2008 nineth biennial con-
ference on engineering systems design and analysis: American society of
mechanical engineers; 2008. p. 157-66.
[57] Jošt D, Mežnar P, Lipej A. Numerical prediction of Pelton turbine efficiency.
Twenty fifth IAHR symposium on hydraulic machinery and systems: IOP
Conference. Ser: Earth Environ Sci 12 (2010) 012080; 2010.
[58] Barstad LF. CFD Analysis of a Pelton turbine. Trondheim: Norwegian University
of Science and Technology; 2012.
[59] Židonis A, Panagiotopoulos A, Aggidis GA, Anagnostopoulos JS, Papantonis DE.
Parametric optimisation of two Pelton turbine runner designs using CFD. J
Hydrodyn Ser B 2014.
[60] OpenFOAM Foundation. OpenFOAM User Guide, Version 2.1.1. 〈http://www.
openfoam.org/version2.1.1/〉; 2012 [accessed 19.09.14].
[61] Rygg JR. CFD analysis of a Pelton turbine in openFOAM. Trondheim: Norwe-
gian University of Science and Technology; 2013.
[62] Marongiu JC, Leboeuf F, Caro JË, Parkinson E. Free surface flows simulations in
Pelton turbines using an hybrid SPH-ALE method. J Hydraul Res 48:40-49.
[63] Furnes K. Flow in Pelton turbines. Trondheim: Norwegian University of
Science and Technology; 2013.
[64] Koukouvinis PK, Anagnostopoulos JS, Papantonis DE. Flow analysis inside a
Pelton turbine bucket using smoothed particle hydrodynamics HYDRO inter-
national conerence 2010.
[65] Koukouvinis PK, Anagnostopoulos JS, Papantonis DE. SPH method used for
flow predictions at a Turgo impulse turbine: comparison with fluent. World
academy of science, Eng Technol 55 20112011.
[66] Anagnostopoulos JS, Koukouvinis PK, Stamatelos FG, Papantonis DE. Optimal
design and experimental validation of a Turgo model Hydro turbine. ASME
2012 eleventh biennial conference on engineering systems design and
analysis: American society of mechanical engineers; 2012. p. 157-66.
[67] Anagnostopoulos JS, Papantonis DE. A numerical methodology for design
optimization of Pelton turbine runners. . Hydro; 2006. p. 25–7.
[68] Anagnostopoulos JS, Papantonis DE. Flow modeling and runner design
optimization in Turgo water turbines. World Acad Sci Eng Technol
2007;28:206–11.
[69] Anagnostopoulos JS, Papantonis DE. A fast Lagrangian simulation method for
flow analysis and runner design in Pelton turbines. J Hydrodyn Ser B.
2012;24:930–41.
[70] Nakanishi Y, Fujii T, Kawaguchi S. Numerical and experimental investigations
of the flow in a stationary Pelton bucket. JFluid Sci Technol 2009;4:490–9.