Renewable Energy 95 (2016) 356e366

Contents lists available at ScienceDirect

Renewable Energy
journal homepage: www.elsevier.com/locate/renene

Experimental characterization of a ve blade tubular propeller turbine

for pipe inline installation
Irene Samora a, b, *, Vlad Hasmatuchi c, Ce

cile Mnch-Alligne

c, M a

rio J. Franca b,
Anton J. Schleiss b, Helena M. Ramos a
a
Civil Engineering Research and Innovation for Sustainability, Instituto Superior T

ecnico, Universidade de Lisboa, Lisboa, Portugal
b

Laboratory of Hydraulic Constructions, Ecole Polytechnique F
ed

erale de Lausanne, Lausanne, Switzerland
c
University of Applied Sciences and Arts Western Switzerland Valais, Engineering School, Sion, Switzerland

a r t i c l e i n f o a b s t r a c t

Article history: The interest in micro-hydropower in existing infrastructure is increasing since this is a technology with
Received 20 October 2015 low environmental impacts and potential for energy recovery in different types of installation. The
Received in revised form technologies available for these schemes are still restricted and it is a current subject for research. In this
31 March 2016
work an experimental characterization of an inline tubular propeller suitable for pressurized systems,
Accepted 12 April 2016
Available online 22 April 2016
such as water supply and distribution networks, is presented. In the framework of the European Project
HYLOW, started in 2008, a rst prototype has been developed using CFD analysis and tested. Never-
theless, optimization and validation were still necessary. An improved model has been now adequately
Keywords:
Micro-hydropower
tested in the laboratory and proves to be interesting in terms of potential application. The experimental
Turbine investigation evidenced that the new turbine has efciencies of around 60% for low-headed operations,
Experimental investigation below 50 m.
Performance measurements 2016 Elsevier Ltd. All rights reserved.
Energy production

1. Introduction to connect to the national grid, stand-alone small hydro-systems

The investment in small hydropower is growing all over the
world as it is a clean, sustainable and emissions-free source of
renewable energy. Small hydropower can be classied in mini,
micro or pico, depending on the output power and on the type of
the adopted scheme [25,26]. There are not yet globally accepted
boundaries to dene these classes, it depends on the country, but
micro-hydro typically refers to schemes below 100 kW [28] while
pico-hydro usually produces less than 5 kW [1].
A micro-hydropower scheme can generate energy from small
rivers, water supply systems, irrigation channels, wastewater
treatment plants and pluvial drainage systems. These micro
schemes can be used to produce energy to supply the national grid
or for local consumption. Consuming on site has the advantage of
reducing transmission losses, since energy transportation is an
inefcient energy conversion process [40]. Also, in situations such
as remote communities, where it is not economical or even possible

cnico, Universidade de Lisboa, Lis-
* Corresponding author. Instituto Superior Te
boa, Portugal.
E-mail address: irene.almeidasamora@ep.ch (I. Samora).
can be used to respond to the energy needs [17,18].
The turbine technology for micro-hydropower is still much
unexplored. Conventional turbines are often not cost-effective for
micro-hydro, mostly due to their large diameters and expensive
civil engineering works [15,41]. Many investigators have attempted
to use different turbines in the micro production range. However
their experience was not much encouraging as it resulted in
complicated arrangements and high costs of installation [17].
Consequently, special converters are needed for the exploitation of
this source of renewable energy [15].
There are several hydropower technologies currently being
employed in low-headed schemes. The most frequently used are
the Zuppinger water wheel, the Archimedes screw working in
reverse, the crossow turbine and pumps operating as turbines
(PAT).
The Zuppinger or undershot wheel was developed in the 1850's
and can be used for very small head differences (0.5e2.5 m) and
large ow volumes (0.5e0.95 m3/s/m). This wheel employs only
the potential energy of the ow as the principal driving force and
the water enters over a weir, so that the cells can be lled rapidly
[22]. Its peak efciency is 70e74%, and power ratings range from 3
to 100 kW. With diameters from 4 to 7.5 m and rotation speed from

http://dx.doi.org/10.1016/j.renene.2016.04.023
0960-1481/ 2016 Elsevier Ltd. All rights reserved.
 I. Samora et al. / Renewable Energy 95 (2016) 356e366
3 to 6 rpm, the wheels have a slow speed, which means that
expensive gearing is obligatory [16].
The Archimedes screw is a technology for pumping that exists
since antiquity. Its application in reverse operation has led to more
research on its efciency and optimal congurations. It consists of a
helical screw that can be applied on rivers with low heads [19,23].
Schleicher et al.[34] proposed a geometry of the blades optimized
by CFD analysis in which a peak efciency of 72% was reached.
A PAT is a hydraulic pump running in reverse mode so that, in
conjunction with an induction generator, it recovers energy. The
mass production of pumps and its smaller complexity makes them
a less costly solution [17]. Nevertheless, the main disadvantage of
using a PAT is the difculty of nding the characteristics needed to
select the correct pump for a particular site. Fixed ow rate PATs are
only suitable for sites where there is a sufcient supply of water
throughout the year [42] and there is still no standard design
criteria available for highly variable ow-rates and pressure heads
[5]. The efciency of a PAT is very sensitive to changes of ow,
which means that an inappropriate pump selection will result in an
undesired output, and ultimately in the failure of the project [28]. In
situations where multiple discharge values are plausible, it is also
possible to consider the use of two or more PATs [6,10].
Other than these well-known technologies, research has
recently been carried for the development of new turbines for low
heads. Some examples are herein presented.
The hydroelectricity group in University of Applied Sciences and
Arts Western Switzerland in Sion (HES-SO VS) and the Laboratory
of Hydraulic Machines (LMH) at the Ecole
Polytechnique Fe
de

rale
de Lausanne (EPFL), Switzerland, have designed and developed an
axial counter-rotating turbine to be installed inline pipes. Numer-
ical simulations were performed with ANSYS CFX and three pro-
totypes were built for laboratory testing. The rst turbine prototype
with a 50 mm radius was designed to recover 10 mWc on each
runner for 8.7 l/s at the best efciency point [24] and tested in an
elbow conguration at EPFL-LMH. A second prototype in a bulb
conguration has been developed and tested by the Hydroelec-
tricity group of the HES-SO VS. The bulb conguration is particu-
larly interesting as it allows encapsulating the two generators, one
for each runner, internally [21]; [4]. Now a multi-stage strao
design is under development to provide a family range of turbine
from 5 to 25 kW.
The free vortex propeller was designed by Singh and Nestmann
[38] to recover gross heads from 1.5 to 2 mWc for open channel
ows. The blades of the propeller were optimized using the free
vortex theory considering ows from 60 to 75 l/s. An experimental
test-rig was assembled and tests were satisfactory, resulting in
810 W of maximum power with 73.9% of efciency.
A exible foil vertical axis turbine has been studied by Zeiner-
Gundersen [43] for river, ocean, and tidal applications. Its exible
behavior was inspired on the dynamic characteristics of aquatic
animals. A model with 9 m of diameter was tested, with ve wings
connected to a 7 m tall central main shaft holding the blades. Its
best efciency was achieved for a velocity of 0.79 m/s and up to
17 kW were generated.
The EU-Project HYLOW (HYdropower converters with very LOW
head differences, 2008 to 2012), aimed studying new hydropower
converters for very low head differences. Within the scope of this
project, a few designs were proposed, among them:

The Rotary Hydraulic Pressure Machine (RHPM), developed in
the University of Southampton, in the UK. This machine has a
wheel with a diameter between 1.5 m and 7.5 m which rotates
about a horizontal. It was envisaged that the RHPM could be
employed with fall heights under 5 m in any conventional
diversion or run-of-river installation and would also be
357
particularly suited for installation into bays of existing weir
structures [35].

The Hydrostatic Pressure Machine (HPM), also developed in the
University of Southampton, for river applications with head
differences between 1 and 3 mWc. Prototypes of this converter
were investigated, including a full-scale one installed in a river
[44].

The Free Stream Energy Converter (FSEC), inspired by boat mills,
was tested at various scales at the University of Rostock, in
Germany. Its application is also envisaged for small rivers [3].

And a Five Blade Tubular Propeller (5BTP), developed in Insti-
tuto Superior Te
cnico (IST), in Portugal [15], which is the object
of the present work.
The presented state of the art reveals that research is being
carried to design new technologies for micro-hydro since the
traditional large-scale turbines are not easily applicable for such
low heads and ows. The presented ongoing research can be
divided in two main elds of application, for small rivers and for
pressurized systems. In the case of pressurized systems, crossow
turbines, although considered expensive, can work on a wide range
of net heads, from 1.75 to 200 mWc. PAT are less costly but they are
sensitive to ow variations. The development of a tubular turbine
that can be installed inline in pressurized pipe is thus considered
interesting.
The installation of turbines inline a pipe is particularly attractive
for applications in water supply systems. It has been veried that
the incorporation of turbines in these systems can contribute to
improve the energy efciency of the system and to solve problems
of leakage [7,10,20]. The optimization of this type of micro-
hydropower generation is currently a topic of research [7,33].
The goal present work focuses on the presentation of the results
of laboratorial experiments from the testing of the 5BTP. The
development of this turbine started during the HYLOW Project
with numerical simulations and a few primary experimental tests,
but laboratorial optimization and validation with wide ranges of
discharge and rotational speed were still needed. In Section 2 the
background development of the turbine is summarized. The
experimental methodology is presented in Section 3, followed by
the results in Section 4. In Section 5 the results are discussed and
the main conclusions are presented in Section 6.
2. Background and optimization of the 5BTP
The HYLOW Project (2008e2012) was a research project funded
by the European Commission's 7th Framework Program with the
aim of developing novel hydropower converters for very low heads.
IST, one of the partners, was involved with the study of small
converters for pressurized ows. Initially, turbines and positive
displacement concepts were investigated but it was found that
positive displacement converters are only suitable for vary small
discharges and high heads, leading to unfavorable pressure surges.
Therefore, the research on turbines was deepened further for tur-
bines, based on CFD modeling and laboratorial tests.
Propeller turbines are axial turbines that are usually adequate
for operation under low head and high ow rates [6,9]. They have
been mainly used in small and mini hydro schemes, but their
application for micro hydro is still at the beginning. Nevertheless,
scaling-down from a large turbine cannot be directly applied on
this case due to associated scale effects, as it has been veried by
Ramos et al. [29]; but also because the economic and
manufacturing constraints are not the same [38]. To overcome this
limitation, numerical simulations were done in the FLUENT model,
a Navier-Stokes solver, considering different turbine congurations
[28]. A k turbulence model was used and the meshes were
358 I. Samora et al. / Renewable Energy 95 (2016) 356e366
generated with a TGrid unstructured mesh technique with rene-
ment in areas of difcult modeling. TGrid requires a discretized
boundary mesh consisting of triangular and quadrilateral faces in
3D. Four types of boundary conditions were dened: inlet and
outlet pressure, impeller, dened as moveable walls with rotational
speed around the rotation shaft, and the tubular wall, with the
imposition of the condition of impermeability [30,32,36].
The initial geometry of the 5BTP is presented in Fig. 1 for the
case of a 200 mm impeller. It consisted of a runner with ve curved
blades and two bulbs. It was conceived to be installed in a 90
curved pipe, to allow the connection of a rotating axis with an
external generator. The geometry of the axial turbine, including the
blade slopes, was numerically investigated and optimized for pro-
totypes with impeller diameters of 100 mm and 200 mm and with a
maximum thickness of 1 mm [28,31].
The CFD analysis was performed for discharges between 5 and
20 l/s for the 100 mm impeller and for discharges between 70 and
155 l/s for the 200 mm impeller. Flow velocities and total pressures
were analyzed, as well as turbulence, including Reynolds shear
stress distribution [36]. Finally, theoretical characteristic and ef-
ciency curves were obtained, as presented in Fig. 2. It was
concluded that this turbine could be the most attractive among the
studied machines in IST during the program, since it showed to
have efciencies of 20e80% in CFD modeling.
To compare and validate the results obtained numerically, a few
laboratorial tests were performed in IST [30,32] using a prototype
with a diameter of the impeller of 100 mm developed in PVC (the
bulbs) and brass (the impeller) e Fig. 3. A 500 W DC permanent-
magnet machine, derived from an electrical scooter, was used to
simulate a generator and a 6 U rheostat was connected to force an
electrical load in the system [6].
The laboratorial experiments demonstrated that the 5BTP could
be economically interesting, as the behavior seemed to be consis-
tent with the CFD simulations [6,30,32]. Nevertheless, the instal-
lation was limited in terms of ow and electrical control, so it was
not possible to have complete hill diagrams. Also, the monitory of
the tests was not sufciently reliable to provide denite results
regarding the turbine performance.
To have full hill diagrams and reliable empirical data of the
turbine performance, a new experimental installation was pre-
pared at the HES-SO Valais Laboratory, which is specialized in small
turbine testing. A new conguration of the hydraulic circuit was
adopted. The turbine has been placed upstream a curved (45 ) pipe
to minimize ow head losses.
Fig. 1. Geometry of the 5BTP for 200 mm of diameter and its disposition inline the pipe.
3. Methods
3.1. Experimental setup and instrumentation
A new model of the 5BTP was tested in the facilities of HES-SO
Valais, in a universal test rig designed to assess the hydraulic per-
formance of different types of small-power turbomachines (up to
10 kW) and other hydraulic components [12,13]. With the new
installation, a number of changes in the conguration of the ma-
chine were introduced for optimization of the turbine performance
and for adaptation to the existing setup, making it distinct from the
previous one in IST.
Firstly, the nominal diameter of the turbine model, previously
planned for 100 mm as at IST, was reduced since the maximum
available ow-rate in the installation (50 m3/h), not enough to
cover the whole range of application that was veried numerically
to be interesting. The new model was hence designed for 85 mm of
diameter.
The distance between the blades and the pipe walls was reduced
to 0.1 mm in order to minimize the leak losses and to allow a better
comparison with the CFD model. Indeed, the latest did not consider
any tip clearance thus the totality of the ow participating to the
energy conversion in the system.
Changes were also applied to the shape of the turbine (Figs. 4
and 5). The downstream bulb disappeared, the diameter of the
hub being maintained up to the exit from the elbow. The hub
houses the ceramic bearings and the main shaft of the turbine. The
upstream bulb was elongated to obtain straighter velocity
streamlines approaching the runner, which has the second
advantage of providing a better reading in the pressure transducer.
The runner and the bulb were produced separately in different
materials, the runner in bronze, while the bulb in black POM
(polyoxymethylene). The thickness of the blade of 1.7 mm was
chosen after an analysis of the structural resistance, corresponding
to a safety factor of 3 considering a pressure of 30 m and the mo-
mentum for 1600 rpm.
The angle of the curve, necessary to allow the communication
between the runner and an external generator, was changed from
90 to 45 to reduce ow head losses. Additionally, the turbine was
placed upstream from the curve to ensure a uniform distribution of
the upstream ow velocity, and also with the goal of reducing c-
tions losses. In Fig. 6 the nal experimental setup is presented.
As showed in Fig. 6, four pressure transducers are installed, one
upstream from the runner, two immediately upstream and
 I. Samora et al. / Renewable Energy 95 (2016) 356e366 359

0.5 100
0

90
0

0.4 80
0

70
0

[%]
0.3
H [bar]
60
0

50
0
0.2 40
0

30
0
0.1 20
0

10
0
0.0 0
0 2
20 40 60 0 10 20 30 40 5
50 60
Q [m3/h]] Q [m
m3/h]

a) Relaationship bettween dischaarge (Q) and b) Relationship

R b
between disccharge (Q) annd
headd (H) effficiency ()

1400 40
00

35
50
1200
30
00
1000
25
50
Pmec [W]
N [rpm]

800
20
00
600
15
50
400 10
00

200 5
50

0 0
0 2
20 40 60 0 10 20 30 40 5
50 60
Q [m3/h] Q [m
m3/h]

c) Relaationship bettween dischaarge (Q) and d) Relationship

R b
between disccharge (Q) annd
rotattion speed (N
N) m
mechanical poower (Pmec)
Fig. 2. Characteristic and efciency curves obtained from CFD numerical analysis of a 5BTP with a 100 mm impeller diameter.

Fig. 3. Laboratorial apparatus at IST.

downstream the runner and a fourth downstream the curve. These
transducers were connected to three absolute and two differential
pressure sensors, with accuracies between 0.05%, and 0.5%.
The ow was measured with an electromagnetic owmeter
with 0.5% of accuracy, the temperature of the water being
controlled with a probe. A permanent magnet generator with
1.23 kW and a maximum speed of 4880 rpm was used to impose
the rotation speed. The torque was measured with a sensor work-
ing from a range of 4 to 4 Nm and a measurement accuracy of 1%.
All instrumentation was calibrated before the experiments.
3.2. Similarity analysis
During the experimental procedure, only the turbine with
360 I. Samora et al. / Renewable Energy 95 (2016) 356e366

Fig. 4. Final design of the 5BTP. Left: design scheme. Right: 3D visualization.

Fig. 5. Pictures of the 5BTP. a) Comparison of the initial design, on the left, and the nal design, on the right. b) Runner, bulb and shaft of the nal design.

Fig. 6. Experimental setup. Top: scheme of components. Below: photo of the installation.

diameter of 85 mm was tested. Nevertheless, the results obtained if it is considered that there is a kinematic and dynamic similarity
can be extrapolated to other geometrically similar turbines with between them, then the following relation is valid if the efciency
different diameters. In fact, for two machines geometrically similar, is also the same [11].
 I. Samora et al. / Renewable Energy 95 (2016) 356e366 361

1 2
=
3 4

= 1 2
=
1 2

= 5 4
=
n1 H1 D2 H1 Q2 P2 H1
(1)
n2 H2 D1 H2 Q1 P1 H2
In this relation, called turbomachines similarity or afnity law, n
is the rotation speed, H is the head, D is the diameter, Q is the
discharge and P is the electric power, for two similar turbines 1 and
2.
Hence, from the performance of the measured machine, i. e. the
characteristic and efciency curves, the performance for other di-
ameters can be obtained. The state in which the machine has its
maximum value of efciency, corresponding to specic values of
discharge and head drop, is called the Best Efciency Point (BEP).
Since cavitation can affect the performance of turbomachinery,
the inuence of this phenomena on the hydraulic characteristics of
the turbine was also investigated. In this sense, the pressure level of
the test rig should be imposed so that cavitation occurs in similar
conditions as in a typical pressure system for drinking water dis-
tribution. For the similarity between cavitation in the model and
cavitation in a prototype, the Thoma's cavitation factor was
considered. The Thoma's cavitation factor is a dimensionless
number given by
NPSE
s (2)
E
where NPSE is the net positive suction energy (J/kg) and E the
specic hydraulic energy downstream from the runner (J/kg). The
NPSE is given by
rgHs patm  tva Cm
2
NPSE (3)
r 2
where r is the water density (kg/m3), g is the acceleration due to
gravity (m/s2), Hs is the setting level of the machine (m), patm is the
atmospheric pressure (Pa), tva the vapor pressure of water (Pa) and
Cm is the mean velocity downstream from the runner (m/s). If this
value is bellow a critical value, cavitation is likely to occur [2].
To compare the behavior of the model towards the cavitation, it
was considered that the number of Thoma should be the same
between the model and the prototype. Considering an average
pressure level was of 35 mWc, which is a reasonable value for
pressure in a water supply network, the test rig should then have an
imposed pressure of 22.62 mWc.
Nevertheless, after a few tests, it was veried that for an
imposed pressure equal to the atmospheric one, there was no
cavitation occurrence for the range of ows planned to be tested. In
conclusion, there is no effect of cavitation in the efciency of the
turbine. Hence, the pressure level was left at an imposed value of
10 mWc.
3.3. Experimental procedure
To assess the performance of the turbine, measurements were
performed in the installation covering a range of ows between 5
and 50 m3/h and keeping the rotating speed constant. For each
value of rotational speed imposed by the generator, between 50
and 2750 rpm, the ow was progressively varied to generate
measurement points of ow, pressure and torque.
From the pressure measurements, the characteristic curves, i. e.
the relationship between ow and head, were obtained for each
xed rotational speed. To dene the efciency curves, i.e. the
relationship between ow and efciency, the hydraulic power Ph
(W), the mechanical power Pmec (W) and the efciency hh were
calculated according to the following equations:
Ph rgQH (4)
2pN
Pmec ,Tmec (5)
60
Pmec
hh (6)
Ph
where Q is the ow (m3/s), H is the head (m) obtained by the dif-
ference of pressures measured with the transducers at the down-
stream and upstream edges of the circuit, r is the water's density
(kg/m3), N is the rotation speed (rpm) and Tmec is the mechanical
torque (Nm), measured with the torque sensor. All the tests were
performed with water temperatures between 15 and 25  C.
4. Results
4.1. Performance
The results of the performance assessment are presented in
Fig. 7, where we can analyze the total head between the pressure
transducers further upstream and downstream, the efciency, the
rotation speed and the mechanical power as function of the
discharge for the model with 85 mm of impeller diameter. For
twelve xed values of rotation speed, a total of 203 operation points
were measured. The efciency was calculated according to equa-
tion (6).
For a xed rotation speed, the efciency increases rapidly with
the increase of the discharge until a maximum is reached and then
it slowly starts to decrease. Nevertheless, the mechanical power
always increases with the increase of the discharge, showing that
the augmentation of the head compensates the decreasing of
efciency.
The hill charts of the turbine can be obtained by interpolation
between the measured data. In Fig. 8a) the hill chart relating ow,
rotation speed and efciency is represented, showing the peak of
efciency between 750 and 1000 rpm, for ows between 15 and
25 m3/h. In Fig. 8b) the hill charts relating ow, head and efciency
is presented, showing that the peak for discharges between 15 and
25 m3/h occurs for a head bellow 0.2 bar.
The obtained BEP corresponds to an efciency of around 64%,
held at a rotation speed of 750 rpm, a ow of 16 m3/h and 0.34 mWc
of head. The maximum mechanical power measured was of around
330 W, for a maximum ow of 48 m3/h, 0.48 bar of head and
1500 rpm of rotation speed.
The turbine specic speed NS is dened as the speed of a
geometrically similar turbine that, according to the afnity law
(equation (1)) produces a unit power (here considered in horse
power) under a unit head (1 m). It is given by the turbine's BEP
according to the expression:
1=
2
PBEP
NS NBEP 5= 283 m; kW 330 rpm m; h:p: (7)
4
HBEP
where PBEP is the power, NBEP the rotation speed and HBEP the head,
all corresponding to the BEP. The turbine presents a high specic
rotation speed, as it would be expected from a propeller. Axial
turbines have the highest specic rotation speeds, usually over
310 rpm m; h:p: [8].
4.2. Runaway conditions
Additional measurements were performed in runaway
362 I. Samora et al. / Renewable Energy 95 (2016) 356e366

Fig. 7. Characteristic and efciency curves obtained from the experimental results.

Fig. 8. Hill charts: a) Efciency as a function of the discharge and rotation speed; b) Efciency as a function of the discharge and head.

conditions, where the generator was physically detached to elimi-
nate any supplementary friction. In this conditions, no power is
produced and the rotation speed increases with the discharge. It is
imperative to know the rotation speed of the turbine in runaway
conditions when selecting the generator. In case of a sudden elec-
trical load rejection, both turbine and generator will rotate without
control until the problem is solved and this rotation should not
overheat the generator.
Additionally, it has been shown that reaction turbines, in
particular when the specic speed is low, runaway conditions can
induce dangerous overpressures during load rejection [27,39]. As
can be seen on Fig. 9, for turbines with specic speeds over
250 rpm, which is the case of Kaplan turbines and high specic
speed Francis, the transient discharge tends to increase, while for
lower specic speeds, the ow drops with the transient overspeed
[26]. To test if transients are formed within the considered range of
discharges, pressure measurements were also performed with no
control over the rotation speed.
 I. Samora et al. / Renewable Energy 95 (2016) 356e366
Fig. 9. Overspeed effect on discharge variation of reaction turbines [26].
The results of both pressure and rotation speed measurements
are presented in Fig. 10 by the red lines. The pairs of measurements
of discharge, head and rotation speed presented before are herein
presented for comparison.
For the same discharge, the available head is lower, resulting in a
higher downstream pressure along with a higher rotation speed.
Nevertheless, it was not registered any pressure surge and
discharge tends to increase, as expected from Fig. 9. The 5BTP is an
axial turbine, hence there is no ow drops in overspeed conditions,
and therefore no instabilities were noticed. Thus, within this range
of ow rates, there is no danger of pressure surges due to load
rejection.
4.3. Uncertainty
To conclude the experimental campaign, the measurements
were repeated for one xed rotation speed to estimate an average
variability of the results and obtain an estimation of the error
associated with the experimental procedure. The rotation speed
was xed at 1000 rpm, as it is a value for which it was possible to
cover a wide range of ows with adequate precision, giving a better
outlined efciency curve (Fig. 7b)). The results of this repetition is
given in Fig. 11, where it is visible a good superposition in all
characteristic and efciency curves.
5. Discussion
Considering the performance of the turbine, it is noticeable in
Fig. 7b) that the maximum efciency reached for each constant
rotation speed test is not the same. Since the turbine running with
different rotation speeds can be considered as two similar ma-
chines, it is noticeable that this means the afnity law, equation (1),
is not veried. The reason for this is the fact that the afnity law
Fig. 10. Characteristic curve and rotation speed in runaway conditions.
363
does not take into account factors, such as viscosity, that do not
scale with velocity and whose magnitude depend on the machine
size [37]. Nevertheless, the dimensions of the model are quite
small. For prototypes with larger diameters, the scale effects will be
reduced, possibly resulting in improvements on the hydraulic
efciency.
If the performance results are made dimensionless, considering
the rotation speed factor NED, the ow factor QED and the torque
factor TED [14], we obtain the curves of Fig. 12. The advantage of this
is to visualize all operating conditions of a turbine model, regard-
less of the diameter [14]. These parameters are given by the
following equations:
NDe
NED p (8)
60 E
Q
QED p (9)
D2e E
Tmec
TED (10)
rD3e E
where De is the external diameter of the runner (mm) and E the
specic energy (J/kg).
The dimensionless characteristic curves allow us to visualize
that the operating points of the turbine are approximately aligned
over a single curve although the rotation speed varies. This can be
justied by the fact that the blades of the turbine are xed, unlike
the Kaplan turbines. On the contrary, we can verify different curves
for the efciency depending on the rotation speed of the runner.
To compare the measurements with the characteristic and ef-
ciency curves obtained in CFD (Fig. 2), these had rst to be
transformed from a 100 mm runner to 85 mm. This was accom-
plished with the afnity law for turbomachines (equation (1)). As
we can see from Fig. 13, the model was able to predict approxi-
mately well the characteristic curves, but it largely overestimates
the efciency. This overestimation is due to the limitations of the
CFD model, which did not accounted for scale effects, anisotropy in
the zones of high ow circulation and leaks between the impeller
and the external envelop [30,32].
However, the design in CFD is fairly different from the tested
model, hence the direct comparison must be done with
reservations.
Finally, in order to maximize the mechanical power generated
for each ow-rate, the rotation speed of the turbine can be adjusted
according to Fig. 14. In all the operation points where the energy
production for the discharge was the maximum, the efciencies
were below the maximum efciency that could be achieved with
364 I. Samora et al. / Renewable Energy 95 (2016) 356e366

Fig. 11. Characteristic and efciency curves repetition for 1000 rpm.

Fig. 12. Dimensionless characteristic and efciency curves.

that rotation speed (overload operating regime). The efciencies
are compensated by the higher head turbinated, according to the
characteristic curve of Fig. 15.
6. Conclusion
In conclusion, an experimental campaign was conducted on a
turbine for micro-hydropower applications that has been in
development since 2008. A new model with a diameter of 85 mm
was optimized and tested, allowing recovering energy from ows
between 5 and 50 m3/h by extracting heads under 7.5 mWc. The
best efciency point was identied for a 750 rpm rotation speed,
with an operation point of 15.95 m3/h and 0.34 bar of head,
resulting in a 63.75% of efciency. The maximum power obtained
was of 328 W, for 1500 rpm of rotation speed, 48.15 m3/h of ow,
4.76 m of head and an efciency of 51.45%. These results can be
extrapolated for higher diameters through the afnity laws of tur-
bomachines. For example, the point of maximum power measured
would correspond to 71 kW for a diameter of 250 mm
Q ; H 1225 m3 =h; 4:1 bar and to 2.3 MW for a diameter of
500 mm Q ; H 9801 m3 =h; 16:5 bar.
The turbine has proved to have potential for further develop-
ment since the results of the experimental tests were satisfying,
with efciencies around 60%, which are considered good for such
 I. Samora et al. / Renewable Energy 95 (2016) 356e366 365

0.8 100

Measured points 90
0.7
80
0.6 CFD and affinity laws
70
0.5

[%]
60
H [bar]

0.4 50

40
0.3
30
0.2
20
0.1
10

0.0 0
0 10 20 30 40 50 0 10 20 30 40 50
Q [m3/h] Q [m3/h]

Fig. 13. Comparison between measured performance points and CFD results.

Fig. 14. Curve for maximum mechanical power production and respective efciencies and rotation speeds.

0.6

0.5

0.4
H [bar]

0.3

0.2

0.1

0
0 10 20 30 40
0 50 60
Q [m3//h]

Fig. 15. Characteristic curve considering the maximum mechanical power production.

small machines. 2012 issued by FCT under the IST-EPFL Joint PhD initiative. The
authors thank SFOE, the Swiss Federal Ofce of Energy, for the
Acknowledgements nancial support the experimental tests of the turbine.

This research is supported by PhD grant ref. SFRH/BD/51931/

366 I. Samora et al. / Renewable Energy 95 (2016) 356e366

References [23] G. Mller, J. Senior, Simplied theory of Archimedean screws, J. Hydraulic Res.
47 (5) (2009) 666e669, http://dx.doi.org/10.3826/jhr.2009.3475.
[24] C. Mnch-Alligne
, S. Richard, B. Meier, V. Hasmatuchi, F. Avellan, Numerical
[1] M. Arriaga, Pump as turbine e a pico-hydro alternative in Lao People's
simulations of a counter rotating micro turbine. Advances in Hydro-
Democratic Republic, Energy Strategy Rev. 35 (5) (2010) 1109e1115, http://
informatics, Springer Hydrogeol. (2014) 363e373, http://dx.doi.org/10.1007/
dx.doi.org/10.1016/j.renene.2009.08.022.
978-981-4451-42-0_30.
[2] P. Balachandran, Engineering Fluid Mechanics, PHI Private Learning Limited,
[25] O. Paish, Small hydro power: technology and current status, Renew. Sustain.
New Delhi, 2011.
Energy Rev. 6 (2002) 537e556, http://dx.doi.org/10.1016/S1364-0321(02)
[3] W.M.J. Batten, G.U. Mller, Potential for using the oating body structure to
00006-0.
develop head difference to increase the efciency of a free stream energy
[26] H. Ramos, Guidelines for Design of Small Hydropower Plants, CEHIDRO,
converter, in: 34th International Association of Hydraulic Engineering &
WREAN (Western Regional Energy Agency & Network) and DED (Department
Research (IAHR) Biennial Congress, Brisbane, Australia, 2011.

, Design & performance of Economic Development), Belfast, North Ireland, 2000.
[4] D. Biner, V. Hasmatuchi, F. Avellan, C. Mnch-Alligne
[27] H. Ramos, B. Almeida, Dynamic orice model on waterhammer analysis of
of a hydraulic micro-turbine with counter-rotating runners, in: 5th Interna-
high or medium heads of small hydropower schemes, J. Hydraulic Res. 39 (4)
tional Youth Conference on Energy 2015, Pisa, Italia, 2015.
(2002) 429e436.
[5] A. Carravetta, G. Giuseppe, O. Fecarotta, H. Ramos, Energy production in water
[28] H. Ramos, A. Borga, M. Sim~ ao, New design for low-power energy production
distribution networks: a PAT design strategy, Water Resour. Manag. 26 (13)
in water pipe systems, Water Sci. Eng. 2 (4) (2009a) 69e84, http://dx.doi.org/
(2012) 3947e3959, http://dx.doi.org/10.1007/s11269-012-0114-1.
10.3882/j.issn.1674-2370.2009.04.007.
[6] G. Caxaria, D. Mesquita e Sousa, H. Ramos, Small Scale Hydropower: Generator
[29] H.M. Ramos, A. Borga, M. Sim~ ao, Cost-effective energy production in water
Analysis and Optimization for Water Supply Systems, World Renewable En-
ping, Sweden, 2011. pipe systems: theoretical analysis for new design solutions, in: 33rd IAHR
ergy Congress 2011, Linko
Congress. Water Engineering for a Sustainable Environment, Managed by
[7] L. Corcoran, A. McNabola, P. Coughlan, Optimization of water distribution
EWRI of ASCE on Behalf of IAHR, British Columbia, Canada, 2009.
networks for combined hydropower energy recovery and leakage reduction,
[30] H. Ramos, M. Sima ~o, A. Borga, Experiments and CFD Analyses for a new re-
J. Water Resour. Plan. Manag. (2015), http://dx.doi.org/10.1061/(ASCE)
action microhydro propeller with ve blades, J. Energy Eng. 139 (2) (2013a)
WR.1943-5452.0000566, 04015045.
109e117, http://dx.doi.org/10.1061/(ASCE)EY.1943-7897.0000096.
[8] M.V. Desphande, Elements of Electrical Power Sation Design, PHI Learing
[31] H. Ramos, M. Sim~ ao, K. Kenov, Low-head energy conversion: a conceptual
Private Limited, New Delhi, 2010.
design and laboratory investigation of a microtubular hydro propeller, Int.
[9] A. Elbatran, O.B. Yaakob, Y.M. Ahmed, H.M. Shabara, Operation, performance
Sch. Netw. Mech. Eng. (2011), http://dx.doi.org/10.5402/2012/846206.
and economic analysis of low head micro-hydropower turbines for rural and
[32] H. Ramos, C. Theyssier, I. Samora, A. Schleiss, Energy recovery in SUDS to-
remote areas: a review, Renew. Sustain. Energy Rev. 43 (2015) 40e50, http://
wards smart water grids: a case study, Energy Policy 62 (62) (2013b)
dx.doi.org/10.1016/j.rser.2014.11.045.
 , A. Carravetta, R. Martino, H.M. Ramos, Hydropower 463e472, http://dx.doi.org/10.1016/j.enpol.2013.08.014.
[10] O. Fecarotta, C. Arico
[33] I. Samora, M.J. Franca, A.J. Schleiss, H.M. Ramos, Simulated annealing in
potential in water distribution networks: pressure control by PATs, Water
optimization of energy production in a water supply network, Water Resour.
Resour. Manag. 29 (2014) 699e714, http://dx.doi.org/10.1007/s11269-014-
Manag. (2015) under review.
0836-3.
[34] W. Schleicher, H. Ma, J. Riglin, Z. Kraybill, W. Wie, R. Klein, A. Oztekin, Char-
[11] R.S.R. Gorla, A.A. Khan, Turbomachinery: Design and Theory, CRC Press, New
acteristics of a micro-hydro turbine, J. Renew. Sustain. Energy 6 (1) (2014),
York, 2003.
http://dx.doi.org/10.1063/1.4862986.
[12] V. Hasmatuchi, F. Botero, S. Gabathuler, C. Munch, Design and control of a new
[35] J. Senior, O. Wiemann, G. Mller, The Rotary Hydraulic Pressure Machine for
hydraulic test rig for small hydro turbines, Hydropower & Dams 22 (4)
Very Low Head Hydropower Sites, University of Southampton, U.K, 2008.
(2015a) 54e60.
[36] M. Sim~ ao, Hydrodynamics and Performance of Low Power Turbines:
[13] V. Hasmatuchi, F. Botero, S. Gabathuler, C. Munch, New Hydraulic Test Rig for
Conception, Modeling and Essays. e in Portuguese, MSc Thesis, Instituto Su-
Small-power Turbomachines, IAHR World Congress, Hague, Netherlands,
perior Te
cnico, Lisbon, 2009.
2015b.
[37] A. Simpson, A. Marchi, Evaluating the approximation of the afnity laws and
[14] International Electrotechnical Committee (IEC), Hydraulic Turbines, Storage
improving the efciency estimate for variable speed pumps, J. Hydraulic Eng.
Pumps and Pump-turbines e Model Acceptance Tests, second ed., Interna-
139 (12) (2013) 1314e1317, http://dx.doi.org/10.1061/(ASCE)HY.1943-
tional Standard IEC 60193, 1999.
7900.0000776.
[15] HYLOW (Hydropower converters with very low head differences) (2010)
[38] P. Singh, F. Nestmann, Experimental optimization of a free vortex propeller
Periodic Report Summary. http://cordis.europa.eu/result/rcn/45354_en.html.
runner for micro hydro application, Exp. Therlam Fluid Sci. 33 (2009)
[16] HYLOW (Hydropower converters with very low head differences) (2012)
99e1002, http://dx.doi.org/10.1016/j.expthermusci.2009.04.007.
Periodic Report Summary 2. http://cordis.europa.eu/result/rcn/53053_en.
[39] J.P. Taulan, Pressure surges in hydroelectric installations: peculiar effects of
html.
low specic speed turbine characteristics, in: 4th International Conference on
[17] S. Jain, N. Patel, Investigations on pump running in turbine mode: a review of
Pressure Surges, Bath, England, 1983.
the state-of-the-art, Renew. Sustain. Energy Rev. (2014) 841e868, http://
[40] R. Weijermars, P. Taylor, O. Bahn, S.R. Das, Y. Wei, Review of models and
dx.doi.org/10.1016/j.rser.2013.11.030.
actors in energy mix optimization e can leader visions and decisions align
[18] S. Joshi, A. Holloway, L. Chang, H. Kojabadi, Development of a stand alone
with optimum model strategies for our future energy systems? Energy
micro-hydro system using pump as turbine technology for low head sites in
Strategy Rev. 1 (1) (2012) 5e18, http://dx.doi.org/10.1016/j.esr.2011.10.001.
remote areas, in: 20th International Power System Conference (PSC), 12-E-
[41] P. Wiemann, G. Mller, J. Senior, Risk Management and Resolution Strategies
EPG-130, Tehran, Iran, 2005.
for Established and Novel Technologies in the Low Head Small Hydropower
[19] M.W.K. Lyons, Lab Testing and Modeling of Archimedes Screw Turbines, MSc
Market, Hidroenergia 2008, Ljubljana, 2008.
Thesis, University of Guelph, Canada, 2014.
[42] A. Williams, Pumps as Turbines for Low Cost Micro Hydro Power, World
[20] A. McNabola, P. Coughlan, L. Corcoran, C. Power, A.P. Williams, I. Harris,
Renewable Energy Congress 1996, 1996.
J. Gallagher, D. Styles, Energy recovery in the water industry using micro-
[43] D.H. Zeiner-Gundersen, A novel exible foil vertical axis turbine for river,
hydropower: an opportunity to improve sustainability, Water Policy 16
ocean, and tidal applications, Appl. Energy 151 (2015) 60e66, http://
(2014) 168e183, http://dx.doi.org/10.2166/wp.2013.164.
dx.doi.org/10.1016/j.apenergy.2012.12.022.
[21] D. Melly, R. Horta, C. Mnch, H. Biner, S. Chevailler, Development of a PM-
[44] I. Andreev, K. Azmanov, S. Bozhinova, S. Petkova, D. Kisliakov, Installation of
generator for a counter-rotating micro-hydro turbine, in: XXI International
large scale model of a hydrostatic pressure machine on Isakar River. 5th
Conference on Electrical Machines, Berlin, Germany, 2014.
Bulgarian-Austrian Seminar, 2012. Small Dams and HPP.
[22] G. Mller, Water wheels as a power source, in: Seminar e Hydraulic Aspects
of Renewable Energy, Scottish Hydraulics Study Group, 2004.