Vous êtes sur la page 1sur 18

Renewable Energy 113 (2017) 461e478

Contents lists available at ScienceDirect

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

Performance analysis of a single stage modified Savonius hydrokinetic


turbine having twisted blades
Anuj Kumar, Research Scholar *, R.P. Saini, Professor
Alternate Hydro Energy Centre, Indian Institute of Technology, Roorkee, 247667, India

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

Article history: Savonius hydrokinetic turbine is one of the prominent vertical axis turbines for tapping hydro potential
Received 15 November 2016 available in flowing streams in rivers or canals. In spite of their simple design, Savonius turbines have the
Received in revised form problem of poor performance. This study aims to enhance the performance of turbine through modifi-
15 May 2017
cation in the blade shape. Under the present study, geometrical parameters namely blade arc angle and
Accepted 3 June 2017
blade shape factor are considered to modify the blade shape of Savonius hydrokinetic turbine. A com-
Available online 6 June 2017
mercial unsteady Reynolds-Averaged Navier-Stokes (URANS) solver in conjunction with realizable k-ε
turbulence model has been used for numerical analysis. Using CFD analysis, blade arc angle and blade
Keywords:
Modified Savonius hydrokinetic turbine
shape factor are optimized on the basis of coefficient of power. Fluid flow distributions found around the
Computational fluid dynamics (CFD) rotor has also been analyzed and discussed. Based on the present investigation, the maximum power
Power coefficient coefficient value of 0.426 is obtained for blade arc angle of 150 and blade shape factor of 0.6 corre-
Micro hydropower sponding to TSR value of 0.9 at flow velocity of 2 m/s.
Numerical modeling © 2017 Elsevier Ltd. All rights reserved.
Blade arc angle
Blade shape factor

1. Introduction outflows, irrigation canals, rivers and tidal streams [3]. In order to
harness energy from river or canal, turbines are broadly classified
Since last couple of decades, intensive research has been stim- into two categories i.e. axial-flow turbines and cross-flow turbines
ulated for development and promotion of renewable energy gen- [4]. Horizontal axis turbines are more expensive for small power
eration due to combination of social, environmental and economic applications [5]. Guney and Kaygusuz [6] reported that vertical axis
interests. Shifting to renewable energy can ensure for reliable, turbine is more suitable for the cases of limited water flow rate.
timely, and cost-efficient delivery of energy without environmental Khan et al. [7] also stated that vertical axis turbines can be installed
threats. Renewable energy technologies have also been serving as either as a single unit in small rivers or stacked together in large
economical sources of electricity for rural and remote areas since rivers. Several advantages i.e. less expensive, independent of flow
last couple of years. To produce electricity in such area, hydropower direction, easy installation and maintenance are driving the
is emerged as most cost effective, efficient and reliable solution. But intensive research into the design and development of improved
large-scale hydropower generation is always criticized because of vertical axis turbine. Savonius turbine (semi-circular), Gorlov tur-
its negative environmental impact while small scale hydropower bine (helically shaped blades), Darrieus turbine and H-shaped
generation is considered as environmentally friendly technology. Darrieus (straight blades) are the popular turbines in the category
Therefore, due to clean, sustainable and locally available, harness- of vertical axis turbines [8].
ing naturally-occurring high energy flows (i.e. river stream, canal Savonius hydrokinetic turbine, drag-type rotor starts rotating at
flow) is the best alternative to electricity the rural areas [1]. This very low fluid velocity as compared to the conventional hydraulic
kind of generation is known as in-stream hydrokinetic energy turbine. It has an ability to accept fluid from any direction with
production [2]. Hydrokinetic technologies have their geneses in good starting characteristics at low speed. Despite such advantages,
wind energy extraction technologies. Hydrokinetic turbines can Savonius hydrokinetic turbines face low efficiency and large static
also be set up for a wide range of applications including industrial torque variation [3]. In the last couple of decades, a lot of studies
including experimental, numerical studies were carried out to
enhance the power coefficient of Savonius hydrokinetic turbine.
* Corresponding author. Khan et al. [9] tested experimentally single and multi-stage Savo-
E-mail address: anuj.iitr02@gmail.com (A. Kumar). nius hydrokinetic turbine in a water channel for Reynolds numbers

http://dx.doi.org/10.1016/j.renene.2017.06.020
0960-1481/© 2017 Elsevier Ltd. All rights reserved.
462 A. Kumar, R.P. Saini / Renewable Energy 113 (2017) 461e478

Nomenclature S assumed source terms


k turbulent kinetic energy
H Height of turbine ε dissipation rate of the turbulent kinetic energy
D Diameter of turbine Р density of fluid
D0 Diameter of end plate uj velocity components
q radius of circular arc xj Cartesian coordinate
p straight edge of the blade T time
j Blade arc angle m viscosity
a Twist angle mt turbulent viscosity
u Angular velocity of rotor sk constant of the standard k-ε turbulence model
CP Coefficient of power Gk generation of turbulence kinetic energy due to the
L Tip speed ratio (TSR) mean velocity gradients
U Free stream velocity Gb generation of turbulence kinetic energy due to
CM Coefficient of moment buoyancy
Uij is the mean rate-of-rotation tensor viewed in a rotating YM is the effect of the changing dilatation in compressible
reference frame with the angular velocity uk turbulence to the overall dissipation rate

ranging from 0.98  105 to 1.96  105. The maximum coefficient of develop a workflow, starting from generation of 3D CFD model,
power for single stage, two stage, and three stage Savonius rotors is mesh generation, setting fluent solver to results analysis in CFD-
reported as 0.038, 0.049 and 0.04 respectively. Nakajima et al. [10] Post. This software is widely used to simulate the flow around
carried out an experimental study to test Savonius water turbine the turbine. In order to accomplish the proposed objective, the
for Reynolds Number of 1.1  105 and reported the maximum po- procedure adopted under the present study is given in a flow chart
wer coefficient of 0.25. Further, single stage and double stage (with as shown in Fig. 1.
0 and 90 phase shift) Savonius water turbine have been investi- As shown in Fig. 1, the component of optimization process is
gated experimentally to visualize the flow field around the rotor carried out by determining the optimal value of blade arc angle (j)
[11]. The study concluded that double stage rotor with 90 phase and blade shape factor (p/q). In order to obtain the maximum value
difference produces 10% larger power coefficient. Golecha et al. of CP, blade arc angle varied for a given value of fixed blade shape
[12,13] reported the improvement in coefficient of power by using factor (p/q). The simulations have been carried out for different
deflector plate. The coefficient of power is improved by 50% with values of TSR for a given range of flow velocity. Corresponding to
the deflector plate on the returning blade side and 150% with maximum value of CP, the blade arc angle is identified as optimal
deflector plates on the both blade side respectively. Yaakob et al. arc angle. On similar lines, for a given value of blade arc angle, the
[14] investigated conventional Savonius water turbine numerically, blade shape factor is optimized.
and maximum CP is reported as 0.275 at TSR of 0.7, which were in
good agreement with the experimental results. Kumar and Saini 2. Parameters investigated for Savonius hydrokinetic turbine
[15] carried out numerical investigation to improve the power co-
efficient by using twisted blade profile. Under the present study, the geometrical parameters consid-
The literature review as mentioned above emphases on the ered are blade arc angle (j) and blade shape factor (p/q). The
studies carried out on Savonius hydrokinetic turbine. In earlier schematic view of Savonius hydrokinetic turbine having twisted
studies, the problem of having low efficiency of Savonius wind blades along with geometrical parameters considered (j and p/q)
rotor have also been addressed by modifying blade shape like for the study is presented in Fig. 2. Blade cross section keeps on
modified Bach, Benesh, semi-elliptical [16e18]. However, no in- changing certain degree in each step along the vertical axis (z-axis)
formation is available on modification of blade shape in case of in the case of twisted blade profile as shown in Fig. 2(iv).
twisted bladed Savonius hydrokinetic turbine. Hence, there is a Table 1 gives geometrical parameters of twisted blade Savonius
need to analyze the effect of blade arc angle and blade shape factor hydrokinetic turbine considered for numerical simulations. Opti-
on the performance of twisted blade Savonius hydrokinetic turbine mized twist angle for Savonius hydrokinetic turbine was reported
in an open channel flow. as 12.5 in an earlier study [15], and same is taken for the present
In order to study the effect of blade arc angle and blade shape study. In order to study the effect of blade arc angle (j) and blade
factor on the performance of twisted blade Savonius hydrokinetic shape factor (p/q), Table 2 presents configurations considered in
turbine, unsteady and rotating 3D CFD simulations are carried out the present study.
under the present study. It is, therefore, the objectives of the pre- Under the present study, the range of water velocity is consid-
sent study are as follow: ered as 0.5 m/s to 2 m/s. A cylindrical shape encompassing the
turbine is created which represents a rotating zone of the domain
1. To study the effect of blade arc angle on the performance of for numerical simulations. In other studies [19e21], the size of a
twisted Savonius hydrokinetic turbine rotating zone has been considered as 1.28 times of turbine diameter
2. To investigate the effect of blade shape factor on the perfor- to 3 times of turbine diameter (1.28D-3D). The diameter of the
mance of twisted Savonius hydrokinetic turbine rotating zone is considered as 1.5D for the present investigation.
3. To study the velocity and pressure distributions across the tur- Computational domain similar to an open channel having di-
bine under different velocity conditions. mensions of 0.55 m  0.65 m  3.0 m is also designed for numerical
investigation. This domain has water surface similar to open
Under the present study, commercially available software channel flow; the normal water level in the channel is set at 0.65 m.
ANSYS 15.0 is used. ANSYS workbench provides a platform to Savonius hydrokinetic turbine is considered to be completely
A. Kumar, R.P. Saini / Renewable Energy 113 (2017) 461e478 463

Fig. 1. Flow chart of methodology adopted for present study.

submerged in an open channel. Table 3 presents working param- condition (no-slip wall) is assigned to the turbine. Symmetry con-
eters considered for simulation. dition is given to the free surface of the open channel [23]. For CFD
study, effects of free surface on the performance of turbine are not
considered. Tip speed ratio (TSR) is dependent on angular velocity
3. Modeling and numerical analysis
and free stream velocity of water for a given turbine diameter. In
order to obtain results at considered TSR range, a number of sim-
3.1. Computation domains and boundary settings
ulations are carried out by varying angular velocity given to
rotating zone.
For numerical analysis, creating a computational domain is
considered as a time-consuming and challenging task. Under this
study, three-dimensional model of Savonius hydrokinetic turbine is 3.2. Grid generation
generated in ANSYS “Design Modeler” as per the parameters pro-
vided in Table 1. Cylindrical volume enclosing Savonius hydroki- Under the present study, ANSYS MESH is used for generating
netic turbine is created about Z-axis in such a way that Savonius mesh of 3D domain. A non-conformal unstructured grid having
turbine can rotate at the specified angular velocity during numer- tetrahedral elements is considered to generate mesh for the com-
ical analysis. In order to ensure the continuity of the flow field, the plete flow domain as shown in Fig. 4. Fine mesh has been created in
interface is also created between stationary zone and rotating zone. the rotational region as compared to stationary zone. In order to
Complete computational domain including stationary zone and perform grid independence test, three different mesh elements
rotating zone bear a resemblance to the immersed turbine in an ranging from 12.4e18.7 million are created for the complete
open channel, as shown in Fig. 3. domain. As the results, 15.3 million elements are estimated as grid
Table 4 gives boundary conditions considered for CFD analysis elements beyond which negligible change in power coefficient re-
and the same are shown in Fig. 3. It is seen from the figure that left sults is observed. Accordingly, the refinement of mesh is stopped at
side of the channel (inlet) has been set as velocity inlet with a ve- 15.3 million grid elements. In order to capture flow near bound-
locity of free stream velocity. The outlet of the open channel has aries, inflation (prism layer) was applied to the rotor wall. Under
been set as outflow condition [22,23]. The side and bottom walls of earlier studies [24,25], yþ value less than one was recommended to
the channel are provided as slip boundary, and rotating wall fix the height of first prism layer above the blade surface to apply
464 A. Kumar, R.P. Saini / Renewable Energy 113 (2017) 461e478

Fig. 2. Schematic of twisted blade Savonius hydrokinetic turbine for different geometrical parameters.

the realizable k-ε turbulence model. Accordingly, the value of yþ has been used as flow solver to solve unsteady incompressible
has been considered as less than one under the present investiga- Navier-Stokes equations with the help of finite volume discretiza-
tion. Table 5 presents various mesh parameters which affect quality tion scheme. To capture the continuous change in the flow field,
of mesh generated are found to be in good agreement. unsteady Reynolds-Averaged Navier-Stokes (URANS) in conjunc-
tion with realizable k-ε turbulence model is employed. In order to
obtain a numerical solution, the case with required boundary
3.3. Simulation procedure conditions is defined and solved. In this study, FLUENT solver runs
in parallel processing using HPC clusters.
ANSYS FLUENT (v 15.0), wherein generated mesh is imported,

3.3.1. Turbulence model


Table 1
Turbulence is considered as critical which should be accounted
Geometrical parameters of twisted blade Savonius hydrokinetic
turbine. precisely in the case of water flow. There are a number of models
available to take care of turbulence factor in the simulation. Rey-
Parameter Value
nolds number and flow geometry are the prevailing parameters to
Number of blades 2 select an appropriate turbulence model. Under the present study,
Aspect ratio (H/D) 1.58
the range of Reynolds number has been computed as 7.36  104 to
Overlap ratio (e) 0
Rotor diameter (D) 0.160 m 37.52  104 considering the range of flow velocity and turbine
Rotor height (H) 0.253 m diameter. For the range of Reynolds number considered, the flow
Endplate diameter (D0) 1.1D field exhibits small eddies and turbulence. It is therefore, there is a
Twist angle (a) 12.5 need to select an appropriate turbulence model. In order to select
Blade arc angle (j) 110 e150
suitable turbulence model for present study, the results of earlier
Blade shape factor (p/q) 0e0.6

Table 3
Table 2 Working Parameters considered for Numerical analysis.
Configurations of Savonius rotor covered in the present study.
Parameter Value
Configuration Twist angle Blade arc angle (u) Blade shape
Rotating zone diameter 0.24 m
factor (p/q)
Open channel height 0.65 m
I 12.5 110 , 124 ,135 0 Open channel width 0.55 m
and 150 Open channel length 3.0 m

II 12.5 150 0.2, 0.4 and 0.6 Water velocity in channel (U) 0.5e2.0 m/s
A. Kumar, R.P. Saini / Renewable Energy 113 (2017) 461e478 465

Fig. 3. Boundary conditions for CFD analysis.

Table 4 turbulence model under the present study to simulate the rotating
Boundary conditions for Numerical Analysis. behavior of turbine blades. This model has two major differences
Name Boundary type Boundary condition from the Standard k-ε. It comprises a new formulation for the
Inlet Velocity Inlet 0.5e2.0 m/s, uniform flow
turbulent viscosity and a new transport equation for the dissipation
Outlet Outflow outflow rate ε, derived from an exact equation for the transport of the
Channel side and bottom walls Free slip wall Stationary mean-square vorticity fluctuations. Further, this model does not
Top surface of channel Symmetry Symmetry rely on the assumed relationship between the Reynolds stress
Turbine No slip wall Rotates at desired RPM
tensor and the strain rate tensor. In the realizable k-ε model, the
coefficient of the model is expressed as a function of mean flow and
turbulence properties, rather than assumed to be constant as in the
study carried out by Hayashi et al. [26] are referred. The results of standard model.
their study is shown in Fig. 5 which demonstrates the excellent After selecting a realizable k-ε model, boundary conditions are
agreement obtained between CFD and experiments results for the set as given in Table 3. Turbulence quantities like turbulence in-
realizable k  ε turbulence model. Accordingly, realizable k  ε tensity and turbulent viscosity ratio are to be defined at the inlet of
turbulence model has been considered for the present study. It was the open channel. Turbulent intensity, accountable for the quality
suggested under earlier studies [27e34] that Realizable k-ε turbu- of fluid flow around the device, may affect the performance of a
lence model is better to simulate the rotating behavior of blades or Savonius rotor. A turbulence intensity of 1% or less is generally
airfoil, flow through the channel, a boundary layer or separated considered low, and turbulence intensities greater than 10% are
flows. Accordingly, the Realizable k-ε model has been employed as considered high [35]. Akwa et al. [36] also reported that maximum

Fig. 4. Mesh showing boundary layer and dense mesh in rotating zone.
466 A. Kumar, R.P. Saini / Renewable Energy 113 (2017) 461e478

Table 5 step size along with a number of the time steps is defined for nu-
Mesh details for rotating zone and stationary zone. merical investigations. Under the present study, TSR range is
Quality aspect Rotating zone Stationary zone considered from 0.5 to 1.0 which covers the complete operating
Number of element 5856361 9444159
range for Savonius turbine operation. Time step size considered is
Skewness 0.84 0.84 taken as 15 /time step which means the rotor will complete its one
Orthogonal quality 0.86 0.86 revolution in 24-time step. When quasi-steady state for the
Aspect ratio 14 16 magnitude of torque fluctuations is achieved, the simulation is
assumed to be completed. To achieve this quasi-steady state, the
total number of time steps is set as 120. Coefficients of momentum
averaged power coefficient decreases with the increase of turbu- (Cm) obtained from last revolutions of the turbine are sampled to
lence intensity in the fluid flow. It is, therefore, turbulence intensity obtain average torque. The last revolution under the present study
is specified as 5% or less depending upon tip speed ratio in the has been considered based on results reported under earlier studies
present study whereas turbulent viscosity ratio is specified by [15,20,21 & 38]. The power of coefficient is obtained by multipli-
default value as 10%. cation of TSR and average coefficients of momentum over an entire
Pressure based Solver with absolute velocity formulation is revolution.
adopted for simulation. Multiple reference frame (MRF) is applied
for rotating zone in which equations are solved in a rotational 4. Results and discussion
reference frame in terms of relative velocity, acceleration [37]. With
the help of interface, rotating zone and stationary zone is associated In order to study the performance of Savonius hydrokinetic
with each other to model the rotation of Savonius hydrokinetic turbine, simulations have been carried out to investigate the effect
turbine. In Fluent Solver, the rotating zone is set up by providing a of blade arc angle and blade shape factor for different values of TSR
rotational velocity in the absolute reference frame. Then wall (ranging from 0.5 to 1.0) and flow velocity (0.5 m/se2 m/s).
boundary condition for Savonius rotor has a zero relative speed Simulation results have been obtained for all the values of pa-
with respect to adjacent cells [37]. rameters considered. However, flow distribution around the tur-
In order to obtain results with more accuracy, the CFD investi- bine is described corresponding to maximum value of CP.
gation is carried out adopting a SIMPLE (Semi-Implicit Method for
Pressure-Linked Equations) pressureevelocity coupled solver along 4.1. Velocity contours and streamlines
with second order upwind scheme for all convection terms i.e.
equation of momentum, turbulent kinetic energy, and turbulent Based on numerical investigation, velocity contours in different
dissipation rate. The least squares cell based algorithm is consid- region in the vicinity of blade have been obtained and presented in
ered to evaluate all the gradients. Convergence criteria for all re- plots as shown in Fig. 6. Velocity contours obtained for twisted
siduals of momentum, continuity, and turbulence equations were Savonius turbine for blade arc angle of 150 at TSR value of 0.9 and
defined as 1  105 for each time step. In FLUENT solver, reference for 0.5 m/s is shown in Fig. 6(a). The maximum and minimum
values are given as given in Table 6. values of velocity are indicated by red and blue colors respectively
For the transient analysis, the numerical iterative process is in the figures. It is apparent from velocity contour that velocity
significantly influenced by time step size. Therefore, optimal time magnitude at left side of plot is 0.5 m/s which is same as given at
inlet boundary of domain. At the tip of advancing blade, the
rotating vortices of high speed are also noticed. It is observed that
as soon as water approaches blades of the rotor, flow velocity de-
creases at trailing edge of blade and wake zone is formed on the
downstream side of the channel. Periodic high speed zones are also
found near the wake zone which means flow regains its velocity
after passing the wake zone. The velocity distributions around
turbine blades are validated with the results of earlier study carried
out by Zhang et al. [39], as shown in Fig. 6(b). The velocity contours
obtained for the present study exhibits the similar trends obtained
for earlier study.
Fig. 7 shows the flow distribution around the turbine for blade
arc angle (j ¼ 150 ) and blade shape factor (p/q ¼ 0.6) for a typical
value of flow velocity of 2 m/s. All velocity contours on XY and YZ
plane are shown at a TSR value corresponding to the maximum
coefficient of power. Higher velocity region is observed at the top
and bottom side as shown in YZ plane near end plate. For blade arc
angle value of 150 , water velocity increases up to 5.535 m/s as it
reaches to the blade. This maximum velocity is observed at tip of
advancing blade, as shown in Fig. 7(a). However, maximum velocity

Table 6
Reference value for simulation to calculate moment coefficient.

Reference parameter Value

Length Turbine radius


Area Swept area of turbine
Velocity Free stream velocity (U ¼ 0.5e2 m/s)
Fig. 5. Comparison of turbulence models results with earlier published experimental
Density 998.2 kg/m3
results [26,32].
A. Kumar, R.P. Saini / Renewable Energy 113 (2017) 461e478 467

Fig. 6. (a) Velocity contours at 0.5 m/s and TSR ¼ 0.9 and j ¼ 150 (b) Velocity contours given by Zhang et al. [39].

is found as 4.55 m/s near the advancing blade corresponding to maximum. At stagnation point, the fluid velocity is zero, and all
blade shape factor of 0.6, as shown in Fig. 7(b). Details of velocity kinetic energy has been converted into pressure energy. Again, the
contours obtained for other values of blade arc angle and blade pressure starts to increase after passing the Savonius rotor and flow
shape factor in the considered range of flow velocity, blade arc experiences an adverse pressure gradient. Due to the adverse
angle and blade shape factor are presented in the Appendix A. pressure gradient, the flow separates from the surface and creates a
The lines which are drawn tangential to velocity vector at each highly turbulent region behind the rotor. Due to flow separation
point in the flow at a given instant, are known as velocity and recirculating flow, wake zone appears in the flow domain
streamlines. Velocity Streamlines help in understanding flow behind rotating Savonius rotor. The pressure inside the wake region
around the rotor. In velocity vector, the color of vector gives the remains low as the flow separates and a net pressure force (pres-
magnitude of velocity while vector direction represents flow di- sure drag) is produced. Streamlines and vector path for blade arc
rection. In order to study the complete flow profile around the angle of 110 , 124 , and 135 ; blade shape factor of 0.2 and 0.4 are
Savonius rotor, streamlines and velocity vector are observed. For presented in Appendix B.
the optimize value of geometrical parameters studied, streamlines
and vector are shown in Fig. 8 at flow velocity of 2 m/s. It is clear 4.2. Pressure contours
from these figures that high randomness is found in the flow
behind the rotor which causes turbulence in the flow. High tur- Fig. 9 presents pressure contours for the optimized value of
bulence can be observed after the rotor as shown in streamlines. blade arc angle (j ¼ 150 ) and blade shape factor (p/q ¼ 0.6)
From vector path, it can be observed that high velocity vortices respectively. All pressure contours on XY and YZ plane are shown
(shown by the yellow and red line of vectors) are observed at the for flow velocity of 2 m/s at a TSR value corresponding to the
edge of advancing blade for all twist angle studied. Tip of advancing maximum coefficient of power. In order to predict the pressure
blade acts as the stagnation point where pressure is found to be variations in various regions near the blades within the flow
468 A. Kumar, R.P. Saini / Renewable Energy 113 (2017) 461e478

Fig. 7. Velocity contours around the turbine at V ¼ 2 m/s.

domain, pressure contours plots are used. The maximum and an earlier study [40,41], as shown in Fig. 10. A good agreement is
minimum values of pressure are indicated by red and blue colors found between the results. Pressure contours for blade arc angle
respectively in pressure contours. From these figures, it can be of 110 , 124 and 135 ; blade shape factor of 0.2 and 0.4 are
observed that pressure drop occurs across the rotor from up- presented in Appendix C.
stream to downstream side. At the inlet of the channel, pressure
is found uniform, whereas higher pressure values are found at 4.3. Performance characteristics
the concave side of advancing blade. Lower pressure region is
developed on the convex side of advancing blade and the Under the present study, a numerical investigation on the per-
concave side of returning blade. Thus, two pressure regions i.e. formance of Savonius hydrokinetic turbine with geometrical pa-
higher and lower are found near the blades within the flow rameters considered is carried out. In order to complete the
domain which creates a pressure drop. Turbine blades start objectives of this study, modeling, and meshing of the computa-
rotating due to this pressure drop across the rotor which causes tional domain is completed as narrated under the previous section
power extraction by the Savonius hydrokinetic turbine from the of the paper. The value of the coefficient of power (CP) corre-
flowing water in the channel. For j ¼ 150 , pressure is found as sponding to its TSR (l) is used to describe the performance of the
7991 Pa at the concave surface of advancing blade and negative turbine. Characteristics curves (CP vs. l) are obtained for different
pressure of 3129 Pa at the convex side of the blade. Due to this geometrical parameters and flow parameters using the following
pressure difference, Savonius hydrokinetic turbine starts to Eqs. (1)e(3).
rotate. Similarly, pressure difference of 11618 Pa is observed
across the rotor blades in case of blade shape factor of 0.6, which 4T
Cm ¼ (1)
generate drag force and consequently torque. Pressure contours rU 3 D2 H
obtained are compared with the pressure contours obtained in
Fig. 8. Streamlines and Vector path at 2 m/s.

Fig. 9. Pressure distribution around the turbine at 2 m/s.


470 A. Kumar, R.P. Saini / Renewable Energy 113 (2017) 461e478

Fig. 10. Pressure distribution around the turbine blade (a) given by Wahyudi et al. [40] (b) given by Zhao et al. [41].

Pturbine 2T u 2T u
uD CP ¼ ¼ ¼ ¼ Cm l (3)
TSRðlÞ ¼ (2) Pavailable rAS U 3 rHDU 3
2U

Fig. 11. Variation of CP with respect to TSR for blade arc angle studied.
A. Kumar, R.P. Saini / Renewable Energy 113 (2017) 461e478 471

obtained for the Savonius hydrokinetic turbine having blade arc


angle of 150 with the results of earlier study [15] carried out for
conventional Savonius hydrokinetic turbine at water velocity of
0.5 m/s. This comparison has been made to establish the fact that
proposed changes in blade arc angle offer better performance in
terms of coefficient of power at higher TSR, as shown in Fig. 12.

4.3.2. Effect of blade shape factor (p/q)


Numerical analysis of Savonius hydrokinetic turbine was carried
out with a twist angle of 12.5 and blade arc angle of 150 with
varying blade shape factor (p/q). Blade shape factor (p/q) studied
are 0.2, 0.4 and 0.6 with fixed value of aspect ratio of (H/D) of 1.58,
overlap ratio of 0.0. Fig. 13 represents the effect of blade shape
factor (p/q) on the coefficient of power for a complete range of flow
velocity. Savonius hydrokinetic turbine with a blade shape factor
(p/q) of 0.6 is found to have the higher coefficient of power
compared to rotors having other blade shape factor (p/q) of 0.2, 0.4
and 0.6. Based on the numerical investigations, it can be concluded
that Savonius hydrokinetic turbine with twist angle (a) of 12.5
having blade arc angle (j) of 150 and blade shape factor 0.6 results
Fig. 12. Comparison of CP with results of earlier study [15]. in highest CP of 0.426. Therefore, the twisted Savonius hydrokinetic
turbine having blade arc angle (j) of 150 and blade shape factor
4.3.1. Effect of blade arc angle 0.6 can be termed as single stage modified Savonius hydrokinetic
Numerical analysis of Savonius hydrokinetic turbine was carried turbine having twist blades. Table 8 presents the peak value of the
out with a twist angle of 12.5 and varying blade arc angle (j). coefficient of power (CPmax) and corresponding tip speed ratio (TSR)
Blade arc angle (j) studied are 110 , 124 ,135 and 150 with fixed at flow velocities considered. Fig. 12 shows comparison of results
value of aspect ratio of (H/D) of 1.58, blade shape factor (p/q) of 0.0. obtained at 0.5 m/s for blade shape factor of 0.6 with conventional
It is noteworthy here that the blade with blade arc angle of 180 Savonius hydrokinetic turbine tested in open channel at 0.5 m/s.
corresponds to a normal blade shape. In an earlier study, twisted This comparison shows that blade shape factor of 0.6 offer better
Savonius hydrokinetic turbine with blade arc angle of 180 has been performance in terms of coefficient of power. Modified blade shape
investigated [15]. The twist angle of 12.5 , aspect ratio of 1.58 with (j ¼ 150 , p/q ¼ 0.6) contributes in average lifting force due to
zero overlap between blades was considered for the study. Fig. 11 change in moment arm and large surface area at concave surface of
represents the effect of blade arc angle on the coefficient of po- blade, consequently increases average torque on blade which
wer for a complete range of flow velocity. Results of twisted improve the overall performance of turbine.
Savonius hydrokinetic turbine with Blade arc angle of 180 [15] is
taken for comparative study as shown in Fig. 12. Table 7 presents 5. Conclusion
the peak value of the coefficient of power (CPmax) and corre-
sponding tip speed ratio (TSR) at a flow velocity considered. It was Under the present study, numerical simulations have been car-
observed that Savonius hydrokinetic turbine with a blade arc angle ried out to investigate the effect of blade arc angle and blade shape
of 150 is found to have the higher coefficient of power compared to factor for different values of TSR (ranging from 0.5 to 1.0) and flow
rotors having other blade arc angle of 110 ,124 ,135 and 180 . An velocity of water ranging from 0.5 m/s to 2 m/s. Two different
attempt has also been made to compare the power coefficient (CP) configurations of turbine blade are investigated to identify the

Table 7
Maximum coefficient of power and the corresponding tip speed ratio for different blade arc angle.

Blade arc angle (j) Flow velocity (m/s) Maximum Coefficient of power CPmax TSR corresponding to CPmax

110 0.5 m/s 0.26 0.9


1.0 m/s 0.26 0.9
1.5 m/s 0.28 0.9
2.0 m/s 0.29 0.9
124 0.5 m/s 0.29 0.9
1.0 m/s 0.30 0.9
1.5 m/s 0.33 0.9
2.0 m/s 0.35 0.9
135 0.5 m/s 0.26 0.9
1.0 m/s 0.29 0.9
1.5 m/s 0.32 0.9
2.0 m/s 0.35 0.9
150 0.5 m/s 0.35 0.9
1.0 m/s 0.36 0.9
1.5 m/s 0.37 0.9
2.0 m/s 0.41 0.9
180 (normal shape) 0.5 m/s 0.34 0.9
1.0 m/s 0.35 0.9
1.5 m/s 0.36 0.9
2.0 m/s 0.39 0.9
472 A. Kumar, R.P. Saini / Renewable Energy 113 (2017) 461e478

Fig. 13. Variation of CP with respect to TSR for blade shape factor studied.

optimal values of geometrical parameters for the range of param- been analysed under the present paper. The results are also vali-
eters considered. In configuration I, blade arc angle was varied from dated with published results available in literature. Based on the
110 e150 for a fixed value of blade shape factor (p/q) as 0.0 while investigations, the optimal geometrical parameters for modified
for configuration II, blade shape factor was varied from 0.2e0.6 for Savonius hydrokinetic turbine having twisted blade are found to be
fixed blade arc angle of 150 . as blade arc angle of 150 and blade shape factor of 0.6. It has also
In order to obtain simulation results, a commercial unsteady been found that coefficient of power increases with flow velocity.
Reynolds-Averaged Navier-Stokes (URANS) solver in conjunction It is recommended that present study may be useful for further
with realizable k-ε turbulence model has been used. Based on nu- investigations of the performance of modified Savonius hydroki-
merical simulations, flow distribution around turbine blades has netic turbine for a different system and operating parameters.

Table 8
Maximum coefficient of power and the corresponding tip speed ratio for different blade shape factor.

Blade shape factor (p/q) Flow velocity (m/s) Maximum Coefficient of power CPmax TSR corresponding to CPmax

0 0.5 m/s 0.35 0.9


1.0 m/s 0.36 0.9
1.5 m/s 0.37 0.9
2.0 m/s 0.41 0.9
0.2 0.5 m/s 0.31 0.9
1.0 m/s 0.35 0.9
1.5 m/s 0.36 0.9
2.0 m/s 0.37 0.9
0.4 0.5 m/s 0.35 0.9
1.0 m/s 0.36 0.9
1.5 m/s 0.37 0.9
2.0 m/s 0.37 0.9
0.6 0.5 m/s 0.39 0.9
1.0 m/s 0.40 0.9
1.5 m/s 0.42 0.9
2.0 m/s 0.43 0.9
A. Kumar, R.P. Saini / Renewable Energy 113 (2017) 461e478 473

Appendix A. Velocity contours

Fig. A.1 shows velocity contours for Savonius hydrokinetic tur-


bine having blade arc angle ranging from 110 e135 for different
values of flow velocity from 0.5 m/s to 2.0 m/s. The velocity con-
tours are shown for a typical value of TSR of 0.9. It can be seen that
patterns of velocity contours are found to be similar for entire flow
velocity range considered. However, a small change in velocity
magnitude has been observed. Similarly, Fig. A.2 shows the velocity
contours for Savonius hydrokinetic turbine having blade arc angle
of 150 and blade shape factor of 0.2 and 0.4.

Fig. A.1. Velocity distribution around the turbine for blade arc angle.

Fig. A.2. Velocity distribution around the turbine for blade shape factor.
474 A. Kumar, R.P. Saini / Renewable Energy 113 (2017) 461e478

Appendix B. Streamlines and vectors

Fig. B.1 present streamlines and vector for Savonius hydroki-


netic turbine having blade arc angle ranging from 110 e135 for
different values of flow velocity from 0.5 m/s to 2.0 m/s. The
streamlines and vectors are shown for a typical value of TSR of 0.9.
Similarly, Fig. B.2 shows the streamlines and vector for Savonius
hydrokinetic turbine having blade arc angle of 150 and blade
shape factor of 0.2 and 0.4.

Fig. B.1. Streamlines and Vector path around turbine with blade arc angle.
A. Kumar, R.P. Saini / Renewable Energy 113 (2017) 461e478 475

Fig. B.2. Streamlines and Vector path around turbine with blade shape factor.

Appendix C. Pressure contours patterns of pressure contours are found to be similar for entire flow
velocity range considered. However, a change in pressure differ-
Fig. C.1 present pressure contours for Savonius hydrokinetic ence across the blades surface has been observed. Similarly, Fig. C.2
turbine having blade arc angle ranging from 110 e135 for different shows the pressure contours for Savonius hydrokinetic turbine
values of flow velocity from 0.5 m/s to 2.0 m/s. The pressure con- having blade arc angle of 150 and blade shape factor of 0.2 and 0.4.
tours are shown for a typical value of TSR of 0.9. It can be seen that
476
A. Kumar, R.P. Saini / Renewable Energy 113 (2017) 461e478
Fig. C.1. Pressure contours around turbine with blade arc angle.
A. Kumar, R.P. Saini / Renewable Energy 113 (2017) 461e478 477

References

[1] C. Boccaletti, G. Fabbri, J. Marco, E. Santini, An overview on renewable energy


technologies for developing countries: the case of Guinea Bissau, in: Inter-
national Conference on Renewable Energy and Power Quality (ICREPQ'08),
2008.
[2] Amir Hossein Birjandi, Eric Louis Bibeau, Vijay Chatoorgoon, Anurag Kumar,
Power measurement of hydrokinetic turbines with free-surface and blockage
effect, Ocean. Eng. 69 (1) (September 2013) 9e17.
[3] A. Kumar, R.P. Saini, Performance parameters of Savonius type hydrokinetic
turbine e a Review, Renew. Sustain. Energy Rev. 64 (2016) 289e310.
[4] Kumar A, Saini RP. Numerical investigation and novel designing of multi-stage
Savonius rotor for harnessing hydro power. In: Annual IEEE India Conference
(INDICON) 2015.
[5] Y.M. Dai, W. Lam, Numerical study of straight-bladed Darrieus-type tidal
turbine, Proc. Inst. Civ. Eng. 162 (2009) 67e76.
[6] M.S. Guney, K. Kaygusuz, Hydrokinetic energy conversion systems: a tech-
nology status review, Renew. Sustain. Energy Rev. 14 (2010) 2996e3004.
[7] M.J. Khan, G. Bhuyan, M.T. Iqbal, J.E. Quaicoe, Hydrokinetic energy conversion
systems and assessment of horizontal and vertical axis turbines for river and
tidal applications: a technology status review, Appl. Energy 86 (2009)
1823e1835.
[8] Kumar A, Saini RP, Investigation on performance of improved Savonius rotor:
an overview, International Conference on Recent Developments in Control,
Automation and Power Engineering (RDCAPE) 2015.
[9] M.N.I. Khan, M.T. Iqbal, M. Hinchey, V. Mase, Performance of Savonius rotor as
a water current turbine, J. Ocean. Technol. 4 (2009) 71e83.
Fig. C.2. Pressure contours around turbine with blade shape factor.

[10] M. Nakajima, S. Iio, T. Ikeda, Performance of Savonius rotor for environmen-


tally friendly hydraulic turbine, J. Fluid Sci. Technol. 3 (3) (2008) 420e429.
[11] M. Nakajima, S. Iio, T. Ikeda, Performance of double step Savonius rotor for
environmentally friendly hydraulic turbine, J. Fluid Sci. Technol. 3 (2008)
410e419.
[12] K. Golecha, T.I. Eldho, S.V. Prabhu, Influence of the deflector plate on the
performance of modified Savonius hydrokinetic turbine, Appl. Energy 88
(2011) 3207e3217.
[13] K. Golecha, T.I. Eldho, S.V. Prabhu, Performance study of modified Savonius
water turbine with two deflector plates, Int. J. Rotating Mach. 2012 (2012)
1e12.
[14] O. Yaakob, Y.M. Ahmed, M.A. Ismail, Validation study for Savonius vertical axis
marine current turbine using CFD simulation, in: 6th Asia-Pacific Workshop
on Marine Hydrodynamics-AP Hydro2012, September 3e4, 2012.
[15] Anuj Kumar, R.P. Saini, Performance analysis of a Savonius hydrokinetic tur-
bine having twisted blades, Renew. Energy 108 (August 2017) 502e522.
[16] S. Roy, U.K. Saha, Wind tunnel experiments of a newly developed two-bladed
Savonius-style wind turbine, Appl. Energy 137 (2015) 117e125.
[17] M.A. Kamoji, S.B. Kedare, S.V. Prabhu, Experimental investigations on single
stage modified Savonius rotor, Appl. Energy 86 (7e8) (2009) 1064e1073.
[18] S. Roy, U.K. Saha, Review of experimental investigations into the design,
performance and optimization of the Savonius rotor, Proc. Insti. Mech. Eng.
Part A-J. Power Energy 227 (4) (2013) 528e542.
[19] S. McTavish, D. Feszty, T. Sankar, Steady and rotating computational fluid
dynamics simulations of a novel vertical axis wind turbine for small-scale
power generation, Renew. Energy 41 (2012) 171e179.
[20] P. Marsh, D. Ranmuthugala, I. Penesis, G. Thomas, Three-dimensional nu-
merical simulations of straight-bladed vertical axis tidal turbines investigating
power output, torque ripple and mounting forces, Renew. Energy 83 (2015)
67e77.
[21] A. Ghosh, A. Biswas, K.K. Sharma, R. Gupta, Computational analysis of flow
physics of a combined three bladed Darrieus Savonius wind rotor, J. Energy
Inst. 88 (4) (2015) 425e437.
[22] N. Kolekar, A. Arindam Banerjee, Performance characterization and placement
of a marine hydrokinetic turbine in a tidal channel under boundary proximity
and blockage effects, Appl. Energy 148 (2015) 121e133.
[23] N.K. Sarma, A. Biswas, R.D. Misra, Experimental and computational evaluation
of Savonius hydrokinetic turbine for low velocity condition with comparison
to Savonius wind turbine at the same input power, Energy Convers. Manag. 83
(2014) 88e98.
[24] W. Tian, J.H. VanZwieten, P. Pyakurel, Y. Li, Influences of yaw angle and tur-
bulence intensity on the performance of a 20 kW in-stream hydrokinetic
turbine, Energy 111 (2016) 104e116.
[25] T. Zhou, D. Rempfer, Numerical study of detailed flow field and performance
of Savonius wind turbines, Renew. Energy 51 (2013) 373e381.
[26] T. Hayashi, Y. Li, Y. Hara, Wind tunnel tests on a different phase three-stage
Savonius rotor JSME, Int. J. Ser. B Fluids Therm. Eng. 48 (1) (2005) 9e16.
[27] M.H. Mohamed, A.M. Ali, A.A. Hafiz, CFD analysis for H-rotor Darrieus turbine
as a low speed wind energy converter, Eng. Sci. Technol. Int. J. 18 (1) (2015)
1e13.
[28] M.H. Mohamed, Performance investigation of H-rotor Darrieus turbine with
new airfoil shapes, Energy 47 (1) (2012) 522e530.
[29] M.H. Mohamed, Impacts of solidity and hybrid system in small wind turbines
performance, Energy 57 (1) (2013) 495e504.
[30] M.H. Mohamed, G. Janiga, E. Pap, D. The venin, Optimization of Savonius
478 A. Kumar, R.P. Saini / Renewable Energy 113 (2017) 461e478

turbines using an obstacle shielding the returning blade, Renew. Energy 35 wind turbines, Renew. Sustain. Energy Rev. 16 (5) (2012) 3054e3064.
(11) (2010) 2618e2626. [37] R. Lanzafame, S. Mauro, M. Messina, Wind turbine CFD modeling using a
[31] M.R. Castelli, A. Englaro, E. Benini, The Darrieus wind turbine: proposal for a correlation-based transitional model, Renew. Energy 52 (2013) 31e39.
new performance prediction model based on CFD, Energy 36 (8) (2011) [38] T. Maître, E. Amet, C. Pellone, Modeling of the flow in a Darrieus water tur-
4919e4934. bine: wall grid refinement analysis and comparison with experiments, Renew.
[32] M.H. Mohamed, G. Janiga, E. Pap, D. The venin, Optimal blade shape of a Energy 51 (2013) 497e512.
modified Savonius turbine using an obstacle shielding the returning blade, [39] B. Zhang, B. Song, Z. Mao, W. Tian, A novel wake energy reuse method to
Energy Convers. Manag. 52 (1) (2011) 236e242. optimize the layout for Savonius-type vertical axis wind turbines, Energy 121
[33] M.H. Mohamed, G. Janiga, E. Pap, D. The venin, Multi-objective optimization of (2017) 341e355.
the airfoil shape of wells turbine used for wave energy conversion, Energy 36 [40] B. Wahyudi, S. Soeparman, S. Wahyudi, W. Denny, A simulation study of Flow
(1) (2011) 438e446. and Pressure distribution patterns in and around of Tandem Blade Rotor of
[34] Y. Chen, Y. Lian, Numerical investigation of vortex dynamics in an H-rotor Savonius (TBS) hydrokinetic turbine model, J. Clean Energy Technol. 1 (2013)
vertical axis wind turbine, Eng. Appl. Comput. Fluid Mech. 9 (1) (2015) 21e32. 286e291.
[35] Fluent User Guide <https://www.sharcnet.ca/Software/Fluent6/html/ug/ [41] Zhao Z, Zheng Y, Xu X, Liu W and Hu G. Research on the improvement of the
node217.htm> (Accessed 17 June 2016). performance of Savonius rotor based on numerical study. International Con-
[36] J.V. Akwa, H.A. Vielmo, A.P. Petry, A review on the performance of Savonius ference on Sustainable Power Generation and Supply 2009; 1e6.

Vous aimerez peut-être aussi