Effect of vortices on power output of vertical axis wind turbine (VAWT)

Effect of vortices on power output of vertical axis wind turbine (VAWT)

Sustainable Energy Technologies and Assessments 37 (2020) 100586 Contents lists available at ScienceDirect Sustainable Energy Technologies and Asses...

4MB Sizes 1 Downloads 9 Views

Sustainable Energy Technologies and Assessments 37 (2020) 100586

Contents lists available at ScienceDirect

Sustainable Energy Technologies and Assessments journal homepage: www.elsevier.com/locate/seta

Effect of vortices on power output of vertical axis wind turbine (VAWT) a

a,⁎

a

a

T

a

Ammar Naseem , Emad Uddin , Zaib Ali , Jawad Aslam , Samiur Rehman Shah , Muhammad Sajida, Ali Abbas Zaidia, Adeel Javedb, Muhammad Yamin Younisc a

Department of Mechanical Engineering, School of Mechanical and Manufacturing Engineering (SMME), National University of Sciences and Technology (NUST), Islamabad 44000, Pakistan US Pakistan Center of Advance Studies in Energy (USPCAS-E), National University of Sciences and Technology (NUST), Islamabad 44000, Pakistan c Department of Mechanical Engineering, Mirpur University of Science and Technology (MUST), Mirpur 10250, AJK, Pakistan b

A R T I C LE I N FO

A B S T R A C T

Keywords: VAWT Bluff body CFD Power augmentation

Performance of a vertical axis wind turbines (VAWTs) is a major focus of research these days. The present research focuses on numerical investigate the performance of a VAWT in the presence of upstream bluff bodies and investigate the impact of vortices shed on the VAWT performance. Designing of wind turbine is performed using Multiple Stream Tube Model which uses turbine power, number of turbine blades, aspect ratio and wind speed as parameter. Simulations are done using Reynolds averaged Navier-Stokes equation. Furthermore, VAWT equipped with three NACA 0018 airfoils is used. In this study, single and two upstream bluff bodies are used and the distance of the VAWT relative to a bluff body is changed to 1D, 2D and 3D (where D is the length and width of bluff body) both along and perpendicular to flow direction in order to find an optimal location for maximum power. This study shows that the deflected wind from the bluff body possess high wind velocity which is 12% higher for single bluff body and 25% higher for two bluff bodies than the free stream wind velocity. Moreover the VAWTs can be deployed downstream of tall urban structures for enhancement of wind turbine power generation.

Introduction Scientist and engineers are developing new methods and technology to efficiently harvest energy from the renewable resources. Wind is both clean and free. Vertical axis wind turbines (VAWT) can harvest energy from wind in any direction. This makes them suitable for urban deployment. VAWTs do not require any yaw or pitch mechanism for the rotor and the blades, respectively. Horizontal axis wind turbines face fatigue due to turbulent and gusty winds; however VAWTs have the ability to take advantage of such wind conditions. Strong support structures are not required for VAWTs as the gearbox, generator and other heavy components are mounted on the ground. This makes the replacement of turbine components simple and significantly reduces the maintenance costs. A study suggests that arrangement of VAWTs in an array that follows the natural phenomenon of fish schooling, the land area required for wind farm is reduced significantly. This results in increased power per unit area, generated by VAWTs as compared to HAWTs in wind farms [1,2]. The efficiency of VAWTs is low as compared to HAWTs as a single blade of the turbine produces maximum torque at a given instance of time in a complete rotation. Other major disadvantage of ⁎

VAWT is the dynamic stall of the blades, as the angle of attack is changing rapidly throughout the rotational cycle of the turbine [3,4]. Turbine efficiency is reduced; noise and structural vibrations are induced due to the phenomenon of dynamic stall. Study had been done using flat plate deflector behind the wind turbine and effects of unsteady flow flied on the turbine [5], but no study has been performed on the wake of one and two bluff bodies (mimicking the urban environment) in order to investigate the power augmentation and performance of wind turbines. In this study, performance of VAWT is investigated in the downstream region of a bluff body mimicking the urban structures. Wind turbines operate in wind farms which are usually situated in remote areas like coastal areas and open fields away from the urban environment. This increases transmission and maintenance costs. Urban structures such as tall buildings shed vortices, the effects of these wakes on the turbine performance is investigated. The distance of the turbine from a bluff body (axial and lateral) is varied to find out the optimal point where power can be maximized.

Corresponding author. E-mail address: [email protected] (E. Uddin).

https://doi.org/10.1016/j.seta.2019.100586 Received 30 March 2019; Received in revised form 26 October 2019; Accepted 12 November 2019 2213-1388/ © 2019 Elsevier Ltd. All rights reserved.

Sustainable Energy Technologies and Assessments 37 (2020) 100586

A. Naseem, et al.

Research background Different airfoils have been studied to find out their effects on the power output of the VAWT. NACA four digit series airfoils used in aviation industry are the most widely used airfoils in the wind turbine industry as well [6,7]. Usually airfoils having thickness, ranging from 12% to 21%, of their chord length are selected for VAWTs [7,8]. Cambering the airfoils has no significant effect on the performance of the VAWT because it causes the tangential force to increase in one half, but it also causes the tangential force to decrease in the other half of the total swept region [9]. Study of the vertical axis H-darrieus wind turbine model in a wind tunnel suggested that the number of blades influence the rotation of the rotor in a wind turbine. Increasing the number of blades makes it easy for the wind turbine to run at low wind speeds. Three blade configuration is more stable than two and four blade configurations [10]. The aspect ratio plays an important role in the performance, as well as it controls the power coefficient of the turbine. High aspect ratio VAWT has a lower power coefficient and vice versa [11]. The blades of VAWT produce maximum torque two times in the cycle that is front and back of the wind turbine. VAWTs have sinusoidal power cycle [12]. The straight blade H-darrieus wind turbines have a simple geometry because their blades are neither twisted nor they have yaw mechanism [13]. Bluff bodies are studied for performance enhancement of VAWTs; it is found that if a flat plate deflector is placed in the path of the wind perpendicular to the direction of flow it becomes a source of vortex. These vortices increase the wind speeds in the downstream of bluff bodies. If a deflector is placed between the wind path and the wind turbine, at a specific location, the turbine rotational speed increases which in turn increases the power output of the turbine. This deflector bluff body acts as an augmentation device which enhances the power coefficient by 33% [14]. The effect of flat plate deflector on the VAWT is also studied using CFD analysis. The detailed 3D analysis is done using CFD simulation in order to study the effect of deflector plate on the performance of wind turbine. It was found out that with increase in unsteady flow due to deflector plate the performance of wind turbine is increased. With addition of flat plate deflector the coefficient of torque is increased by 47.10% [5]. VAWT test case Rotor design has been optimized in the past by researchers and industry to improve the performance of the turbines. To predict the performance of the VAWTs many numerical codes have been developed like single streamtube model, multiple streamtube model, double multiple streamtube model, vortex method, actuator cylinder flow model and computational fluid dynamics method [15–19]. The wind turbine for this study has been designed using the methodology proposed by Brusca, Lanzafame and Messina in their research. This methodology has been derived from Multiple Stream Tube Model (MSTM). Fig. 1 shows a flow chart of the turbine design process [11]. The wind turbines are designed considering the wind speed and desired power output. The wind turbine used for this research is a 3 straight bladed H-rotor capable of producing 10 kW at 12 m/s wind speed. Power coefficient and tip speed ratio are plotted against each other as function of solidity of the rotor. Iterative method was used to calculate the turbine design parameters [11]. The power of a Vertical axis wind turbine is given by Eq. (1).

1 P = ρV 32RHCp 2

Fig. 1. Design process flowchart proposed by Brusca [11].

is selected from that plot. These values are then plugged in the power equation (Eq. (1)) and the wind turbine radius is found. The chord length of the airfoil is calculated using Eq. (2).

c= (1)

σCpmax Nb

R

(2)

Reynolds number of the wind turbine is calculated by Eq. (3).

Power, wind speed, number of turbine blades and aspect ratio are taken as inputs. Then a plot is selected for a specific Reynolds number (Re) and maximum power coefficient, tip speed ratio and blade solidity

Re =

2

cVo λ Cpmax v

(3)

Sustainable Energy Technologies and Assessments 37 (2020) 100586

A. Naseem, et al.

and flow can only past the building from above. For doing 3D simulations, the computation cost increases exponentially keeping in view the effects of vortices on 3D VAWT needs to be studied as well. However the results were changed by 2% on the top and back of the building due to the 3D effects [21]. Similarly we considered high rise and symmetric building to minimum the effects of ground and 3D vortices respectively.

Table 1 Design parameters. Parameter

Reference

Value

Power (kW) Airfoil Wind speed (m/s) Air density (Kg/m3) Kinematic air viscosity (m2/s) Rotor aspect ratio (h/R) Number of blades (Nb) Reynolds number (Re) cpmax, σcpmax, λcpmax Rotor radius (m) Airfoil chord length (m) Rotational speed (rad/s)

– [6–8] – At 300 K At 300 K [11] [10] Eq. (3) [11] Eq. (1) Eq. (2) Eq. (4)

10 NACA 0018 12 1.177 1.568 × 10−5 0.6 3 9.57 × 105 0.475, 0.3, 3.01 4 0.4 9.03

Computational domains VAWT rotor domain Based on the geometric data from design process in previous section, a CAD model of the turbine was generated. A fluid domain was created around the turbine. The domain is designed in two parts rotating fluid domain and fluid domain. This domain has been designed to validate the study. The direction of wind flow and rotation of wind turbine along with the dimensions of the validation domain is shown in Fig. 2. This domain was later meshed using ICEM CFD. A very fine mesh was used near the walls of the blades to capture the gradient of the boundary layer. For this purpose an inflation layer has been introduced. From theoretical derivations it is found that the dimensionless parameter (y+ ≤ 5) for the viscous sub layer [22,23].

This Re is compared with the original Re selected at the time of selection of the wind turbine power coefficient curves. The error is reduced by repeating the same procedure. Usually after two iterations the solution is converged and the error is reduced. VAWT rotational speed is given by Eq. (4).

ω=

λ Cpmax Vo R

(4)

After following the design processes the design parameters of the turbine are finalized and shown in Table 1. Fig. 2b shows the design and geometry of the H-Darreus type vertical axis wind turbine.

Single and two building domains Two more computational domains have been designed for this study. These computational domains have single and two buildings as bluff bodies. Fig. 3 and Fig. 4 show the computation domains for both scenarios. The dimensions of the domain are a multiple of urban structure dimension (D = 20 m). The wind flow in the downstream region of urban structures has been considered while selecting the locations of the wind turbine. A total of nine and twelve different cases have been formulated for single and double building scenarios. In these cases the wind turbine location have been changed in the × and y directions. Table 2 and Table 3 shows the distances of wind turbine from the urban structure for both scenarios. These locations have been set after detailed analysis the wind flow deflected by urban structures. The main strategy behind selecting these locations is to analyze the effects of deflected wind flow characteristics on the performance of the wind turbine appropriately. To make this study and comparison simple the wind turbine locations are related to urban structure dimension D. Starting from 0D and 1D in x and y directions respectively, with the increment of 1. The flow characteristics

Aerodynamic evaluation with CFD A comprehensive aerodynamic evaluation of the wind turbine has been conducted using ANSYS FLUENT and Qblade. The computational machine used for this study had two physical processors (6 cores each) and 16 GBs of RAM. Qblade is an open source software developed at TU Berlin for analysis of wind turbines specifically [20]. In this study the flow is considered incompressible. To analyze the turbulence in the flow many different models are derived from navier-stokes equations. In this study Reynolds Averaged Navier-Stokes equations (RANS) have been used to model the wakes generated in the flow by VAWT blades and the urban structures. In actual 3D simulation, the 3D effects like 3D vortices, ground effects and flow blockage also affects the flow. In 2D simulation the flow cannot go around the sides of the building. Full blockage can be seen

Fig. 2. Computational domain (validation). 3

Sustainable Energy Technologies and Assessments 37 (2020) 100586

A. Naseem, et al.

Fig. 3. Computational domain (single urban structure).

study. Upstream flow is considered laminar. After interaction with the urban structure and wind turbine blades, the flow is considered turbulent. To model the rotation of the wind turbine, a moving reference frame has to be incorporated in the domain. Rotating cell zone capabilities of the numerical solver have been used to visualize the VAWT rotating motion. The numerical solver used for this study can solve the problems using either relative velocity or the absolute velocity as dependent variables. For this particular study relative velocity is considered as dependent variable.

have no significant variation, if the distances between two consecutive wind turbine locations are reduced. If the distances are increased, the wind turbine performance cannot be analyzed appropriately as it will skip some of the key locations at which the wind flow characteristics have been altered due to deflection. These flow characteristics are discussed in detail in the next sections. Boundary conditions In this study the flow is considered incompressible. The upstream wind velocity has been varied from 3 m/s to 12 m/s with an increment of 1 m/s. The rotational velocity of the wind turbine has been calculated using Eq. (4). These rotational velocities have been shown in the Table 4. A 300 K ambient temperature has been considered for this

Grid generation and independence Grid independence study has been carried out on six grids of

Fig. 4. Computational domain (two urban structures). 4

Sustainable Energy Technologies and Assessments 37 (2020) 100586

A. Naseem, et al.

Table 2 Wind turbine locations for single urban structure. Location 1

2

3

4

5

6

7

8

9

X = 1D Y = 0D

X = 2D Y = 0D

X = 3D Y = 0D

X = 1D Y = 1D

X = 2D Y = 1D

X = 3D Y = 1D

X = 1D Y = 2D

X = 2D Y = 2D

X = 3D Y = 2D

different resolutions. Number of cells, corresponding to each grid resolution, has been compared with the coefficient of lift generated on each blade of the wind turbine. Initially a coarse grid of 70 k cells has been generated, the cell count has been approximately doubled for each new grid resolution until an optimum grid resolution has been obtained. The grid independence study is shown in Fig. 5. The grid independence study suggested that coefficient of lift has no significant change when the grid resolution is in between 1 and 2 Million cells. Thus, the domains have been meshed according to the results of the grid independence test. The optimum grid resolution was 0.9 Million cells. The optimum grid is shown in Fig. 6. The figure shows the stationary domain has structured mesh, rotating domain has unstructured mesh and there is an inflation layer of 20 layers with the initial layer height of 2 × 10−5 m and the growth rate of 1.2. The optimum mesh has average skewness 0.12 and average orthogonal quality 0.96.

Table 4 Simulation conditions for free stream velocity and rotational velocity. Index

Wind Speed(m/s)

Rotational velocity (rad/s)

1 2 3 4 5 6 7 8 9 10

3 4 5 6 7 8 9 10 11 12

2.2575 3.01 3.7625 4.515 5.2675 6.02 6.7725 7.525 8.2775 9.03

comparison with numerical solver software and Qblade software. The analytical results for turbine power are calculated using Eq. (1) for wind velocity changing from 3 m/s to 12 m/s (rest of variables in Eq. (1) remain same). For numerical solver software the computational domain for the validation study is shown in Fig. 2. The torque on each blade is calculated using Eq. (5). For the calculation of coefficient of moment Cm of each blade the turbine rotor is rotated for 10 revolutions and Cm is obtained by numerical solver. Cm from numerical solver is averaged for last 2 revolutions of turbine rotor for removing any transient effects. After obtaining torque, power on each blade is calculated using Eq. (6).

Turbulence model In this study SST k-ω turbulence model is used. This model uses k-ω model at the walls and switches to k-ε model in the free stream away from the walls. By switching to k-ε model it averts the high sensitivity to free stream turbulence properties faced by k-ω model [24,25]. This model is based on the Reynolds Averaged Navier-Stokes equations (RANS).

τ=

Yplus of the model

1 ρCm Vo2 HS 2

(5) (6)

P = τω

Yplus is the non-dimensional number, which is used to describe the coarseness and fineness of the mesh. For specific turbulence, model different value yplus is used depending upon the Reynold’s number. There are restriction of choosing reasonable overall y plus wall function for specific turbulence model. Generally, for SST k-ω turbulence model y plus value is taken to be ≤1. In the Table 5 the values for the y plus of each blade is shown along with contours in Fig. 7.

For obtaining total wind power, power on all blades is summed up. In case of Qblade, total wind power is directly obtained by selecting the studied parameters and wind turbine. The power obtained by experiments, analytical model (Eq. (1)), numerical solver software and Qblade is compared in Fig. 8. It can be seen that the power output from the numerical solver follows the same trend as of experimental and Qblade data for different wind velocities. At low wind velocities, the difference is very small for different approaches. However, at high wind velocities the difference of wind power increases but the overall trend remains same.

Results and discussion Validation In this paper, the studied cases are validated by comparing the power output obtained from the experimental data, analytical calculations, numerical solver software and Qblade software. The experimental results are obtained from the study of Aeolos V-10kKw wind turbine in real environment. Although, Aeolos V-10KW wind turbine has different aspect ratio than the studied wind turbine but it shows the same performance trend at different wind speeds and can be used for qualitative

Urban environment performance The performance of the wind turbine has been investigated in two different scenarios. A single building and two buildings have been considered as two urban scenarios in this study. These two scenarios were tested to visualize the vortex shedding of the buildings and their effect on the velocity magnitude around them. Vortex shed from urban

Table 3 Wind turbine locations for two urban structures. Location 1

2

3

4

5

6

7

8

9

10

11

12

X = 1D Y = 0D

X = 2D Y = 0D

X = 3D Y = 0D

X = 1D Y = 1D

X = 2D Y = 1D

X = 3D Y = 1D

X = 1D Y = 2D

X = 2D Y = 2D

X = 3D Y = 2D

X = 1D Y = 3D

X = 2D Y = 3D

X = 3D Y = 3D

5

Sustainable Energy Technologies and Assessments 37 (2020) 100586

A. Naseem, et al.

Table 5 Y plus value of the blades. Blades

Y plus values

Blade 1 Blade 2 Blade 3

0.93818104 0.67257905 0.65848517

been placed in the downstream region of the urban structure. These probes helped to analyze the flow characteristics of the wind in the downstream region. The output of these probes have been recorded and averaged over the complete cycle of 20.16 s. It has been established that, wind speed is augmented after deflection from the urban structures. This fact has been confirmed from previous research work [14]. Keeping the results of this analysis 9 and 12 locations have been picked for the analysis of wind turbines. Single building In the case of single urban structure as wind deflection device, Fig. 11 shows the wind characteristics. After 5.4 s a deflected wave is forming near the building walls shown in Fig. 11(a). This wave will augment the wind speed in the downstream region. Fig. 11(b,c,d & e) show that the defected wave is passing through the wind turbine locations. At these instances, it is found that the wind speed at locations 2,4,5,7 and 8 has been augmented initially, but later wind speed diminishes at locations 2 and 4. Similarly, Fig. 11 also suggests that the wind speed at locations 1, 2, 3, 5, 6 and 9 falls below the free stream wind velocity. The Fig. 9 shows the interaction of vortices generated by the bluff body on the turbine blades. This means that the urban structure has favorable and adverse effects on the downstream wind speeds. This trend of wind speed augmentation and diminution at deferent locations has been shown in the graph in Fig. 10. Fig. 5. Grid independence.

Two building In the case of two buildings, the wakes produced by deflection of wind from both the buildings effect each other. This leads to both augmentation and diminution of wind speed in the downstream region. Fig. 13 shows the wind characteristics in the in upstream and downstream regions of the urban structures. The figure suggests that the urban structures have a favorable effect on the downstream wind

structures require 20.16 s to complete one cycle. This complete vortex cycle is shown in the figures for both the scenarios. The figures show the augmentation in the free stream velocity after deflection by an urban structure. Before analyzing the wind turbine performance in this scenario, various velocity magnitude probes have

Fig. 6. Optimum grid resolution (1.2 million cells). 6

Sustainable Energy Technologies and Assessments 37 (2020) 100586

A. Naseem, et al.

Fig. 7. Yplus contours on every blade.

even falls below the free stream wind speed. This trend of wind speed augmentation and diminution at deferent locations has been shown in the graph in Fig. 12. Velocity magnitudes calculated by the simulation software are shown in Fig. 10. These figures show the behavior of flow for one complete cycle for both the scenarios. The results shown in the figures are at free stream velocity of 6 m/s. The time required for one complete cycle at this free stream velocity is approximately 20.20 s.

speeds. In Fig. 13(a), deflected wave is shown forming near the walls of buildings. In this case two different deflected waves have been observed. One of these waves is formed by the combined deflection of both the urban structures. Fig. 13(b, c, d, e & f) show that the deflected waves are passing through the wind turbine locations. At these instances, the wind speed at locations 1, 2, 3, 6, 7 and 8 augments initially, but later it diminished at locations 1,2,6,8 and 9. Similarly, Fig. 13 also suggests that the urban structures have an adverse effect on the wind speed at locations 1, 2, 3, 4, 5, 6, 7, 8 and 9. The wind speed

Fig. 8. Power comparison (validation). 7

Sustainable Energy Technologies and Assessments 37 (2020) 100586

A. Naseem, et al.

Fig. 9. Vorticity contours in the wake of square building.

Results

and 3D). These locations yield the Cm enhancement from 8% to 56%.

Wind turbine performance has analyzed for both scenarios. Location of wind turbine was changed as shown in the schematics (Fig. 3 & Fig. 4). The coefficient of moments calculated at different locations were compared with the results in free stream. Results mentioned in Table 6 and Table 7 shows that coefficient of moment is effected by the vortex shedding of buildings. The coefficient of moment is enhanced up to 24% and 56% in case of a single building and two buildings side by side used as a deflection obstacles respectively. In case of single building used as deflector the suitable turbine locations are X = (2D and 3D) at Y = 1D and X = (1D, 2D and 3D) at Y = 2D. The Cm enhancement on these locations vary from 10% to 24%. In case of two buildings side by side used as deflectors the suitable locations are X = 3D at Y = 0D and X = (1D, 2D and 3D) at Y = (2D

Conclusion In this study, the performance of an H-rotor vertical axis wind turbine was investigated in an urban environment. The study was conducted by varying the location of the wind turbine in the downstream region of urban scenarios. Two urban scenarios considered were single building and two buildings side by side. The characteristics of wind flow were analyzed and simulated. Details if the simulation setup has been described in the study. Velocity magnitude for urban scenarios and coefficient of moment and power for VAWT were calculated for this study. The VAWT performance in free stream was validated from experimental data literature and analytical calculation. The study shows a potential in power augmentation when used at

Fig. 10. Wind velocities at different wind turbine locations (single urban structure). 8

Sustainable Energy Technologies and Assessments 37 (2020) 100586

A. Naseem, et al.

Fig. 11. Velocity contours (single urban structure).

Fig. 12. Wind velocities at different wind turbine locations (two urban structures). 9

Sustainable Energy Technologies and Assessments 37 (2020) 100586

A. Naseem, et al.

Fig. 13. Velocity contours (Double urban structure).

locations in both urban scenarios, the high wind velocity enhances the torque generated on the turbine blades, which subsequently increases the power output of the VAWT. If the turbine is located in other locations, the wind speed reduces which decreases the power output of the VAWT significantly. The average Cm was increased by 24% and 56% for single building and two buildings side by side respectively. From this study, it is found that the VAWTs can be deployed in the downstream of buildings for enhancement of generated power. For future, this phenomenon may be validated experimentally. More parameters of the buildings such as aspect ratio, different shapes and height needed to explore in the future.

specific location in the downstream region of buildings (33°38′12.5″N 72°59′23.8″E). The buildings act as deflectors and subsequently augment the wind turbine power. The deflected wind from the buildings, near the wake region, possesses high wind velocity. This wind velocity is 12% and 25% higher than the free stream wind velocity in case of single building and two buildings side by side respectively. As the power possessed by wind is directly proportional to the cube of upstream velocity, this high velocity wind enhances the rotor RPM and turbine power output. According to the simulation results, the power of the VAWT is dependent on its location in the downstream of building. At optimum

Table 6 Comparison of turbine Cm calculated at different locations and in free stream (single building). Bare turbine

Averaged Cm Change (%)

1.78 –

Location 1

2

3

4

5

6

7

8

9

X = 1D Y = 0D

X = 2D Y = 0D

X = 3D Y = 0D

X = 1D Y = 1D

X = 2D Y = 1D

X = 3D Y = 1D

X = 1D Y = 2D

X = 2D Y = 2D

X = 3D Y = 2D

1.57 −12

1.41 −21

1.36 −24

1.67 −06

1.96 10

2.05 15

2.21 24

2.00 12

1.96 10

10

Sustainable Energy Technologies and Assessments 37 (2020) 100586

A. Naseem, et al.

Table 7 Comparison of turbine Cm calculated at different locations and in free stream (two buildings side by side). Bare Turbine

Averaged Cm Change (%)

1.78 –

Location 1 X = 1D Y = 0D

2 X = 2D Y = 0D

3 X = 3D Y = 0D

4 X = 1D Y = 1D

5 X = 2D Y = 1D

6 X = 3D Y = 1D

7 X = 1D Y = 2D

8 X = 2D Y = 2D

9 X = 3D Y = 2D

10 X = 1D Y = 3D

11 X = 2D Y = 3D

12 X = 3D Y = 3D

1.75 −02

1.59 −11

2.18 22

1.53 −14

1.33 −25

1.06 −40

2.78 56

2.27 27

1.93 08

2.44 37

2.40 34

2.34 32

Declaration of Competing Interest

2014;5(4):333–40. [12] D'Ambrosio, M. and M. Medaglia, Vertical axis wind turbines: History, technology and applications. 2010. [13] Islam M, Ting DS-K, Fartaj A. Aerodynamic models for Darrieus-type straightbladed vertical axis wind turbines. Renew Sustain Energy Rev 2008;12(4):1087–109. [14] Wong KH, et al. Experimental and simulation investigation into the effects of a flat plate deflector on vertical axis wind turbine. Energy Convers Manage 2018;160:109–25. [15] Strickland JH, Webster B, Nguyen T. A vortex model of the Darrieus turbine: an analytical and experimental study. J Fluids Eng 1979;101(4):500–5. [16] Strickland J. The Darrieus turbine: a performance prediction model using multiple streamtube. SAND 1975:75e0431. [17] Paraschivoiu I. Double-multiple streamtube model for studying vertical-axis wind turbines. J Propul Power 1988;4(4):370–7. [18] Beri H, Yao Y. Double multiple streamtube model and numerical analysis of vertical axis wind turbine. Energy Power Eng 2011;3(03):262. [19] Lanzafame R, Mauro S, Messina M. 2D CFD modeling of H-Darrieus wind turbines using a transition turbulence model. Energy Proc 2014;45:131–40. [20] Marten D, et al. QBLADE: an open source tool for design and simulation of horizontal and vertical axis wind turbines. Int J Emerging Technol Adv Eng 2013;3(3):264–9. [21] 21. Buckingham, S., et al., CFD simulation of the three-dimensional effects induced by the nox photocatalytic degradation around an isolated building. [22] Lam HF, Peng H. Study of wake characteristics of a vertical axis wind turbine by two-and three-dimensional computational fluid dynamics simulations. Renewable Energy 2016;90:386–98. [23] Tahani M, et al. A novel heuristic method for optimization of straight blade vertical axis wind turbine. Energy Convers Manage 2016;127:461–76. [24] Mohamed M, Ali A, Hafiz A. CFD analysis for H-rotor Darrieus turbine as a low speed wind energy converter. Eng Sci Technol Int J 2015;18(1):1–13. [25] Moshfeghi M, Song YJ, Xie YH. Effects of near-wall grid spacing on SST-K-ω model using NREL Phase VI horizontal axis wind turbine. J Wind Eng Ind Aerodyn 2012;107:94–105.

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper. References [1] Whittlesey RW, Liska S, Dabiri JO. Fish schooling as a basis for vertical axis wind turbine farm design. Bioinspiration Biomimetics 2010;5(3). [2] Dabiri JO. Potential order-of-magnitude enhancement of wind farm power density via counter-rotating vertical-axis wind turbine arrays. J Renewable Sustainable Energy 2011;3(4). [3] Buchner A-J, et al. Dynamic stall in vertical axis wind turbines: scaling and topological considerations. J Fluid Mech 2018;841:746–66. [4] Buchner A, et al. Dynamic stall in vertical axis wind turbines: comparing experiments and computations. J Wind Eng Ind Aerodyn 2015;146:163–71. [5] Wong KH, et al. 3D CFD simulation and parametric study of a flat plate deflector for vertical axis wind turbine. Renewable Energy 2018;129:32–55. [6] Holdsworth B. Options for micro-wind generation: part two. Renewable Energy Focus 2009;10(3):42–5. [7] Brulle R. Feasibility investigation of the giromill for generation of electrical power. St. Louis, MO (USA): McDonnell Aircraft Co.; 1975. [8] Tjiu W, et al. Darrieus vertical axis wind turbine for power generation I: Assessment of Darrieus VAWT configurations. Renewable Energy 2015;75:50–67. [9] Deglaire P, et al. Analytical solutions for a single blade in vertical axis turbine motion in two-dimensions. Eur J Mech B/Fluids 2009;28(4):506–20. [10] Sunyoto, A., F. Wenehenubun, and H. Sutanto. The effect of number of blades on the performance of H-Darrieus type wind turbine. in QiR (Quality in Research), 2013 International Conference on. 2013. IEEE. [11] Brusca S, Lanzafame R, Messina M. Design of a vertical-axis wind turbine: how the aspect ratio affects the turbine’s performance. Int J Energy Environ Eng

11