Elsevier

Renewable Energy

Volume 87, Part 1, March 2016, Pages 332-345
Renewable Energy

Tow testing of Savonius wind turbine above a bluff body complemented by CFD simulation

https://doi.org/10.1016/j.renene.2015.10.015Get rights and content

Highlights

  • Wind over a cuboidal building generates a high velocity circular flow above it.

  • This flow field is best exploited by a drag-based Savonius wind turbine.

  • A three-bladed turbine installed above a bluff body has been tow tested.

  • Tow test results have been used to validate CFD simulation.

  • Installation of the turbine above a bluff body improves its cp at high wind speeds.

Abstract

A simple way to improve its power coefficient (cp) of a Savonius turbine is by its installation above a cuboidal building as the building will redirect the wind and increase its speed significantly. To determine the gain, a turbine was constructed and installed above a bluff body and tow tested. Detailed measurements of vehicle speed and turbine power were made. Tow test speeds were 8, 10 and 12 m/s, while TSR range was 0.6–1.1. Most importantly, wind speed at the position beside and slightly above the turbine was measured during test runs. The cp calculated using this measured wind speed was used to validate CFD simulation results. Simulation results were also used to obtain the relationships between the wind speed of the free stream and at the anemometer position. Typically, wind speed at the anemometer position is about 9% higher than those of the free stream. These relationships were used to derive the free stream wind speed of each experimental run. The cp calculated using these derived free stream wind speeds showed an increase of 25% at 12 m/s wind speed, compared to the cp reported by previous researchers for a similar turbine operating in unmodified air flow.

Introduction

The Savonius turbine invented by Finnish engineer Sigurd Johannes Savonius is a simple machine that extracts energy from the wind using the principle of differential drag between the advancing (moving in the same direction as the wind) and returning halves of the rotor. In spite of its lower maximum power coefficient, it still finds application in circumstances where simplicity of construction, high torque at low wind speeds, ability to withstand extreme wind conditions and self-limiting characteristic are overriding considerations. A typical large axial flow wind turbine in the megawatt range has power coefficient (cp) of more than 0.5 [1]. In contrast, the cp of drag-based turbines does not exceed 0.2, according to a summary of various devices by Saha and Rajkumar [2]. For a three-bladed rotor, the cp is even lower, at 0.16 [3], [4].

Therefore many researchers have attempted to improve the performance of the Savonius turbine. A commonly employed principle is the shielding of the returning blade. Examples are the use of a shielding plate [5], a deflector plate [6], guide vanes [7], etc. These prior works all involve additional components, which detracts from the one of the advantages of the Savonius turbine: its simplicity.

The present research focuses on the installation of a Savonius turbine with its axis horizontal above a flat-roofed cuboidal building. Our prior computational fluid dynamics (CFD) simulation has shown that such a building standing in the presence of wind generates a circular flow field just behind the windward top edge due to flow separation. The upper part of this circular flow region has high local wind speed, typically up to 25% higher than the free stream wind speed [8]. Just adjacent to the roof surface, there is backflow, i.e., the direction of the air flow is opposite to the direction of the wind. Since a drag-based wind machine produces useful torque when its blade is moving in the same direction as the wind but at a slower speed, hence a rotary drag-based turbine would be the best machine to exploit the circular flow field. An axial flow turbine would derive no benefit from the back flowing region even though it would also benefit from the high local wind speed at the upper part of the circular flow region. This concept of improving the performance of a Savonius turbine does not require additional components. Furthermore, existing structures can be used to install and optimise the wind turbine, in particular in urban areas, where the produced electricity is needed.

In this study, we performed tow tests of a Savonius wind turbine. This was after initial tests in our University's wind tunnel produced unsatisfactory results. The wind tunnel had a limited test section size of 0.78 m wide and 0.72 m high, which meant that when a forward facing step was installed, the turbine size was limited to 0.1 m in diameter. Such a small turbine produced very low torque and thus the relative error in torque measurements were quite large. In contrast, tow testing offers an infinitely large 'test section' without the blockage effect found in a wind tunnel. However, tow testing is subject to prevailing wind which is beyond our control. Fortunately, by conducting the tests during the inter-monsoon season, such disturbances were minimised.

A bluff body was constructed and installed around a lorry. A three-bladed Savonius turbine was installed with its axis horizontal above the bluff body. Instruments to measure vehicle speed, turbine power and wind speed were installed. Tow tests at vehicle speeds of 8, 10 and 12 m/s were conducted, with the turbine TSR ranging from 0.6 to 1.1. Full details are given in Section 3 of this paper.

Section snippets

Analytical power estimate

Erich Hau [9] gives the fundamental derivation of the cp of a drag-based wind machine. A blade subjected to wind experiences drag force proportional to its drag coefficient CD. In a wind machine, the effective wind seen by the blade is the difference between the wind speed (using a stationary reference) and the blade speed. The drag force multiplied by the blade speed gives the power output of the blade. Hence:P=CD[12ρA(UUb)2]Ub

When the blade is loaded so heavily that it is stationary, it

Design of the Savonius turbine

The Savonius turbine used in the present study is a three-bladed device with its axis of rotation installed horizontally. It has a diameter of 1.1 m and width of 1 m. The blades are made of aluminium sheet of thickness 3 mm rolled into a semi-cylinder. Each blade has a flange welded on each end of the semi-cylinder for it to be attached with counter sunk screws to the end plates. The medium density fibre end plates consist of an interior frame shaped like a spoked wheel with circular boards

Experimental results

Fig. 11 shows the results for the tow tests at 12 m/s vehicle speed. An important observation was that if the vehicle speed is used as the basis for calculating cp (shown by the blue triangles (in the web version) and diamonds), the plotted points scatter wildly without showing the expected trend of a quadratic curve with a maximum. This wild scatter is present even if the northbound and southbound data are considered as separate groups. Upon closer examination of the experimental data, it is

Conclusion and future work

To explore the possibility of improving the cp of a Savonius turbine simply by installing it on a building, a turbine was constructed and mounted on a bluff body and tow tested. Test results were used to validate CFD simulation satisfactorily by selectively using those that were obtained under low cross wind conditions. Simulation results were used to derive the free stream wind speed of the experimental runs and these were used to obtain composite results of the cp. It was found that at higher

Acknowledgements

The authors would like to thank the Maritime and Port Authority of Singapore for generously funding the present research. Additionally, they also express their appreciation to Edwin Ong Han Ze for his assistance during tow testing.

References (16)

There are more references available in the full text version of this article.

Cited by (35)

  • Overview of the development and application of wind energy in New Zealand

    2023, Energy and Built Environment
    Citation Excerpt :

    Due to the increasing interest in urban wind energy, installing wind turbines on the roof of cuboidal buildings became one of the possible scenarios. In order to understand the dynamic behavior of the turbine in such situations, Goh et al. [90] conducted the tow testing of a test rig equipped with a three-bladed Savonius mounted above a bluff body. The experimental results from that work were used to validate some of the simulation setups used for the present case study.

  • Improvement of startup performance of the drag-type hydrokinetic rotor through two-stage configuration

    2021, Ocean Engineering
    Citation Excerpt :

    The drag-type rotor is featured by simple rotor structure, low cost of infrastructure and high torque output, which facilitate its application in offshore regions and rivers (Zhang et al., 2019a,b). Drag type turbines are generally less efficient than properly designed lift type turbines because of the principle on which they operate (Hau, 2013; Goh et al., 2016). With the rotation of the drag-type rotor, the hydraulic torque on the rotor varies periodically (Kumar and Saini, 2016).

View all citing articles on Scopus
View full text