Acknowledgements

ConclusionThe project required the design of a 20W wind turbine; the final product was capable of producing up to 35-40W of power, therefore, overall the project was a success. The Lenz2 type wind turbine works very well in low wind conditions; it is self starting at about 4mph winds and can maintain its rotation at low speeds.

Design and Fabrication of a Solar-Wind Hybrid Power System

AbstractRural coastal communities in developing countries along the equator lack electricity due to the expenses involved with connecting them to the power grid. An affordable local solution, such as on-site micro power generation, would benefit these communities and fulfill their basic lighting and ventilation needs. This project involves the development such a system; consisting of a vertical axis wind turbine integrated with a solar panel to generate direct current (DC) power.

Prad Pathirana Advised by Prof. Richard Wilk

Lenz2 Wind TurbineThe Lenz2 turbine design is a gyro mill type wind turbine with three scoops at 120o to each other. Although the scoops have an aerodynamic design it’s mainly a drag based wind turbine, like the Savonius turbine. An advantage that the Lenz2 has over the Savonius turbine is the fact that it can reach much faster rotational speeds. SymLab4 was used to model lenz2 blades under different wind speeds and showed promising results as seen in figure 10.

References1. NASA . 18 Mar. 2009 <http://www.nasa.gov>. 2. "Topic Ten : Local Winds." WRGB CBS 6 Albany. 10 Nov. 2008 <http://www.cbs6albany.com/sections/weather/research/topics/topicten/>.3. Elliot, D., and M. Schwartz. Wind Energy Resource Atlas of Sri Lanka and Maldives. Rep. National Renewable Energy Laboratory. 17 Mar. 2009 <http://www.nrel.gov/wind/pdfs/34518.pdf>.4. "SymLab | Symscape." Symscape | Computer-Aided Engineering for All. Fall 2008 <http://www.symscape.com/product/symlab/>.

Introduction and Background

This project was motivated by the predicament of rural coastal communities in Sri Lanka that have no access to the power grid. Sri Lanka happens to be a developing island nation with vast stretches of coastline. Many of the communities along the coast have very poor living standards and one of their main concerns is the lack of electricity for lighting and other uses. Currently, they use kerosene lamps as a source of lighting and car batteries to power basic electronics. These are the cause of both fire, and other health hazards. Children from these coastal communities tend to study with kerosene lamps for lighting; mishaps and fires are common place.

Coastal areas in Sri Lanka have two sources of energy that are yet to be harnessed; the high incident solar radiation, and the convection wind flows that occur due to the proximity to the ocean. This project involves development a system that will harness these two “free” sources and supply the power to light and ventilate these small houses.

Figure 1: Solar energy incident on surface1

Sri Lanka

Figure 2 : Sea breeze and land breeze2 Figure 3 : Wind data map3

Table 1: Load and power generation calculations

Initial Tests and Calculations

Power (W)

#Usage (hrs)

Energy (W.hrs)

Houses

LED 6 3 6 1083 houses supplied

Fan 10 1 10 100

Total power required with LED bulbs per day 208

CFL 14 3 6 2522 houses supplied

Fan 10 1 10 100Total power required with CFL bulbs per day 352Solar Panel 50 1 6 300

Power Output

Wind Turbine 20 1 20 400Total power supplied by system per day 700

Ametek 38VDC Motor TestingThe ideal wind turbine generator would reach high voltage levels at low RPM’s, while at the same time the torque should remain relatively low. The Ametek 38VDC motor showed the best characteristics and the results of its motor test is shown below in table 2. The table was truncated for the sake of clarity.

Tach (V) Speed(RPM)

Torque(oz-in)

Gen Voltage Out (V)

2.39 114.90 10.14 4.003.15 151.44 10.65 5.303.66 175.96 11.15 6.306.84 328.85 13.35 11.207.04 338.46 13.69 12.007.41 356.25 13.86 12.70

12.50 600.96 17.41 21.2013.25 637.02 18.08 22.4014.35 689.90 18.93 23.90

Table 2: Ametek 38VDC motor test results

Size the LoadDifferent types of lighting such as CFL and LED were considered. A small DC fan will be used for ventilation. It was assumed that the house will be lit for six hours a day, on average; with three lighting fixtures providing the entire lighting requirement for these small dwellings. The wind turbine was assumed to generate 20W of power for 20 hours, while the solar panel generated 50W for 6 hours per day. The resultant power supply and demand is tabulated in table 1.

Wind Tunnel TestingA blade testing apparatus that can be mounted on to the wind tunnel was designed and built. Through this apparatus the rotational speed of different blade designs at varying wind speeds was measured; as well as torque and power generated by the mini-turbine. Figures 4a-b shows the design schematic and the final apparatus.

Figures 4a-d: Design Schematic, Savonius blade, pulley, bevel gear arrangement

5a) Rotational Speed 5b)Power

Torque vs. Wind Speed

0

0.01

0.02

0.03

0.04

0.05

0.06

7 8 9 10 11 12

Wind Speed (m/s)

Torq

ue (N

.m)

5c)TorqueFigures 5a-c: Rotational Speed, Power, Torque measures for a Savonius type blade under different wind speeds.

4a

4b

4d4c

These results predicted that a larger scale Savonius type wind turbine would self start at very low wind speed, but it may have structural issues at very high wind speeds.

Figure 6a-b: Lenz2 Wind turbine simulated in SymLab at 15 mph wind in x directiona) Shows the entire pressure coefficient spectrumb) Shows the areas with higher pressure coefficient

Overall Design

As shown in figure 7, the two components generating power are the solar panel and the wind turbine. They are connected to a charge controller which would be connected to the battery pack. The vertical axis wind turbine is connected to a sprocket and chain with a four to one transmission ratio. This was sized depending on the expected power and torque of the VAWT. The VAWT will drive an Ametek 38VDC motor that would generate a certain voltage and current. If the said voltage is below 13V or excessively higher from that number, the voltage will be converted up or down to 13.3V using a Linear LTM4605ev buck/boost converter. The charge controller will protect the battery from surges and make sure it does not get overcharged. The battery used for testing purposes was a 12V - 12Ah Maintenance-Free Rechargeable Sealed Lead-Acid Battery.

Figures 7: Overall Design Schematic

6a 6b

Final Wind Turbine Design

Testing the Wind TurbineThe finished wind turbine was assembled on top of the Union College facilities building as seen in figure 8c. Now, it’s ready for field testing. Performance tests were carried out on a particularly windy day(16mph average). The experimental setup is seen in figure 9a-b.

The voltage generated by the DC motor averaged at about 5-7V at low-medium winds; and it reached up to 14V. The buck/boost converter regulates the voltage to 13.3 Volts. After the charge controller this voltage dropped back down to about 12.5V. The current delivered to the battery reached up to 2.9A. Therefore, by multiplying current by voltage, the power generated in the system is about 37W.

Figures 8a-c: Design Schematics of the wind turbine, and the final finished product assembled on top of the facilities building

8a 8b

8c

Figures 9a-b: Yellow multi-meter measures Vgenerated , Green-1 measures current supplied to the battery, Green-2 measures Vregulated to the battery. Inside the box you can see the charge controller and the buck/boost voltage converter circuit.

V V A Power (W)

Min 0 0 0 0 No windMax 14 12.6 2.9 36.54 Medium-high wind

Average 5 12.6 1.25 15.75 Medium-low wind

Table 3: Multi-meter readings

Future WorkIntegrate a solar panel to the system. Test the entire system performance over time using a data acquisition device; while keeping track of wind speed. Develop electrical system such that multiple batteries can be charged; while, allowing certain batteries in the battery pack to discharge. Experiment with higher gear ratios. Test other generators and compare data.

Prof. Richard Wilk, Prof. James Hedrick, Prof. John Spinelli, Prof. William Keat, Paul Tompkins , Stanley Gorski , Rhonda Becker , Malysa Cheng, Kevin Donovan, James J. Howard , Roland Pierson

Department of Mechanical Engineering, Union College, NY

Senior Project - 2009