SUBJECT: M.S. Thesis Presentation
   
BY: Michael Schmidt
   
TIME: Wednesday, November 7, 2007, 1:15 p.m.
   
PLACE: Tech Square - Management Building, 4249
   
TITLE: The Economic Optimization of Wind Turbine Design
   
COMMITTEE: Dr. Samuel Shelton, Chair (ME)
Dr. Bert Bras (ME)
Dr. Sheldon Jeter (ME)
 

SUMMARY

This thesis studies the optimization of a variable speed, three blade, horizontal-axis wind turbine. The design parameters considered are the rotor diameter, hub height and generator capacity. The levelized cost of energy and simple payback are the figures of merit being minimized. Blade element momentum theory is used to calculate the power produced by the wind turbine rotor. Increasing the rotor diameter increases the power delivered to the generator at all wind speeds up to the limit of generator capacity. Increasing the generator capacity raises the limit on maximum power output. Increasing the hub height of a wind turbine increases power output due to the higher wind speeds at increased heights. However, all of these design changes involve an increase in capital cost. Furthermore, wind characteristics vary between wind resources. Therefore, the optimal wind turbine design will change depending on the wind resource. The model developed in this thesis is used to minimize the levelized cost of energy for various wind resources. The results of this study provide a guideline for the optimum wind turbine design in various wind resources. The model is also used to compare the difference between minimizing the levelized cost of energy and minimizing simple payback of a wind turbine located off the coast of Georgia. Simple payback is calculated by considering not only the total annual electricity produced and capital cost of the turbine but also the revenue the turbine will generate. Revenue is calculated from a time-dependent valuation of electrical power. The results of this study show that minimizing levelized cost of energy and minimizing simple payback result in the same optimum design for this particular site. The results show, however, that using a time-dependent valuation of electricity results in a different simple payback than when an average value of electricity is used.