Date of Award

Spring 1-1-2018

Document Type

Dissertation

Degree Name

Doctor of Philosophy (PhD)

First Advisor

Julie K. Lundquist

Second Advisor

George H. Bryan

Third Advisor

Katja Friedrich

Fourth Advisor

Peter Hamlington

Fifth Advisor

Branko Kosović

Abstract

Extreme fluctuations of wind cause damaging loads on wind turbines and pose significant difficulties in wind power forecasting. Increased global interest to move from conventional to renewable energy sources highlights the importance of successful wind energy integration into the electrical grid. Knowledge of extreme atmospheric phenomena that may affect wind turbines and the power they produce is critical to a wind farm’s success. This dissertation addresses two types of extreme events: extreme winds in the hurricane boundary layer, and forecasting challenges of rapid changes in wind speeds, or wind ramping events.

First, this dissertation quantifies the extreme wind conditions that future offshore wind turbines, placed in hurricane-prone regions, could experience. The analyses include novel large-eddy simulations of major hurricanes evaluated with real hurricane reconnaissance flights. Investigations into turbulence characteristics, mean wind speeds, 3-s gusts, wind veer, and wind shear at different locations of the hurricane provide first-ever estimates of how these variables could impact wind turbines. Comparisons of these variables to current wind turbine design standards suggest that modifications to the standards may be required to represent extreme hurricane conditions.

Work presented herein also addresses the difficulty in predicting power ramp events, which are large and abrupt changes in wind power production. These ramp events often lead to costly fees imposed on wind farm operators and decreased reliability of wind as an energy source. An exploration of four multivariate statistical post-processing techniques is performed to improve power ramp forecasts by the High-Resolution Rapid Refresh model. Specifically, the techniques seek to reduce under- and over-forecasting biases of ramp events and provide valuable uncertainty information through the generation of numerous forecast scenarios.

All of the research discussed herein alludes to modifications of current turbine design or forecasting practices to account for extreme wind conditions. Therefore, recommendations to change design codes and to increase forecast skill are provided throughout the main chapters.

Share

COinS