Multi-Fidefity Wind Turbine Simulation and Wind Assessment at Masdar City

Abstract

Wind energy is undergoing unmatched growth amongst renewable energy portfolios with over 25% annual increase in deployment. Since the regional wind pattern has tremendous influence on wind project development, characterizing the wind at a given site becomes an indispensable step. BEM based engineering models are still dominating among wind turbine R&D community due to their low computational requirement and high accuracy. Furthermore, advanced development in computation power, computational algorithms and memory storage render the migration into high fidelity CFD wind turbine simulation. This work began by first, characterizing the wind in Masdar City. Both low and high resolution wind data were collected. Intermittency was identified by initially utilizing FFT then by wavelet analysis to remedy the shortcomings of FFT by preserving the time-scale information. Then the analysis focused on the wind power assessment starting by fitting the wind data with Weibull distribution. Two wind turbines were considered to explore their power generation under the measured Masdar City low density wind pattern. The second part of the work focused on developing a BEM based MATLAB code as a low fidelity tool to assess wind turbine performance. The code was first validated against the NREL phase VI experimental data before applying it to the 3.5KW Windspot. The power generations under different operational conditions were obtained and several key parameters, i.e. TSR, Twist angle, and Pitch angle, were genuinely studied and discussed. Results suggested that designing wind turbine blades with proper twist angle, operating wind turbine under proper TSR and Pitch angle can lead to a substantial increase in power generation. Finally, high fidelity CFD simulation was explored to provide better insight for the flow over the 3.5KW Windspot. The flow domain was discretized using ICEM CFD and the model was comprised of 4.5 million hexahedral elements. The blade was wrapped with high resolution O-grid mesh to achieve reasonable y+ value. The model features SRF formulation and RANS based SST k-ω turbulence model. The results showed good agreement with experimental data and the extended running cases suggest 3.5KW Windspot has a high power coefficient within the TSR range between 4.0 and 6.0.

Date of Award	Dec 2012
Original language	American English
Supervisor	Isam Janajreh (Supervisor)