Engineering an Optimal Wind Farm

Engineering an Optimal Wind Farm

Mahulja, S.

Larsen, G.C. (mentor)
Elham, A. (mentor)

Aerospace Engineering

Wind Energy

European Wind Energy Master

2015-09-25

The project covers creation of a framework for optimisation of wind farm layouts. The optimisation is done with an objective of maximising the Net Present Value of the farm. With that aim, both power production and the fatigue driven component degradation are considered. The work is split in two main parts. First, in order to circumvent the use of full DWM model simulations, surrogate models of lifetime equivalent loads and power production are created. In the second stage, an optimisation platform is assembled as a two-stage, multi-fidelity strategy encompassing genetic algorithm for global and gradient based algorithm for local optimisation. The analysis of a wind farm is performed using DWM surrogate model which significantly contributes to the speed of computations. On the other hand, the use of surrogates makes it possible to capture the in-stationary effects of wake meandering with sufficient quality. The objective of the thesis is to show the steps required in assembling a framework such as this. Moreover, through the tests performed on Middelgrunden wind farm, the aim is to show the importance of setting the number of turbines as a design variable.

Windenergy
optimisation
DWM model
surrogate modeling
wind farm

http://resolver.tudelft.nl/uuid:01b7f1f5-bdad-4a44-a028-78fdd7960f8f

Student theses

master thesis

(c) 2015 Mahulja, S.