Improved Design of a Large Offshore Wind Farm by Using Biogeography based Optimization

Abstract

This paper investigates optimal placement of wind turbines (WTs) within a large offshore wind farm (LOFWF). 88 wind farm (WF) configurations are invested to search the optimal layout for the Horns rev 1 offshore WF using twenty years of wind data knowing that wind characteristics are modeled from long term reanalysis data based on MERRA-2. The regular and the irregular placement of the Horns Rev 1 (HR1) offshore WF are investigated with Jensen wake model. Therefore, the objective is to assess the effect of wind turbine spacing (WTS) on the power output loss in a LOFWF and, also to find the best configuration that gives the maximum power with minimum investment cost. The use of the Biogeography based optimization (BBO), as a bio-inspired evolutionary approach, represents the advantage of being effective on strongly non-convex spaces such as a large offshore WF. Due to the iterative process applied to the initial population, and the multiplicity of the population, the BBO process limit the risk of getting stuck in a local optimum, by distributing the individuals in the whole solution space. The results obtained show that the wind data extracted from MERRA-2 can be applied reliably to any existing WF to simulate wind power production. The results also demonstrate that significant power losses occur when the turbines are arranged in a condensed manner and the significant power gains are not obtained from too large configurations, but rather from the best placement in configurations. The proposed approach shows promise in terms of applicability with MERRA-2 and is effectively suitable for arranging turbines and assessing wind resources in an offshore wind farm project (OFWFP).