A novel circle-packing NLP model for offshore wind farm layout and cable optimization

Abstract

The proper siting of wind turbines and cable routing in offshore wind energy systems can be used to prevent wake effects and electrical losses. This study adopts a continuous-domain optimization model based on nonlinear programming for offshore wind farms' simultaneous layout and electrical placement. In particular, it uses a circle-packing formulation to optimize the placement of turbines within a flexible, unconstrained spatial domain while incorporating a radial cabling strategy to evaluate and minimize power losses. In particular, the model positions turbines in a flexible, unconstrained spatial domain with circle packings and implements a radial cabling strategy to evaluate and minimize power losses. The method considers a Gaussian-based wake model and losses due to dips and resistivity of the electrical cables. The resulting model was solved using a global NLP solver (GAMS/BARON) for several scenarios. The result shows a 43 % reduction in the occupied area and a 0.42 % decrease in annual energy production. It also has more spatial compactness, shorter cable length, and more stable performance than the traditional grid-based and heuristic models. Its scalable and flexible formulation makes it suitable for planning offshore wind farms in earlier stages.