9th International Conference on Computational Management Science, Londres (Reino Unido). 18 abril 2012
Electrical layout design is a critical element for offshore wind farms, and one of the main drivers of both cost and reliability. The model OWL (Offshore Windfarm Layout optimizer) produces optimal electrical layouts in reduced computation times. OWL considers all the most relevant design possibilities and factors, with the inclusion of HVDC connection and consideration of stochasticity in both wind inputs and component failures as well as losses. A real case study demonstration of the algorithm has been carried out based on Barrow Offshore Wind Farm in the East Irish Sea. OWL produces a substantially improved design, with total realizable savings of EUR 800k over the actual implementation of the farm. In addition, the optimal layout presents a remarkable lack of symmetry and redundancy in certain elements, supporting the need of a full optimization rather than the selection of a pre-established configuration. OWL exploits the structure of the problem to improve its MIP implementation with Benders’ decomposition. In addition, efficiency has been increased by scenario aggregation and the dynamic addition of partially relaxed cuts. Moreover, an algorithm that has been named Progressive Contingency Incorporation accelerates convergence by solving a series of simplified instances of the same problem.The developed mechanisms reach time savings of two orders of magnitude.
Palabras clave: offshore wind farm, stochastic optimization, Benders decomposition.
Fecha de publicación: abril 2012.
S. Lumbreras, A. Ramos, Decomposition strategies applied to the optimal design of offshore wind farms, 9th International Conference on Computational Management Science, Londres (Reino Unido). 18-20 Abril 2012.