This thesis aims at proposing novel bilevel models to tackle the generation expansion planning problem in liberalized electricity markets, to point out the difference and advantages that such an approach has with respect to single-level models, and to solve realistic instances of such bilevel models satisfactorily, which can be challenging since bilevel models are known to be complicated. Ever since the liberalization of the electricity sector, generation expansion planning has moved away from a centralized planner and has become the responsibility of generation companies, which has greatly increased the complexity of the investment decision process. Therefore generation companies require the most adequate tools to assist them to take and evaluate generation expansion decisions while coping with an uncertain and highly competitive environment. Therefore, in this thesis we propose bilevel generation expansion models which allow us to represent a sequential decision making process where first capacity decisions are taken and then there is the market clearing, as opposed to simplified single-level approaches which assume simultaneous decision where both capacity and production decisions are taken at the same time.
Universidad Pontificia Comillas. Madrid (España)
26 de junio de 2013