Under the expected decarbonization of the power system (where high voltage network use is diminished due to the expected increase of self-satisfied DG), the economical sustainability of the electrical grid activities might put at risk without increasing the electricity tariffs. New models are therefore necessary to properly address and study the design of these tariffs under this zero-emission scenario. In this paper a mathematical bilevel model is proposed to achieve this aim. While the objective of the upper level is to obtain the long-term evolution of the tariffs needed to sustain the grid activities, the second level focuses on the representation of the electricity generation activity. In particular, in the lower level a generation expansion problem is formulated by means of the maximization of the traditional GENCOs' profits and the minimization of the DG' costs, both simultaneously and in the sense of Nash, taking into account the energy and reserve commodities, the network tariffs of DG customers and the classical UC constraints. Since the proposed model is a non-convex bilevel mathematical problem that has been applied to a realistic large scale power system, the resolution methodology is based on a sensible heuristic. It uses parallel computing for each solution evaluation of the first level while satisfying the equilibrium conditions of the second level and without setting its KKT conditions. The case studies ?
Keywords: Bilevel models, Nash equilibrium, generation expansion, distributed generation, parallel computing, heuristic optimization.
Registration date: 2020-01-14