Analysis of Transitions and Investment Strategies under Uncertainty for Smart Grids.

Summary Smart grids are considered as an efficient way to accommodate more renewable energies, to better control demand, to improve the operating conditions of the electrical system, to increase its performance and to facilitate the development of new uses such as the electric vehicle. However, these potential benefits of the smart grid also bring uncertainties for the electrical system and its actors. The thesis is part of the GreenLys project (a smart grid demonstrator addressing innovations from the energy producer to the consumer, including the actors of the transmission and distribution network). The objective of this study is to propose technical and economic steps for the transition to the smart grid by 2050.In the perspective of this thesis, three uncertainties that can significantly influence future investment strategies are addressed in three separate models. The three models are applied to the GreenLys scenarios (a conservative 'Grenelle' scenario that respects the European energy commitments and a more ambitious '100% RE' scenario that aims at 100% renewable electricity production) to propose smart grid investment tiers and recommendations. First, the uncertainty in the regulation of the public distribution network is studied in a model using a real options approach combined with a fuzzy logic algorithm. Second, a real options approach based on a classical binomial tree is used to analyze the uncertainty on the evolution of the flexibility pool. Finally, the uncertainty on the information system costs is modeled in a Monte-Carlo based algorithm.