Coordinated operation of electric power transmission and distribution systems

In the modern context of seeking sustainability through the large-scale integration of renewable energy, the amount of renewable resources located in the distribution part of the electric power network has been growing and becoming a progressively important component of the electric power supply chain. These renewable resources are originating mainly from rooftop solar panels. The distribution part of electric power supply chain is composed of low-voltage transmission lines to which residential and commercial loads are connected, unlike the transmission part of the network which consists of high-voltage transmission lines, to which large generators and industrial loads are connected. For the moment, only the transmission part of the system is taken into account in the optimal scheduling of power systems. This creates an opportunity to optimize the low-voltage part of the network in order to generate savings in the consumption of energy. Indeed, the distribution network has an important impact on the scheduling of the consumption of energy. However, the operation of the network's distribution part is in some aspects more complex than the one of the transmission network, since the linear approximation of Kirchhoff’s laws does not provide an adequate representation of the physical laws governing the flow of power, voltage needs to be controlled and can not be assumed to fluctuate around its nominal value, and reactive power needs to be accounted for explicitly. The goal of this thesis is to analyze if the co-optimization of the transmission and distribution part of the network justify the required increase in computational effort. For this purpose, the first chapter of this document is devoted to the settlement of all the optimization models needed for the remaining part of this thesis. An example also proves the possibility of an improvement in the operation of the electrical network due to the co-optimization. After that, some distributed algorithms are presented in order to deal with the co-optimization problem in a distributed fashion. Among these algorithms, the ADMM algorithm, the Caramanis and the Level method are presented. The final objective of this project is to compare various algorithms for performing co-optimization, on large-scale power systems, in order to determine the resulting efficiency gains and assess the scalability of the developed algorithms for networks of realistic scale.