Smart grids and buildings : Master thesis

 

The final semester is exclusively dedicated to the End of Studies Project.
This occurs from February to June, either overseas or in France, in a research laboratory or in industry. 

This allows students to use the methods and knowledge they have acquired in a professional environment.  They will experience engineering working methods and learn how to manage a project, test solutions and develop their autonomy and initiative.

Renewable energy sources, especially wind generators are considered as an alternative to generate electrical power, as a non-polluting source and thus low environmental impact. The big boom of Distributed Generation (DG) has motivated the development of methodologies to evaluate their effects on the electrical system behaviour, where the distribution networks contain both uncertain generation and load.
Whereas the power supplied by sources of generation take advantage of wind energy, which is highly unpredictable, (as a result of the unpredictable nature of weather conditions), it is appropriate to consider this uncertain characteristics to plan and operate power systems. For example, in order to assess the technical impact of the wind parks integration into the generation of a power system.
In this research work the Monte-Carlo Simulation (MCS) approach is used to resolve the probabilistic load flow problem, based on the modelling of the one dimensional marginal distributions. A second approach is also explored based on the solution of linear equations of the load flow by using the probabilistic arithmetical laws. This last approach is possible by supposing a power system without losses. To illustrate the application of these methods and validate results, the IEEE 28-bus radial system is employed.


ALVAREZ Marie-Cécile, G2Elab
Architecture of the future distribution grids: long term planification

Deregulation of electricity and the increasing will of protecting environment will enhance the development of dispersed generation. Transport networks, already saturated, need to be helped by distribution networks where dispersed generation will be connected. But a massive introduction of dispersed generation could modify the behaviour of distribution networks. The present report shows a way to integrate those productions:  finding new ways to switch electrical flows. Then, knowing the load, we have developed an automatic tool searching the optimal structure between the radial scheme, the grid, and the meshed scheme. Those structures have been selected regarding their ability to be looped and also to evolve with the load and the insertion rate of dispersed generation.