Address:
Ms Sara Fakih from the Energy Systems and Environment Department (DSEE) and GEPEA laboratory, will present her research about
"Optimization of Renewable Energy Sources and storage integration to support existing electricity grids considering variable and uncertain demand"
Due to the energy transition worldwide electrical energy systems are subject to a set of deep transformations in the energy production and distribution systems illustrated by increased usage of decentralized renewable or zero-carbon emissions production units. In particular, electricity generation from renewable sources (hydro, wind, and solar PV) is on track to grow strongly around the world to meet the forecasted increase in global electricity demand and to fulfill the local authorities’ commitments to energy transition policies. The hard prediction and control of these renewables increase the complexity of the electrical system planning process. To tackle these complexities, Electrical Power System Models (EPSM) for planning and operation optimization are developed. In this thesis, a dynamic linearized AC-optimal power flow (DLOPF) model is developed based on a
successive linear programming resolution approach of the AC OPF model. In the first step of the work, the DLOPF model enables the identification of the need for power production for each bus of the grid on an hourly basis by using the concept of virtual generators. This information helps generate and assess interesting scenarios of renewable energy and storage unit deployment. This methodology is tested on a case study of an undersized network in the context of increasing electricity demand. In a second step, based on the DLOPF model coupled with an optimization stage, an EPSM is developed following a bi-level approach. This model optimizes the size and planning of Renewable Energy Sources (RES) and Battery Energy Storage (BES) while taking into account cost minimization and carbon constraints. The upper level implements a Particle Swarm Optimization (PSO) model and the lower level the DLOPF. The PSO sizes and identifies the placement of the BES systems and DLOPF places and sizes RES in a spatial-temporal framework. This combination helps to dispatch power generation and storage in a network in the context of increasing demand, avoiding the need to reinforce the grid. Finally, an Uncertainty Analysis (UA) coupled with a Sensitivity Analysis (SA) is used to assess the impact of demand uncertainty (due to its unpredictable variability) at each bus on the EPSM and the proposed grid performance indicators.
Organizer(s)
Thesis acreditation from IMT Atlantique with the Doctoral School SPIN
Key-words: Energy planning, Bi-level optimization, Dynamic linearized OPF, Electrical network, Renewable energy sources, Battery energy storage, uncertainty analysis