PhD#5 at Mines Paris in Data Science & Energy: "Optimization of flexibility services under multiple local uncertainties in the context of smart grids"

Title:  "Optimization of flexibility services under multiple local uncertainties in the context of smart grids"

Context and challenges:

In the context of the energy transition, power grids integrate massive amounts of renewable generation (mostly wind and solar) whose volatility and uncertainty bring unprecedented challenges to the grid operation. Flexible generation and demand, as well as storage or storage-like resources, are key for the efficient and reliable management of future power systems. The quest for flexibility is paramount at different temporal but also spatial scales (at the transmission level, at the perimeter of an aggregator or, more locally, at distribution networks) and has expanded to multi-energy systems, e.g., the coupling between electrical and gas networks. Existing methodologies that propose flexibility indicators at a national level need to be revisited at the local level, by considering local characteristics and uncertainties in production and demand at a given territory (district, region), and accounting for events that deviate from normal operating conditions (e.g., peaks due to electric vehicle charging, low renewable availability during long periods, etc.). In a given territory, flexibility valorization raises a multitude of territory-specific questions, e.g.: Should a local flexibility market be deployed? What is the potential of local energy communities? How do local conditions affect territory-level decisions for the flexibility provision and use?

Main objective of the thesis:

The overarching objective of this research project is to develop an approach for the optimal provision and use of flexibility at the level of a territory, which accounts for the uncertainties associated with local renewable production and local energy consumption of the potential flexible consumers (residential, commercial, industrial).

Methodology and expected results:

The first step of this research project is to define flexibility provision indicators, based on production/consumption adequacy and contextual assessment at the level of a territory, relying on predictive methodologies to quantify the local flexibility potential. These indicators will be used as inputs to produce a risk-aware analytical decision-aid methodology of flexibility valorization in multi-energy systems (second step), employing forecasting models and optimization. The third step is to simplify the arguably complex modelling chain by integrating forecasting and optimization via end-to-end learning of flexibility decisions based on AI, thus predicting directly flexibility decisions that are optimal as a function of the predicted local weather conditions (e.g. uncertain mixed-cloudy day) and the local socio-economic context (e.g. high commercial activity expected due to fair/sales etc.).

Funding category: Autre financement public
Project PEPR TASE "FLEX TASE"
PHD title: Doctorat en Énergétique et Procédés
PHD Country: France