Modelling and Control of Active Distribution Networks
M. Antonin Colot will publicly defend his thesis entitled "Modelling and Control of Active Distribution Networks".
Summary
Modern power systems face massive integration of distributed energy resources (small and modular energy generation or storage) interfaced with electrical distribution networks through power electronic converters. Active distribution networks are the result of this integration, where the distribution network is no longer a passive load sink, but actively participates in the power system operation.
Power converters are electronic devices that enable the integration of renewable energy resources in power systems by converting electrical energy from one form to another. The increased integration of these devices alters the power system dynamics; the fairly slow electromechanical phenomena driven by synchronous machines are now dominated by the complex fast-response of power converters. Traditional control strategies are not able to cope with the new dynamics, and may fail at guaranteeing the safe operation of the electrical network. Furthermore, power converters usually interface renewable energy production, or flexible loads with the grid, and their intermittent nature requires new fast control strategies to continuously track the time-varying grid conditions.
In the first part of this thesis, we derive new modelling tools to represent the dynamic behavior of power converters. In particular, we focus on models that are particularly well-adapted to system-level studies. They are designed to analyze the system's dynamics in different operating conditions. Also, they are computationally lightweight compared to detailed models, which allows for simulations of large systems with many different assets.
In the second part, we consider the problem of controlling inverter-interfaced distributed energy resources to ensure voltage regulation in a distribution network while minimizing resource usage. Voltage issues can occur due to the massive integration of renewable energy resources. In particular, excessive production of solar energy in residential low voltage networks can lead to overvoltages.
We first propose a centralized online feedback optimization method that drives controllable power setpoints to a solution of an optimization problem, while ensuring anytime satisfaction of voltage constraints. We then propose a decentralized incremental Volt/Var control strategy, where the gains are derived in order to minimize resource usage and satisfy voltage constraints with a prescribed probability. The optimization problem is solved centrally, but the method requires only occasional communication. Finally, we propose a distributed controller where an optimization problem is decomposed and solved node by node, with only local communication between neighboring nodes.
New advanced control strategies run at a faster timescale, so that there is no longer a timescale separation between the plant dynamics and the controller dynamics. This can lead to unstable closed-loop systems that should be investigated with care. Modelling tools help to validate new control strategies and ensure the secure operation of the power system. These are crucial aspects for the integration of new control strategies in the operation of active distribution networks, that would allow for a greater integration of renewable energy resources in distribution networks.
Practical information
Defence will take place on Friday May 23rd at 16:00, to all at amphitheatre Mania Pavella of Institute Montefiore (Bât. B28, au Sart Tilman) or via le FSA PhD Channel.
