Low-carbon power system planning under uncertainty
Project Leader: Pierluigi Mancarella
Staff: Dr Sebastian Puschel
Collaborators: Dr Alysson Costa (Mathematics and Statistics)
Primary Contact: Pierluigi Mancarella (email@example.com)
Keywords: machine learning; optimisation; Smart Grids
Disciplines: Electrical & Electronic Engineering
Future power systems dominated by renewables (primarily wind and solar) and other low-carbon technologies (eg: battery energy storage systems, electric vehicles, etc) are subject to huge long-term uncertainty when it comes to infrastructure requirements. In particular, new technologies such as various distributed energy resources, microgrids, virtual power plants, etc, could potentially displace ‘conventional’ transmission and distribution network assets; however, from a planning perspective it is unclear what methodology would be most suitable to allow a like-for-like comparison of network and ‘non-network’ solutions, what the role played by uncertainty is, and how investment risk could be mitigated though development of an optimal technology mix.
On these bases, following on work conducted for the UK National Grid, this project aims at developing a planning tool for optimal infrastructure development under long-term uncertainty which considers network, storage, and different smart grid technologies. The modelling will be based on stochastic optimisation, real options theory, and risk-constrained programming, and consider state-of-the-art mathematical programming, decomposition techniques, and machine learning algorithms suitable for parallel and cloud computing.