Data Science applications in whole energy systems

Main Organizers: Harry van der Weijde, Ken McKinnon, Richard Gibbens

Dates: 28-29 January 2016
Location: Edinburgh

A secure, sustainable, and cost-efficient energy system is of crucial national importance. To understand how such a system can be realised and needs to evolve, there is a clear need for integrated energy system models that consider whole energy systems, including energy demand, energy transport, energy supply, and interactions between these sectors. This workshop will examine how Data Science concepts, methods and analytics can be applied to further these modelling efforts, and explore in which areas new methods have to be developed, in such a way as to allow system operators, policy makers, regulators and the private sector to use the vast amounts of data coming from all elements of the energy system in an integrated, efficient and effective way to make reliable and robust decisions.
To do so, it will a) explore how the Data Science concepts and methods mentioned above can be applied to whole energy systems and in which areas theory needs to develop to make the greatest contribution to practical problems, and b) explore the needs for a UK energy systems data repository, which would hold raw and processed data of all parts of the energy supply chain.

Areas that will be discussed include:

• Optimal placement and use of sensors and other methods to gather data, including large-scale estimation.

• Filtering, reliability screening, and processing of large, uncertain datasets, including fusion of data from different energy vectors or different parts of the energy supply chain and from meteorological sources.

• Methods for integration of new data sources in energy system operations and investment models and other energy applications.

• Implications of data uncertainty for optimal, robust and risk aware decision making for whole energy systems operations, and quantification of the value of data sources and improved data quality.

For more information, please visit: