Planning the future electricity system for Great Britain is a complex task due to the scale of the system and the many uncertainties that surround its future use. As the system grows, while also transitioning to be less carbon intensive, many choices will need to be made by government, the system operators, asset owners and investors. These choices will affect the carbon emissions of the system, the costs and the security of supply for consumers.

This talk will demonstrate how we have used open energy data and tools to develop an AI system to support energy policy planning. We have trained ‘reinforcement learning’ algorithms to optimise a design for the GB power sector to understand which mix of technologies policy makers should incentivise over the next five to ten years. Our aim with this work is to highlight the potential for AI techniques to support energy system planners and policy makers.

We will cover the steps taken to develop the AI system, what we found out and our plans for the future.