Civic AI Toolkit: Participatory energy blueprint

www.nesta.org.uk/toolkit/civicai-energy/
Skip to content

Civic AI Toolkit: Participatory energy blueprint

Community energy could play a vital role in decarbonising the national grid and boosting local economies. Yet since government support through the feed-in-tariff ended in 2019, the sector has struggled, relying heavily on volunteers to set up, administer and operate the projects. Current installed capacity in the UK is 264.9 MW, representing less than 1% of the total electricity production. If communities are to help drive the energy transition and dramatically increase installed capacity, they require new tools and methods to help organise, fund and replicate community energy projects.

Participatory energy - overview layout.png

This blueprint uses augmented collective intelligence to address the following key challenges:

  • Reducing admin and operation costs - reducing the administrative burden of setting up and operating projects.
  • Capacity & expertise sharing - sharing expertise to more easily sustain and replicate projects without relying on volunteers.
  • Modelling impact - measuring economic and social impact for individual and aggregated projects.

How it works in practice

In this use case, machine learning algorithms help to remotely identify promising sites during the planning stage of a project. AI is also used to model financial and social outcomes for new community energy projects by learning from the measured impacts of completed projects. Once a project has been funded through a crowdfund or community share offer, digital ownership contracts link directly to the title deeds of the freeholder or leaseholder to facilitate easy transfer if the leaseholder changes. When up and running, a digital twin of the network helps to balance local demand and supply. Open data standards ensure that the data remain accessible by and useful to multiple stakeholders across different software platforms. People and machines act collaboratively to manage energy use, with AI agents helping to shift consumption to times when there is excess local supply through auto-scheduling (e.g. charging an EV overnight) or subtle prompts. Predictive models forecast demand and supply to help identify mismatches, and combine this with other data to inform the community members about the likely lifetime revenues and social impact of the project. This is fed back into the system to help estimate the impact of other planned community energy schemes to de-risk upfront investment and subsequent operating costs.

It will be critical that the digital twin of the local energy system and any connected devices or appliances is technically robust and secure. All electricity consumption and supply data, across a community, must adhere to strict protocols to protect householder privacy and the predictive modelling functions will require explicit attention to human agency and oversight as well as accountability, to ensure reliable and fair comparisons can be made across multiple projects.