What should electricity load profile of new supermarkets look like?

What should electricity load profile of new supermarkets look like?

Senior researcher Ramon Granell presented a paper at the 2nd International Conference on Sustainable Energy & Resource Use in Food Chains, in Paphos, Cyprus. The conference ran from 17th to 19th October 2018.

The paper is "A data-driven approach for electricity load profile prediction of new supermarkets"; authored by Ramon with co-authors Professor David Wallom (University of Oxford Department of Engineering Science) and Colin Axon and Maria Kolokotroni (both Brunel University London).

The work asks what the typical daily electricity load profile of a new supermarket should look like.

Typical approaches to estimating electricity load are based on thermal engineering models of the building, however the group’s approach is a data-driven method to find stores with similar (though not identical) basic building and retail-related features and combine their profiles.

The method was to use a modification of the 'K-nearest neighbours algorithm' to identify similar stores, selecting the K-nearest and calculating the profile of the new store by averaging the profiles of these K stores. The approach was evaluated over a real data-set of 196 stores.

Future work will focus on improving the model with more sophisticated methods to combine the K most similar stores, incorporating external temperature and accounting for uncertainty in the predictions. The group plan on analysing gas consumption in a similar way and testing over other supermarket data-sets.