Multilayer Agent-Based Modeling and Social Network Framework to Evaluate Energy Feedback Methods for Groups of Buildings

Abstract

Important energy reductions can be achieved in the building sector by providing occupants with feedback about their energy-consumption levels. Recent studies link the success of energy-feedback methods to the level of occupant engagement with people in their social circles and the resulting peer pressure to conform to certain social norms. Despite promising results, the literature remains limited in scope to individual rather than groups of buildings. This has limited the design of feedback initiatives leveraging social connections that exist, or that can be induced, within and between buildings. The current paper addresses the identified gap by proposing a multilayer agent-based model that serves as a test bed to simulate and optimize feedback methods applied on any building stock (e.g., community and city). Monte Carlo and sensitivity analyses show that connecting occupants of different buildings, while increasing their engagement with the feedback mechanism, can lead to energy reductions exceeding 10%. The findings confirm the role of social networks in energy-conservation diffusion, setting the stage for large-scale and socially engaging energy-conservation initiatives.