A Multi-Agent Framework for Smart Grid Simulations: Strategies for power-to-heat flexibility management in residential context

Abstract

The complexity of a system, such as the Smart Grid, comes from the interaction of different fields of expertise and actors cooperating together. As a result, co-simulation frameworks are emerging to meet the need for analysis from a broader perspective. The feature to better understand the Smart Grid environment is the ability to simulate its different components simultaneously. Therefore, we propose a multi-agent co-simulation framework that can serve as a test-bed for multiple Smart Grid strategies. In particular, we focus on Demand Response programs that exploit the thermal behaviour of residential buildings. The proposed framework is modular to test Plug&Play models. Moreover, it is highly flexible and configurable to evaluate realistic scenarios. We tested the platform on a case study of 1,000 buildings, performing an analysis of the effects of micro temperature deviations in buildings on the primary grid substation balancing problem. The results show the flexibility of the platform in testing different strategies. Moreover, power imbalances could be mitigated simply by acting on the indoor temperature set-points with smaller deviations.