Coordinating residential flexibility resources in a socio-technical co-simulation design (CO-FLEX)

 In a system with high penetration of renewables and electrification of heating and mobility sectors, Demand-Side Management (DSM) solutions- that is, technologies, actions and programs on the demand side that aim to change the volume and timing of energy demand to optimize the system overall- are considered a cost-effective way to guarantee power supply-demand balance and avoid grid congestion. However, although energy consumption and flexibility are shaped by intertwined socio-technical dynamics, current assessment and optimization methods still tend to neglect social components. As a result, methods of aggregating and managing distributed flexibility also lack proper emphasis on these aspects, resulting in unrealistic and optimistic quantification of residential demand flexibility at grid distribution level.

In order to provide the basis for a more accurate and reliable assessment of residential demandflexibility, here we developed a modeling framework capable of simulating energy demand through a socio-technical model and coordinating its flexibility. Specifically, we coupled an existing high resolution socio-technical model of residential energy demand and flexibility (i.e., demod) with a new flexibility aggregator model (i.e., coflex) through the mosaik co-simulation framework.

The co-simulation design proposed here has been shown to handle a wide variety of types of DSM schemes, targeting consumers at the individual or community level through implicit (e.g., price-based) and explicit (e.g., incentive-based) stimuli. In addition, due to the use of mosaik, the modeling architecture is highly modular and adaptable, allowing alternative and complementary models to be coupled, and their performance to be compared without having to change the core of the simulation architecture. We have also shown that the proposed solution has good scalability properties and can be used for large-scale scenario analysis. For example, a full-year simulation with 100 households takes about 1:10 hours or a 4-day simulation with 8192 households takes 4:35 hours. Finally, by publishing open source the co-simulation design and the new flexibility aggregator model coflex, we hope that this tool can be used to test new flexibility aggregation and control strategies, simplifying the analysis and comparison across modelling approaches and scenarios.