Integrated modeling of active demand response with electric heating systems coupled to thermal energy storage systems

• ADR effects can only be quantified considering the interactions between supply and demand side.
• An integrated modeling approach for ADR was developed.
• A comparison among the integrated model and simplified approaches is performed.
• Merit order models combined with a physical model of the demand side yield high quality results at a low computational cost.

Active Demand Response (ADR) can contribute to a more cost-efficient operation of, and investment in, the electric power system as it may provide the needed flexibility to cope with the intermittent character of some forms of renewables, such as wind. One possibly promising group of demand side technologies in terms of ADR are electric heating systems. These systems could allow to modify their electrical load pattern without affecting the final, thermal energy service they deliver, thanks to the thermal inertia in the system. One of the major remaining obstacles for a large scale roll-out of ADR schemes is the lack of a thorough understanding of interactions between the demand and supply side of the electric power system and the related possible benefits for consumers and producers. Therefore, in this paper, an integrated system model of the electric power system, including electric heating systems (heat pumps and auxiliary resistance heaters) subjected to an ADR scheme, is developed, taking into account the dynamics and constraints on both the supply and demand side of the electric power system. This paper shows that only these integrated system models are able to simultaneously consider all technical and comfort constraints present in the overall system. This allows to accurately assess the benefits for, and interactions of, demand and supply under ADR schemes. Furthermore, we illustrate the effects not captured by traditional, simplified approaches used to represent the demand side (e.g., price elasticity models and virtual generator models) and the supply side (e.g., electricity price profiles and merit order models). Based on these results, we formulate some conclusions which may help modelers in selecting the approach most suited for the problem they would like to study, weighing the complexity and detail of the model.