In the context of increasing renewable penetration and of growing electrification, the electricity grid faces challenges such as costly extension, progression of digitalization and need for resilience. The rapid deployment of controllable loads such as heat pumps, electric vehicles, and home energy storage, combined with smart metering infrastructure, enables coordinated energy usage to relieve grid stress. As of today, numerous approaches exist at different levels to leverage this flexibility potential, such as dynamic tariffs, energy management systems, and energy communities. This work is part of the research project EnQuaFlex that aims to utilize the aggregated flexibility of a neighborhood in a grid-friendly manner. This paper focuses on the development and testing of the Orchestrator, a decision-making tool that acts as an overarching energy management system to coordinate the available flexibility from controllable loads and prevent grid congestion.

In this work, the effect of aggregated neighborhood flexibility on the distribution grid is evaluated by PyPSA simulations of a synthetic medium voltage grid. The interaction between different models and simulations is achieved with the AgentLib Python framework for multi-agent systems. Three main conceptual components emerge: First, the energy system—modeled in simulation or reproduced in a lab environment—represents the neighborhood, while the medium voltage grid is modeled with PyPSA. Then, the flexibility quantification relies on model predictive control optimization to determine the possible deviation compared to a cost-optimized baseline scenario for controllable loads and expresses this flexibility as a standardized FlexOffer. Finally, the Orchestrator decides whether to activate flexibility offers, based on threshold values for electricity prices, weather data and grid status, before propagating that information back to the energy system. The interplay between these different components is tested through a simulation, focusing on the Orchestrator’s decision-making process and grid stress relief effectiveness. Additionally, the energy system was implemented as a Power Hardware-in-the-Loop lab setup to demonstrate real-world applicability.

Results demonstrate that the usage of the Orchestrator reduces the peak load of the neighborhood and diminishes the amount of grid violations. A sensitivity analysis on the decision-making threshold values reveals a trade-off between grid relief effectiveness and flexibility activation costs. Hence, the Orchestrator successfully coordinates neighborhood flexibility while balancing grid stability objectives with economic constraints. The demonstrated modularity—from simulation to lab implementation—enables reuse in similar contexts and potential extension to field deployments. This work contributes to an efficient usage of the available flexibility in the distribution grid, which is an important step for the transition towards a more resilient power system.