Quantifying EnergyNet performance: a simulation-based framework for decentralized energy networks

Preprint, 2026

Modern power systems are increasingly exposed to severe disturbances caused by extreme weather, infrastructurefailures, and geopolitical conflicts. Conventional power systems rely on interconnected transmission and distributioninfrastructure to connect generation and consumers, and failures of this infrastructure can disconnect large groups ofconsumers from their energy supply until service is restored. At the same time, increasing amounts of local generation,storage, and controllable loads are being deployed throughout the power system, creating new possibilities for howelectrical energy systems can be organized and operated.EnergyNet has been proposed as a decentralized, software-defined energy architecture based on interconnectedpower electronic Energy Routers. Each router coordinates local loads, generation, storage, and grid connections, whilebidirectional Energy Links enable energy exchange between routers. Energy exchange is coordinated through theEnergy Protocol, allowing routers to communicate energy needs and availability and establish energy transfer acrossthe network. EnergyNet can operate together with the conventional grid, independently from it, or across combinationsof grid-connected and isolated regions.This paper develops a quantitative framework for modeling and simulating EnergyNet systems and calculating theirperformance under different operating conditions and disturbances. The framework represents network topology,prioritized loads, generation, storage, grid connections, Energy Links, and component constraints. Optimization withina time-stepped, receding-horizon simulation calculates feasible energy allocation and exchange. The same frameworkis also used for disturbance analysis, long-horizon reliability screening, and probabilistic consequence analysis.The framework is first illustrated through a three-router case, demonstrating energy sharing, re-routing, transfer limits,and islanded operation. This case is then extended to demonstrate coordination of a microgrid at the StanfordUniversity campus and an associated off grid microgrid located in Half Moon Bay using an electric bus from the StanfordUniversity bus service as mobile energy resource; under simulated low-solar conditions, scheduled visits support theoff grid microgrid while preserving the primary transportation function of the bus.Finally, the framework is applied to a practical, under-construction neighborhood-scale EnergyNet in Lund, Swedenwith ten buildings comprising 278 apartments. Full electrical service is maintained during representative summer andwinter operation and following the loss of either one grid interface or one Energy Link. Simultaneous loss of two EnergyLinks isolates part of the network, containing the disturbance while service within the island is determined by localgeneration, stored energy, load priority, and repair duration. Reliability screening under assumed component failureand repair rates identifies network partitioning as a rare but credible event. Monte Carlo analysis of the networkpartitioning failure scenario shows full critical service in all sampled summer conditions and a median of approximately92% continuous critical service under sampled winter conditions.
(PDF) Quantifying EnergyNet performance: a simulation-based framework for decentralized energy networks. Available from: https://www.researchgate.net/publication/415031739_Quantifying_EnergyNet_performance_a_simulation-based_framework_for_decentralized_energy_networks [accessed Oct 03 2026].