Benchmarking Co-Simulation Orchestration Engines for Integrated Energy Systems: A Comparative Study of Mosaik and HELICS

Integrated Energy Systems (IES) are inherently complex, requiring co-simulation tools capable of integrating heterogeneous models and capturing diverse dynamics. This paper presents a comparative benchmark of two leading Co-Simulation Orchestration Engines (COEs) for IES: Mosaik and HELICS. We evaluate the performance and scalability of these engines within a containerized, distributed co-simulation platform based on Kubernetes. Following a qualitative comparison of their programming and execution models, we benchmark the execution time of each engine using a realistic district-scale IES scenario, varying the complexity of the simulation and the underlying computational resources. Our results demonstrate that HELICS exhibits superior scalability compared to Mosaik, achieving a more efficient use of resources as the complexity of the simulation increases. We analyze the impact of key architectural features to explain the observed performance differences. We also evaluate the benefits of a multi-broker topology for HELICS, assessing its resilience to overhead and its implications for future geographically distributed co-simulation workflows. Our study may guide the selection and optimization of COEs for the next generation of IES co-simulation tools.