Integration and Modeling of Electrolyzers in Medium-Voltage Grids: Methodological Development and Impact Assessment

Summarise with VIRAT
Authors: Saeed Sohbatloo
Cited by: 0

Summary

No summary yet. VIRAT can read this paper and write a plain-language summary of its aim, methods, key findings and limitations.

Summarise with VIRAT Free to try with your Science Catalogs account.

Abstract

The rapid expansion of variable renewable energy, together with the growing need to decarbonize industrial energy use, has increased the demand for flexible and grid-aware power-to-hydrogen operation. Within this context, electrolyzers should be regarded not only as hydrogen production units but also as controllable electrical loads whose operating behavior directly affects medium-voltage (MV) grid performance, renewable-energy utilization, and economic outcomes. This thesis develops a reproducible methodology for the integration and operation of a grid-connected electrolyzer in an MV distribution network under realistic technical, operational, and market constraints. The proposed methodology integrates an electrolyzer and hydrogen-storage model, an MV grid interaction model, and a predictive control framework within a unified workflow. The aim is to coordinate hydrogen production with local grid limits, renewable availability, and time-varying electricity prices. To ensure practical relevance, the approach is implemented on a real sub-network of the Freiburg 20 kV MV grid and evaluated in a quasi-static time-series (QSTS) simulation environment using the open-source software pandapower. The control concept is formulated as a two-layer architecture consisting of an optimization-based decision core for active-power scheduling and a supervisory logic layer for plant-executable operation. The supervisory layer explicitly represents electrolyzer operating states (OFF, STANDBY, ON), start-up restrictions, and hydrogen-storage constraints. This architecture enables realistic rolling-horizon operation while avoiding the computational burden associated with fully integrated hybrid optimization. The framework is assessed through benchmark comparisons and structured scenario analyses. Relative to reference methods based on dynamic programming and MINLP, the proposed controller achieves substantially lower computation times while maintaining strong economic performance, which makes it suitable for repeated QSTS execution. The scenario analysis further shows that longer prediction horizons improve profitability primarily through better operational timing and more effective use of hydrogen-storage headroom, whereas grid-aware weighting improves voltage and loading performance at a measurable economic cost. In addition, reactive-power capability at the point of common coupling is evaluated, and the results show that smart inverter operation with Q-droop control improves local voltage support with only a minor economic penalty. Overall, the thesis demonstrates that electrolyzer integration in MV grids should be treated as a coordinated techno-economic control problem rather than solely as a grid-connection or market-dispatch problem. The pandapower-based methodology provides a transparent and transferable framework for future planning and operation, particularly for the coordinated deployment and control of multiple electrolyzers in increasingly electrified and hydrogen-coupled distribution systems.

Fields of Study & Subjects

Hybrid Renewable Energy Systems Integrated Energy Systems Optimization Smart Grid Energy Management

Publication details

Journal / source
Opus-HSO (Offenburg University of Applied Sciences)
Publication date
1 January 2026
Language
EN
Document type
Thesis