Modelling of regional hydrogen supply chains -spatially resolved optimisation from electricity generation to hydrogen application under consideration of import and export possibilities

Hydrogen will be a central and indispensable element of climate-neutral energy systems and an important component of sector coupling. However, the development of the hydrogen economy is making slow progress due to a lacking market as well as political, scientific, and economic uncertainties regarding where, when and to what extent it will be used. One promising solution are local ecosystems in which the entire hydrogen value chain is coordinated in close proximity and through cooperation between relevant stakeholders. In an international context, this approach is known as ‘hydrogen valleys’. Hydrogen valleys enable cost reductions through coordinated developments along the value chain, shared infrastructure, economies of scale, and a local development of necessary commitment between producers and purchasers. They can act as nuclei for an (inter-)national hydrogen economy. The planning of hydrogen valleys requires sophisticated tools for detailed calculations of possible hydrogen supply chains and necessary infrastructure for realistic costing under developing external conditions. Here, model-based techno-economic analyses and optimisations can make a valuable contribution. As many aspects of the hydrogen value chain have a clear spatial reference and regional differences, a model with spatial resolution is essential for a realistic depiction of the entire hydrogen system. An important aspect of the model is the consideration of the import and export of electricity and hydrogen, as self-sufficient operation of the regional hydrogen system is only realistic for a transitional period at most. Such a model with the necessary technical, spatial, and temporal level of detail is not yet available. For this reason, a spatially resolved, nonlinear, dynamic modelling environment for the simulation and optimisation of regional hydrogen supply chains is being developed as part of this Ph.D. work. The development is carried out by extending the at Fraunhofer ISE existing modelling environment H2ProSim with regard to a spatial dimension. This requires the implementation of new technologies and control algorithms for material and energy exchange, the processing of spatially resolved data, new methods for evaluation and presentation of the results, especially in the form of maps, as well as an improvement of the optimisation algorithm due to the greatly increased complexity. The development and subsequent application of the modelling environment is motivated by the following key questions:
On what basis can an adequate spatial resolution be derived and solved in the model? Which spatially distributed conditions should it be based on? And how do spatial resolutions and depths of detail differ in the model results?
How should a spatially resolved optimisation model for hydrogen regions be designed? Which technologies in the value chain are central and how can the material and energy flows be controlled?
What results can be derived from the optimisations of a selected case study? How can the hydrogen economy develop in the region and how transferable are the results to other regions?
Answers to the key questions are presented in the three scientific publications of this cumulative dissertation. The publications each focus on a group of key questions (spatial resolution, modelling methodology, and model application in different scenarios). A summary of the most important results from the literature research, modelling methodology and application is presented in this thesis, as well as a critical discussion of the entire project. The development of a hydrogen economy in the Southern Upper Rhine region under various boundary conditions represents the case study for the application of the model environment in this thesis. The results demonstrate the functionality and relevance of the developed modelling environment and illustrate the strong influence of the different model designs and boundary conditions. Hierarchical clustering algorithms show the best suitability for forming the spatial resolution of the optimisation model. A minimum spatial depth of detail, 40 subregions in the case study Southern Upper Rhine, should be used to realistically map and evaluate transport processes and plant locations. Reliable optimisation results can be achieved for this level of detail. The model results also recommend the synergetic combination of bilateral contracts and flexible hydrogen exchange between the locations, as well as the consideration of pipeline installations for hydrogen transport up to 30 km and with an annual transport volume of more than 1000 tonnes. A methodology developed to determine the spatially resolved costs of hydrogen production and supply enables the evaluation and identification of particularly suitable locations. Under the assumed prices on the electricity and hydrogen market, which were developed based on an energy system analysis for the defossilisation of Europe, the import of hydrogen via the planned national hydrogen network represents the most cost-effective supply option for the case study Southern Upper Rhine, enabling hydrogen supply cost of approximately 5-6 €/kg in 2030. Local production is not competitive to hydrogen import, which means that currently planned plants for hydrogen production in the region could no longer be economically viable once the national hydrogen grid is established. However, the result is heavily dependent on hydrogen prices and the cost of electricity from hydropower, which is available in large quantities in the region, as well as on the weather data sets used, as sensitivity analyses show. The model developed as part of this work enables a variety of analyses of hydrogen value chains under changing boundary conditions. In order to improve the quality of results and feasibility and to extend the scope of application, extensions to the modelling of electrical systems, synthetic fuel production technologies and business models are recommended. The most significant contributions of the present work are expected from the application of the modelling environment in local projects for hydrogen valleys. In such case studies, a very high quality of results is achieved only through locally collected input data and direct dialogue with stakeholders. In return, the regions benefit from precise site assessment, infrastructure design and cost optimisation.