Multiscale modeling of solid oxide electrolysis: stack design, process simulation and economic assessment

• Autores: Gonzalo Jiménez Martín
• Directores de la Tesis: Mónica Aguado Alonso (dir. tes.), Iñigo Garbayo Senosiain (codir. tes.), Xabier Judez Lopez (codir. tes.)
• Lectura: En la Universidad Pública de Navarra ( España ) en 2026
• Idioma: inglés
• Número de páginas: 217
• Tribunal Calificador de la Tesis: Julie Mougin (presid.), Luis María Gandía Pascual (secret.), Marek Skrzypkiewicz (voc.)
• Programa de doctorado: Programa de Doctorado en Tecnologías de las Comunicaciones, Bioingeniería y de las Energías Renovables por la Universidad Pública de Navarra
• Enlaces
  • Tesis en acceso abierto en: Academica-e
• Resumen

  • Green hydrogen is a promising solution for decarbonizing the economy, particularly in the power generation, industrial and transportation sectors. It is produced through water electrolysis powered by renewable energies. Among the various electrolysis technologies available, solid oxide electrolysis (SOE) stands out due to its inherent advantages. First, operating at high temperatures reduces electric energy consumption, secures high efficiency and eliminates the need to use noble metals as reaction catalysts. Second, this range of temperatures favors the possible direct conversion of steam and CO2 to produce syngas. Notably, there is a need for a heat supply to reach high temperatures and to generate steam. Nonetheless, SOE can be thermally integrated with various industrial processes or exothermic chemical syntheses to reach unrivalled efficiencies. Despite significant technological advances over the past decade, SOE still faces challenges that have hindered its full deployment. These include degradation problems caused by high-temperature operation and reduced system efficiency linked to reliance on an external heat source. These issues result in high capital and operational costs, raising questions about the economic feasibility of SOE systems. To address these challenges, this thesis aims to advance SOE technology by using modeling and simulation tools at various levels. Specifically, the objectives of this thesis are i) to investigate the impact of interconnector plate (IP) geometry on flow, temperature, and current distribution in SOE stacks to enhance overall performance and minimize degradation, ii) to improve system-level energy efficiency via thermal integration strategies, and iii) to assess the technoeconomic feasibility of SOE systems for hydrogen production. Starting with a SoA IP design, CFD simulations were utilized to evaluate and suggest potential design improvements. The analysis focused on optimizing flow distribution in both the horizontal and vertical directions of the stack as well as on homogenizing temperature and current distribution in the horizontal plane. Based on the results from simulations, it could be concluded that, i) maximizing the feeding and exhausting area in the manifold is crucial for optimal flow distribution, being the two inlet/three outlet holes and its opposite the best configurations among the options evaluated; ii) employing discrete cylindrical ribs as patterning improves flow and temperature distribution and reduces pressure drop. However, the lower contact area in comparison with parallel ribs can compromise the performance. Therefore, metal meshes or foams are essential to guarantee uniform current collection in the IPs. Based on these findings and relevant technical considerations, an optimized shortstack design was subsequently proposed. To model and simulate a MW-scale SOE system, the commercial process simulator Aspen Hysys was employed. The SOE modules behavior was represented by using a custom model developed in Aspen Custom Modeler, while the remaining system components were modeled using standard unit operations available in the Aspen Hysys library. Once the model of the entire system was created and simulated, the mass and energy balances were obtained. The results revealed significant utility demands for the baseline system, indicating that system efficiency could be improved through heat integration. Consequently, a pinch analysis was performed, revealing that for this plant size the heat demand could be reduced up to 59 %, and the cooling demand up to 70 %. To approach these ideal targets obtained from the pinch analysis, a new optimized heat exchanger network (HEN) was proposed, resulting in actual heating and cooling demands 57 % and 67 % lower than the baseline. The new HEN was implemented into the Aspen Hysys flowsheet to simulate the system under different loads. The most energy-efficient operating point was found at 69 % of the maximum system load, resulting in a total specific energy consumption of around 46.2 kWh·kg-1 H2. It was composed of 37.6 kWhe·kg-1 of electrical consumption and approximately 8.6 kWhh·kg-1 of thermal consumption, which is in good agreement with the KPIs set by the Strategic research and innovation agenda. The data from simulating the SOE system was employed to analyze the feasibility of this kind of plants into the Spanish context. Specifically, the case of study consisted in the SOE system coupled with a PV-BESS off-grid plant for hydrogen production. This techno-economic assessment was conducted for different industrial locations in Spain: Puertollano, Tarragona, Algeciras and A Coruña. Based on the SOE system data and the irradiation data at each location, the levelized cost of hydrogen (LCOH) was estimated. It was found that electricity consumption is the primary contributor to OPEX at all locations, with higher costs occurring in regions with lower solar resource availability such as A Coruña. Furthermore, CAPEX and replacement costs together represent approximately 38 % of the LCOH, highlighting the need for technological advancements to reduce stack cost and extend stack lifetimes to reduce overall costs. The trade-off between maximizing the capacity factor and minimizing the LCOH was investigated for different PV and BESS hybrid plant sizes, and operating windows of the SOE system. This analysis showed that there is an optimal PV-BESS plant size that minimizes the LCOH. Additionally, the capacity factor declines as the operating window is reduced. Apart from that, narrowing the operating window leads to a further reduction in LCOH, as the system operates closer to its most efficient point, thereby lowering OPEX. Since stack cost and lifetime are uncertain parameters, its impact on the LCOH was evaluated through a sensitivity analysis. The SOE stack cost has a linear impact on the LCOH, with a variability of approximately 26 % when ranging between 500 and 1,500 €·kW-1. The SOE stack lifetime has a high impact on the LCOH for short lifetimes, while this impact is reduced when the lifetime is extended above 60 kh. The sensitivity analysis helped to identify value combinations needed to meet specific cost targets. The best results were obtained at Puertollano, with values of LCOH from 4.0 to 4.4 €·kg-1 for stacks able to operate more than 56 kh and at a price of 500 to 610 €·kW-1. In conclusion, this work presents a comprehensive approach to advancing SOE technology through multiscale modeling, from component-level design to system integration and economic analysis. The results demonstrate that significant efficiency gains can be achieved through optimized stack and process design, and that cost-competitive hydrogen production is feasible in high-irradiation regions, if stack cost and durability reach targeted thresholds. These findings contribute valuable insights into the deployment of SOE systems in real industrial contexts.