Degradation monitoring and performance optimisation of SOECs

Solid oxide electrolysis cells (SOEC) are electrochemical devices that convert excessive electrical energy directly into fuel components such as hydrogen, syngas, and methane, in a highly efficient and environmentally friendly manner. The non-polluting fuels can be further used for the production of heat and electricity as well as for automotive applications. High efficiency and, remarkably, no need for rare materials like platinum or lithium like for applications in low-temperature electrolysers or batteries, are just some of the attractive properties that make solid oxide electrolysis highly promising. This comes at a price of high operating temperatures of ~550-900C, long start-ups and a range of harmful chemical processes that degrade cell performance and affect its durability. Insufficiently long- term performance and yet poor durability are the major hurdle for broad SOEC commercialization. Apart from the quest for new materials that will help reduce the operating temperatures, two fundamental problems remain open. The first is insufficient understanding of the onset and prediction of the degradation phenomena within the cells. The second is the lack of reliable means to detect, recognise and allocate the evolving degradation modes online (modus operandi) and hence take appropriate counteractions to prolong the operational life of these devices. Exemplary, during a medical operation, it is necessary to know the current health condition of the patient, what kind of health deterioration has occurred, what counteractive steps can be taken, and how long the patient still has to live. The same questions and knowledge are important for SOEC technology. These questions form the basis of the emerging discipline of prognostics and health management (PHM) based on online monitoring, yet almost no research applying methods from that field has been done in the area of SOECs so far. This is what the current project will focus on. Next, quite a difficult problem to be addressed, concerns the design of efficient counter-measures to avoid or slow down the degradation. There are no results available so far in the SOEC domain. Thus, this research project will go beyond the current state-of-the-art in identification methods for both SOEC degradation mechanisms, and novel, to date unavailable, methods of monitoring SOEC systems online. These principles will allow gaining fundamental knowledge of different degradation mechanisms, which can induce the irreversible deterioration of the cells microstructure and performance. In order to gain the required knowledge an interdisciplinary team of highly motivated researchers from TUG and JSI will work together: The experience that the JSI team gained in the design of mitigation action in solid oxide fuel cells will be utilised. The entire methodology will be assessed experimentally at TUG, based on its extensive experience and recent original results in regeneration of solid oxide devices. An important outcome of the project is that long-term measurement data as well as accelerated degradation test results will be publicly released, which will be the first publically available data from SOEC durability tests so far. Thus, the current state-of-the- art in SOEC diagnostics will be moved a step forward by changing the one failure detection approach to the detection of all simultaneously occurring failures approach, their prediction and prevention.

The main objective of this project was to identify and understand various degradation processes that occur during the operation of high temperature electrolysers (SOEC) for hydrogen production. It also aimed to develop techniques to optimize performance, reduce aging and thus extend the lifetime of the technology. In order to achieve these goals, numerous short- and long-term experiments were conducted at laboratory scale using various methods for performance monitoring. Thereby, different types and degrees of performance deterioration and microstructural changes were observed and analyzed in detail using the extensive data set of the different measurements performed. In the course of this, strategies for accelerating various degradation mechanisms have been developed as an alternative to conventional performance assessment by long-term testing. These methods enable fast evaluation of SOEC performance and prediction of the lifetime of the technology. The corresponding experimental results indicate that only a few hundred hours of operation may be sufficient to predict long-term performance if appropriate operating conditions are selected. In addition, strategies were developed to mitigate degradation and regenerate SOEC performance during co-electrolysis of H2O and CO2. The corresponding experimental results demonstrate the potential of the applied operating strategies to mitigate degradation and regenerate SOEC performance in a manageable manner with particular attention to short application times. Furthermore, novel performance monitoring tools to identify early-stage degradation mechanisms were tested and validated. In addition, the great potential of artificial neural networks as a time-efficient and affordable tool to predict SOEC performance with high accuracy under real-world conditions was demonstrated. In summary, the results of this project have enabled us not only to understand why and how various relevant degradation processes occur and develop over time, but also how we can mitigate or prevent them and thus sustainably extend the lifetime of the technology.