Mechanical Investigation of Various Stack Designs for Solid Oxide Cells

To better understand degradation in electrochemical converters and helping to correlate certain phenomena with specific operating conditions, machine learning (ML) methods are increasingly being applied. Success has already been achieved in the field of degradation analysis and prediction of capacity of lithium ion batteries1, for instance. In terms of Solid Oxide Cell (SOC) stacks ML methods have been applied mainly with the aim of identification of faulty operation modes and degradation related fault diagnosis2. ML approaches usually require a considerable amount of real training data, when used for forecasting models. A data consolidation and curation strategy was developed with the aim of processing the historic long-term test bench data of SOCs collected by Forschungszentrum Jülich over the past years. In comparison to other datasets developed in this field3, the one presented in this work contains SOC stack tests in fuel cell operation with significantly longer operating times under load. A compilation of the sample experiments and the consolidation into a hierarchical data format are presented. Further, an essential part of the strategy is the automatic curation and analysis of electrochemical impedance spectroscopy (EIS) measurements, using a specifically developed procedure in Python. The varying quality of measurements from past years, as well as recurring artefacts such as parasitic inductances, can be addressed in this way. Additional distribution of relaxation times (DRT) deconvolutions and equivalent circuit modelling (ECM) are performed, as part of the procedure to automatically retrieve feature values from measurements (cf. Fig. 1). The novel dataset, which to the authors’ knowledge includes some of the longest SOC stack tests available, serves as the basis for several evaluations. In addition to classification and clustering work to derive degradation patterns, in particular based on the EIS data, another focus is on the development of forecasting models. The current work is primarily concerned with long short-term memory (LSTM), as well as regression models that make use of both the time series data and the characterisation measurements, such as EIS. Acknowledgement The authors would like to thank their colleagues at Forschungszentrum Jülich GmbH for their great support and the Helmholtz Society as well as the German Federal Ministry of Education and Research for financing these activities as part of the WirLebenSOFC project (03SF0622B). References 1: Jones, P.K., Stimming, U. & Lee, A.A. Impedance-based forecasting of lithium-ion battery performance amid uneven usage. Nature Communications 13, 4806 (2022).2: B. Yang et al. Solid oxide fuel cell systems fault diagnosis: Critical summarization, classification, and perspectives. Journal of Energy Storage 34, 102153 (2021).3: A.K. Padinjarethil, S. Pollok & A. Hagen. Degradation studies using machine learning on novel solid oxide cell database. Fuel Cells 21, 566–576 (2021). Figure caption: Fig.1: Flow diagram of EIS data curation pipeline and curation results for example EIS measurement. Figure 1

Abstract

Solid oxide cells can deliver highly efficient energy conversions between electricity and fuels/chemicals. A central challenge of upscaling solid oxide cells is the probability of failure of the brittle ceramic components. The failures of the ceramic components may lead to significant degradation or eventual failure of a stack. To predict mechanical failures in a stack, a full stack model is needed, together with a local assessment of stresses at the vicinity of failing regions, e.g. the contact points between the cells and interconnects. A conventional three-dimensional model requires a very fine discretization of the mesh to capture stress intensities. Computational resources needed for such a model are therefore immense, and it is highly unlikely to compute at stack scale, as well describe the evolution over time. In this work, the homogenization modelling framework for solid oxide cell stacks is extended to identify local mechanical failures. Thus, the fracturing within a local failing point is examined by using a localization approach, where stresses in the stack model are linked to the local stresses and the energy release rate at the crack tip of the relevant interface. This is done in a general manner, such that the local stresses and the energy release rate can be evaluated at every point in the stack at every instant of time without loss of computational efficiency. A 100-cell stack can be modelled in three dimensions with all coupled multiphysics in steady state within 3 min on a current workstation computer.

Increasing demand for long-term energy storage, reduced carbon emissions, more resilient grid, and the lowest cost grid balancing solutions promotes solid oxide electrolysis cell (SOEC) and reversible solid oxide cell (RSOC) technologies. These systems can achieve high hydrogen production rates with near 100 percent stack electrical efficiency while co-electrolysis of carbon dioxide and water leads to production of syngas. RSOC can serve as a grid storage device (hydrogen production) and power production plant (provide electricity during peak demand). With over 25 years’ experience building and testing solid oxide cell materials, components and stacks, Nexceris has a proven track-record of commercializing products to major OEMs. The established manufacturing platforms at Nexceris enable smooth expansion into new applications.In this talk, development of cells, stack components and stacks will be discussed. Performance and durability improvements from single cell to short stacks will be addressed. Cells with downselected electrode sets were implemented into short stacks with active cell area of 42 cm2 and 96 cm2. Stacks were operated continuously in SOEC mode at ~1.3 V for ≥ 500 hours with <5%/kh degradation and improving. The resulting short stack demonstrations show promising baseline performance and durability platform for future cell and stack development. RSOC tests were recently initiated with dynamic operation. Nexceris’ ChromLok coatings for bipolar plates were also integrated to stacks to enhance durability.

Numerical simulation acceleration of flat-chip solid oxide cell stacks by data-driven surrogate cell submodels

Fast relaxation of stresses in solid oxide cells through reduction. Part I: Macro-stresses in the cell layers

This study examines the performance of a solid oxide cell (SOC) stack during co-electrolysis of CO2 and H2O at elevated pressures up to 8 bar. Steady-state and dynamically recorded U(i)-curves were performed in order to evaluate the performance over a wide temperature range and to quantify the area specific resistance (ASR) at different pressure levels. Furthermore, the outlet gas composition at various current densities was analyzed and compared with the thermodynamic equilibrium. The open circuit voltage (OCV) was found to increase with higher pressure due to well known thermodynamic relations. An increase of the limiting current density at elevated pressure was not observed for the investigated stack with electrolyte supported cells. The ASR of the stack was found to decrease slightly with higher pressure. It revealed an increase of the cell resistance with lower H/C ratios in the feed at lower temperatures, whereas the performance of the co-electrolysis was very similar to steam electrolysis for temperatures above 820 °C. Within an impedance study for steam, co- and CO2 electrolysis operation it was shown that pure CO2 electrolysis exhibits a higher pressure sensitivity compared to pure steam or co-electrolysis due to significantly increased activation and diffusion resistances.

A new concept was developed to repair leaking solid oxide cell (SOC) stacks. Conventionally, stacks are joined with a glass-ceramic sealant before they are put into operation. Once the sealant starts leaking, they are considered unusable. The approach outlined here aims to prevent the scrapping of valuable aggregates. The method was tested for a composite sealing material based on a crystallizing glass matrix of the BaO-CaO-SiO2 system reinforced with yttrium-stabilized zirconia fibers.

Electrodeposited Co-free spinel ZnFeMnO 4 coating on ferritic stainless steel (SUS) was evaluated as an interconnect/separator material for solid oxide cell (SOC) and benchmarked against Co 2 MnO 4 . Variable-temperature X-ray diffraction (RT–800 °C) confirmed a phase-pure spinel and a tetragonal to cubic transition, with no impurity phases observed. Although ZnFeMnO 4 pellets exhibited an order-of-magnitude lower total conductivity than Co 2 MnO 4 between 600 and 900 °C, coated assemblies showed area-specific resistance (ASR) above 750 °C comparable to those of Co 2 MnO 4 , with higher ASR only below 700 °C. Isothermal oxidation in air at 900 °C followed stable parabolic kinetics without anomalous oxidation to 1500 h; mass gains of ZnFeMnO 4 - or Co 2 MnO 4 -coated SUS were less than 40% of that of uncoated SUS. Energy-dispersive X-ray mapping revealed no detectable Cr in the La 0.6 Sr 0.4 Co 0.2 Fe 0.8 O 3− δ (LSCF) layer after 100–1200 h at 900 °C, attributable to a dense interfacial ZnFeMnO 4 sublayer that blocks outward chromium diffusion and inward oxygen ingress. At 900 °C for 2500 h, the ASR remained essentially constant with negligible increase. Under these conditions, the coating layer densified and the porous overlayer diminished, while Cr diffusion was suppressed and no delamination occurred. These results identify electrodeposited ZnFeMnO 4 as a durable Co-free coating that satisfies oxidation resistance, Cr-leakage suppression, and ASR stability requirements in SOC stacks.

Recently, research on hydrogen technologies has been raised due to ongoing strategy to shift from fossil fuels to renewable energy. Based on this future development, electrolyzers as water to hydrogen converters will play a crucial role in balancing the time and place shift between electricity generation (e.g. by wind and solar power) and necessary electricity demand. The four types of electrolyzers – alkaline (AE), polymer-membrane (PME), anion exchange (AIE) and solid oxide (SOE) – are the backbone of this equalizing possibility. The technology readiness level of those four technologies is different, being for AE state-of-the-art, for PMEs around 8, for SOEs around 6 and for AIEs below 6. Solid oxide electrolyzers have their highest benefit in combination with other high-temperature industrial processes like aluminum, steel, copper, cement, ceramic and glass production. These processes can deliver off-heat or -steam and thus raise the SOEs electrical efficiency to above 85%. In 2030, the European Union aims for having an installed electrolysis capacity of around 120 GW [1], thereof probably 5-10% SOE [2]. Such large annual manufacturing capacities lead, after a certain operation regime, to end-of-life (EoL) stacks.Thus, since recently R&D on recycling possibilities of electrolyzer stacks has been starting with first ideas, strategies and experiments [3–8]. This research covers the whole chain from stack dismantling strategies, separation facilities of stack components, sorting, re-preparation and finally re-usage (re-use, re-furbishment, re-cycling). Based on that, also design strategies for recycling or dismantling friendly stacks evolve.The paper presents an overview of the actual R&D on SOFC & SOEC stack recycling possibilities with respect to the metal and ceramic components. An emphasis is put on fuel-electrode supported cell stacks, but also metal and electrolyte supported stacks are touched. Some fundamental research aspects of metal recovery and ceramic recycling capabilities are presented in more detail.[1] Dolci F. et al. DOI: 10.2760/7606.[2] Flis G. CATF – Solid Oxide Electrolysis: A Technology Status Assessment. Clean Air Task Force; 2023.[3] Mori M et al. DOI: 10.1016/j.ijhydene.2024.08.411.[4] Sarner S. et al. DOI: 10.1002/aenm.202201805.[5] Lastam J. et al. DOI: 10.3390/met14040406.[6] Al Assadi A. et al. DOI: 10.1016/j.rcradv.2023.200172.[7] Valente A. et al. HyTechCycling; 2018.[8] Yenesew G.T. et al. DOI: 10.1016/j.resconrec.2022.106809.

Solid oxide cell stacks (SOCs) usually use a Cr-containing alloy as the interconnect, which requires the air electrode not only to exhibit satisfying electrochemical performance, but also a high resistance to Cr poisoning. This is because during operation volatile Cr species will diffuse from the oxide scales on the interconnect to the air electrode, leading to a fast degradation. Currently, Forschungszentrum Jülich (FZJ) applies an MCF (MnCo1.9 Fe0.1O3–δ) coating by means of atmospheric plasma-spraying (APS) on the interconnect. This coating reduces the degradation rate from 0.9 % / kh to ~ 0.3 % / kh. However, according to distribution of relaxation time (DRT) figures, Cr poisoning is still a major contribution to overall degradation in the beginning of the operation.In order to further alleviate issues of Cr poisoning, an LSC (La0.6Sr0.4CoO3 –δ) air electrode was investigated. Based on previous studies, the LSC electrode was expected to be more resistant to Cr poisoning compared to LSCF (La0.58Sr0.4Co0.2Fe0.8O3 –δ). Long-term tests demonstrated a lower degradation rate for LSC air electrodes compared to LSCF air electrodes and LSC adsorbed less Cr than LSCF even if the operation time of the SOC stack with LSC air electrode was twice longer [1]. For improving the understanding of the degradation mechanisms of Cr poisoning on LSC and LSCF, an accelerated test was carried out. The test was done with a rainbow stack that used LSC (RU A) and LSCF (RU B) air electrodes in different repetitive units (RU) of the stack. To quickly degrade the air electrodes, no protective coatings were added on the interconnects and 1.4509 stainless steel was used instead of Crofer 22 APU. Besides, the operation condition (775 ºC, 0.75 A/cm2) was harsher than that of previous tests (730 ºC, 0.5 A/cm2). Due to fast degradation, the stack was shut down after only 1,700 h operation. Latest analysis of this stack showed the peak attributed to air electrode polarization in the DRT deconvolution for EIS spectra of RU A cells increased evidently while that of RU B cells barely changed. This was a contradictory finding to previous understanding as it indicated it was LSCF instead of LSC that was more resistant to Cr poisoning. A possible reason might be a lower chemical stability of LSC to volatile Cr species at high temperature and current density. To clearly reveal the reason, Cr distribution and deposited Cr amount in the air electrode needs to be determined. Post-test characterization, such as SEM, EDX and ICP-OES, is undergoing.In addition, a steady-state multiphysical three-dimensional model of an SOC was developed with the open-source software OpenFOAM. Momentum transfer, mass transfer and electrochemical reactions were successfully included in the model. A coupled solver was applied to solve the ionic and electronic potential across different regions. The model was verified by the comparison of the simulation results of the OpenFOAM model with the simulation results of an equal model developed with the commercial software COMSOL, whereas the same mesh, geometry, and parameters were used. Also, the OpenFOAM model was validated by a comparison of the simulation results with experimental results under different operation modes (e.g., fuel cell and electrolysis modes) and temperatures. Assisted by the model, the spatial distribution of the local mole fraction of oxygen and local overpotential in the air electrode could be revealed, which was helpful in analyzing Cr poisoning. As expected, it was found that the overpotential mainly became non-zero at the interface between the air electrode and the electrolyte. Since Cr is mainly observed near the interconnect interface of the LSCF air electrode in post-test examinations, the Cr deposition is expected to primarily arise from chemical reactions. Plus, owing to higher current density on the rib, the mole fraction of oxygen was also lower under rib, where more Cr deposition can be triggered owing to lower oxygen partial pressure. Micro-kinetic modelling of the reduction of adsorption sites owing to Cr poisoning is under development.After the governing equations related to degradation will be included, the model will become a transient-state model which is capable of predicting the voltage with evolution of time.

Fatigue resistance of steels: Boardman, B. Metals Handbook 10th Edn (ASM International, Materials Park, Ohio, 1990) Vol. 1, pp 673–688

Mathematical approaches to modelling the mass transfer process in solid oxide fuel cell anode

The aim of this presentation is to address three fundamental questions related to solid oxide cells: Are the oxygen reduction reaction (ORR) and oxygen evolution reaction (OER) indeed the rate-limiting steps in the high-performance of solid oxide cells, and how can we accurately quantify the area specific resistance of the oxygen electrode?What causes the phenomenon known as “electrochemically driven phase transformation,” and how can our understanding of this process be applied to create electrodes that are both high-performing and durable?How can we develop reliable accelerated testing protocols for both solid oxide fuel cells and solid oxide electrolysis cells, and which physical processes are being accelerated through these tests? The accurate quantification of electrode polarization in fuel cells and electrolyzers holds critical importance but presents a significant challenge due to the extensive overlap in polarizations from both the anode and cathode, whether in DC or AC measurements. This overlap creates a notable gap in our understanding of electrode behavior and in the development of electrochemical devices that are both high-performing and durable. In this presentation, I will introduce a reliable method for measuring the polarization of individual electrodes within a solid oxide cell. This technique is versatile and can be easily adapted for use in various types of electrochemical cells. Additionally, we will explore the limitations associated with using the distribution of relaxation times analysis for determining the polarization of individual electrodes.We will then address the origin for the phase transformation observed in an oxygen electrode during operation, as opposed to what is seen with thermally annealed materials. By integrating electronic conduction at the interface layer, we can significantly reduce the occurrence of phase transformation. This understanding of “electrochemically driven phase change” is crucial for the development of reliable accelerated testing protocols. Such protocols are essential in solid oxide fuel cell research to speed up the investigation of durability issues, quickly identify potential failure mechanisms, and ultimately estimate the lifespan of electrochemical cells. In our study, we compared solid oxide fuel cells operated at constant current density with those subjected to accelerated testing methods, which involve intermittent current injections. We have formulated a general accelerated testing framework, which will be presented in detail.Finally, we will present a theoretical model that underlies those three questions and can be used to address stability and durability issues in solid oxide cells. Acknowledgements: The presenter would like to thank Prof. Anil Virkar, Dr. Emir Dogdibegovic, and Dr. Yudong Wang, for their invaluable contributions to the work presented. This work was made possible through the support of projects DE-FE0026097, DE-FE0031667, DE-AR0001346, and DE-FE0032110.

The purpose of the present study was to discuss the effects of local solidification time and thermal gradient on porosity content in long solidification range A201 aluminum alloy plate castings. With systematic change in the riser size, together with variation of thickness and length, were cast in different kinds of molds. The sand molds with end chill for the plate castings were made of 100% silica sand. The porosity content of A201 aluminum alloy was affected by the local solidification time and thermal gradient at same time in this study. The correlation between porosity content and any individual thermal parameter is found to be split into 2 bands, each being associated with one of the two thicknesses of plate castings. The higher the thermal gradient, the lower the porosity content was measured in the A201 aluminum alloy plate casting. And high thermal gradient with short solidification time will get lower porosity content in aluminum alloy castings