Search

Solid Oxide Cell (SOC) technology is considered as an efficient electrolysis technology to produce hydrogen at large scale. It can also operate in fuel cell mode using different fuels (carbon-based or non-carbon based like ammonia), and in reversible mode. Though proofs-of concept have been achieved at different relevant scales for those operating modes, some R&D works still need to be performed to improve performance, durability and cost. Improved and upscaled cells and stacks need to be developed, with a methodology combining multiscale and multiphysics modelling, electrochemical characterization in relevant conditions and advanced post-test analysis. Their integration into modules made of several stacks is also a stepping stone in order to reach multi-MW electrolysers. CEA is working on the whole value chain of SOC technology, from cell development and optimisation to module design and operation through stack upscaling. Recent achievements on those aspects will be presented.

Solid oxide cell technology is currently experiencing a rapid industrialization phase. To investigate operational strategy, CEA/LITEN has designed and constructed its first multi-stack reversible solid oxide cell (rSOC) module. While it is able to host four of CEA’s standard 25-cell stacks, the present work reports on preliminary validation results obtained in a 2-stack configuration. Thermal losses have been quantified and identified. While the hotbox is showing high performances, the overall losses increased twofold when taking into account pass-through piping and current connections. Module fluid distribution was verified to be homogeneous, and does not affect nominal stack operation. A durability test of more than 2 kh is presented. Over the first 1.1 kh, the stacks behavior was compared to that of a stack previously operated on a test bench. The remarkable similarities indicate adequate control of the module. Finally, a detailed analysis of the recorded efficiency was conducted.

This work aims at developing an innovative solution for storing renewable energies, so-called “Smart Energy Hub” (SEH) developed by Sylfen, based on reversible Solid Oxide Cell (rSOC) technology. It is able to operate either in electrolysis (SOEC) mode storing excess electricity to produce hydrogen, or when energy needs exceed local production, in SOFC (solid oxide fuel cell) mode producing electricity (and heat) from hydrogen or any other hydrogenated fuel. Particularly this paper focuses on stack development and optimization, partly performed through the REFLEX project supported by European commission funding. To achieve the project’s objectives, in terms of efficiency and increase of system power density, as well as multi-stack operation, optimizing CEA stack is necessary to upgrade a stack originally designed for SOEC research purposes, to a stack reversibly operated, but also integrated as a part of module. This paper presents results of stack optimisations and integration of cells developed within the REFLEX project. First, a reduction in cell thickness of ~20 % led to adapting the stack manufacturing process parameters. Some tests, here presented, demonstrate that the stack performances were not affected by this change. Then, significant effort was devoted to air distribution at the Single Repeat Unit (SRU) scale to minimize pressure drop, undesirable for both global system efficiency and mechanical resistance of stack sealing. At the SRU scale, air path adjustment reduced the pressure drop by a factor 2; at stack scale, the performances obtained in i-V curves demonstrated reactants distributions were unchanged. Optimized cells (so called G2) developed in the frame of REFLEX project were then integrated in a short stack (5 cells vs. 25 full size) to be compared to short stack comprising regular cells produced by Elcogen. Both stacks were tested in terms of durability over more than 1000 hours at 700°C. A first period of about 800 h of operation was conducted alternating from SOFC to SOEC mode by steps of ≈100 h. During this first part, the stack was supplied by 50/50 H2O/H2 at total flow rate of 12.0 NmL/min/cm² on fuel side and air (clean and dry) on oxygen side. Stack performances were checked at each change of operating mode through i-V curves recorded at 800, 750 and 700°C. For the reference cell, durability tests were operated galvanostatic at 0.36 A/cm² in fuel cell mode and -0.51 A/cm² in electrolysis mode. The fuel utilization (FU) was respectively 42% in SOFC, and Steam Conversion (SC) 59% in SOEC. For the optimized cell, tests were carried out at 0.35 A/cm² (FU=41%), and -0.58 A/cm² (SC=67%). A second stage of more than 250h of operation consisted to alternate daily from SOEC to SOFC by cycles of respectively 8h and 16h. Finally, degradation was evaluated and despite the increase in SOEC current density during G2 cell testing, degradation rate was very similar for both cells in SOEC as well as in SOFC mode. A full size (25-cells) stack made of G2 cells was manufactured integrating the fluidic optimizations but also manifold and an electrical insulation necessary to operate stacks serially connected electrically within the REFLEX modules produced by Sylfen. The change of scale was validated by obtaining similar i-V curves recorded for 5-cells and 25-cells stacks. REFLEX Smart Energy Hub will be operated in ENVIPARK Torino using three modules comprising of four stacks each. Stack production has to be as uniform as possible and consequently, manufacturing process has to be optimized. Dispersion in performance, that results of cell production and stack manufacturing process is checked at final stage by i-V curves recording before delivery. Uniformity in production (in terms of tightness, reactant distribution and so initial electrochemical performance) is found to be satisfactory, the dispersion of median cell stack voltage over the stack production is less than the voltage dispersion observed between the whole cells at stack level. For near future needs in rSOC system, an increase of stack power density is required to both reduce the CAPEX and increase system energy efficiency. In that frame cells produced by Elcogen has an active cell area that was extended from 100 to 196 cm² and tested at stack levels of 5 and 25 cells, respectively. Enlarged stack performances are measured via i-V curves and compared to reference 100 cm² cells stacks. As difference in median cell voltage is less than 10 mV up to a current density of -1.25 V/cm² (80% SC), this paper shows that, from a mechanical and fluidic point of view, the enlarged cell were successfully integrated in full-scale stack. Figure 1

This work aims at developing an innovative renewable energy storage solution, based on reversible Solid Oxide Cell (rSOC) technology. That is to say, one system optimized to operate either in electrolysis mode (SOEC) to store excess electricity to produce H2, or in fuel cell mode (SOFC) when energy needs exceed local production, to produce electricity and heat again from H2 or any other fuel locally available. Firstly, work focused on optimization of the different layers constituting the single SOC cell to reach high initial performance applying state-of-the-art materials as previously reported [1]. Secondly, the initially highest performing cells were selected for long-term reversible SOFC/SOEC single cell tests. Thirdly, these cells were integrated in a stack design optimized for reversible operation at high degrees of H2 and H2O utilization. The long-term single cell tests showed significant degradation in galvanostatic test periods during electrolysis but not in fuel cell mode prior to starting the reversible test operation while the degradation diminished during the subsequent rSOC operation of the cells operating at 700 C, +0.6 and -1.2 A/cm2 in SOFC and SOEC modes respectively, at fuel utilization (FU) up to 80% in both modes. Electrochemical impedance spectroscopy analyses and post-mortem SEM investigations of tested single cells reveal that the fuel electrodes degraded significantly during the long-term single cell tests. Furthermore, long-term stack tests were conducted on 5-cell stacks, integrating both reference cells and optimised cells. The long-term stack tests were conducted applying different switches between SOFC and SOEC modes. Initially long duration tests (100h each mode) were performed in a mixture of 50% H2O and 50% H2 to see the effect of the polarisation only. The alternating cycle SOEC/SOFC was repeated over a 1800 h testing period. Then stack switched daily from SOEC mode (8h in 90% H2O and 10% H2 @ 0.84 A/cm2 FU 54%) to SOFC (16 h in 100% H2 @ 0.26A/cm2@ FU 60%) for full cycles over a testing period of 500 h. Stack tests were operated in reversible mode for up to several thousands of hours and results on the stack performance and lifetime will be presented.

The works presented here aim at developing an innovative renewable energies storage solution, so-called “Smart Energy Hub” (SEH), based on reversible Solid Oxide Cell (rSOC) technology. This systemis able to operate either in electrolysis (SOEC) or in fuel cell (SOFC) mode in order to store excess electricity to produce hydrogen, or when energy needs exceed local production, to produce electricity (and heat) again, from hydrogen or any other hydrocarbon fuel locally available. To be more efficient, the rSOC system is completed with an electrochemical storage solution allowing fast response to the electrical energy needs. In the frame of the REFLEX project, a SEH, made of 3 modules of 4 rSOC stacks, able to produce 16 Nm3/h in SOEC mode and up to 15 kWe in SOFC mode will be installed for an in-field demonstration in a technological park. Here the system will provide electricity and heat to the headquarters and maximize the auto-consumption of local renewable energies mixture. The rSOC system will be coupled to a Battery Energy Storage System (BESS), based on 8 lithium ion battery modules serially connected, able to store 50 kWh. Innovative work was done on each component to maximize, since this first try, the overall efficiency of the system prepared for field operation. Both cells and stacks have been optimised to increase performance (+20%) in both SOFC and SOEC modes, and their behaviour in rSOC operation have been characterized at different conditions. To increase the power density forward, enlarged cells have also been developed. The cell active area was near doubled passing from 100 cm² for reference cell to 196 cm². Concurrently, stack sectional dimension increased of about 40% only. This extension was validated up to stack level by performance testing of the enlarged stack. Voltage evolution, as current density increases, is very comparable for both stacks (reference and with enlarged cells). Voltage scattering is, also, quite similar. Steam starvation effect appears at the same level of steam conversion slightly higher than 80%, which shows that even for enlarged cells appropriate fluidic distribution was obtained. Large power modulations have been validated at stack level in three modes: SOEC, SOFC in H2 and SOFC in CH4, with ranges 58-100%, 23-100% and 13-100% respectively. To increase overall system efficiency by minimizing the energy losses in the balance of plant (BoP) components, specific reversible power electronics have been developed and an efficiency as high as 96% has been obtained, associated with a wide operation window (see figure). High performance heat exchangers and a steam generator able to generate stable steam flows have also been selected. Modelling activities have been performed to support the system design and to find the operating conditions in which the system to be built should/could be operated to ensure achieving maximum efficiency, while still operating the system in safe and lifetime-optimal conditions. The benefit of a recycling loop has been evaluated. Furthermore, techno-economic simulations have been carried out to evaluate market driven requirements on system sizing and costs to be competitive with other storage solutions. The article provides the results associated to these activities. Figure 1

Solid oxide cells are among the best candidates for viable energy conversion and storage solutions. Their capability to operate reversibly as electrolyser (‘SOEC mode’) or as fuel cell (‘SOFC mode’), provides an opportunity to balance out electricity needs. Yet, reversible operation requires an optimized cell and stack design to withstand alternating conditions. In this study, more than ten different cells have been produced and their initial performance evaluated. For selected cells, durability tests were conducted and showed that primarily the oxygen electrode contributes to performance losses in SOFC mode whereas the fuel electrode is mainly affected under SOEC mode conditions. The stack design was optimized in order to withstand the envisaged operating conditions. An optimization of internal pressure drop (reduced by a factor of 2) and sealing resistances towards higher pressure has been achieved while maintaining the electrochemical performance.

This work aims at developing an innovative renewable energy storage solution, based on reversible Solid Oxide Cell (rSOC) technology. That is to say, one system optimized to operate either in electrolysis mode (SOEC) for steam electrolysis to store excess electricity and to produce H2, or in fuel cell mode (SOFC) when energy needs exceed local production, to produce electricity and heat again from H2. In this work rSOC tests were conducted at single cell, in order respectively to finely understand performance and degradation mechanisms and to perform tests in representative conditions; and single cell rSOC test are finally compared to stack test results. Single cell rSOC tests were conducted to investigate the effect of current density, inlet steam content and degree of steam utilization both upon galvanostatic SOFC and SOEC testing, and upon load cycling operation. Cells having high initial electrochemical performance from Elcogen were used for these studies (1). Via four different long-term tests, we have investigated degradation at 700°C during constant galvanostatic SOFC (250 h) and SOEC (250 h) conditions followed by load cycling operation (up to 1000 h) in cycles of 16 h SOFC and 8 h SOEC. The applied current densities were either severe (0.6 A/cm2 in SOFC and -1.2 A/cm2 in SOEC) or moderate (0.4 A/cm2 and -0.6 A/cm2); applying different inlet gas composition ie. p(H2O)/p(H2) of 90/10 in SOEC and dry H2 in SOFC, or applying p(H2O)/p(H2) of 50/50 for both modes. Furthermore, two different steam and H2 utilization degrees of 80% and 40% were used. Figure 1 depicts examples of recorded cell voltage curves over time for two examples of the long-term rSOC testing where the inlet gas compositions were varied between p(H2O)/p(H2) of 50/50 for one test and p(H2O)/p(H2) of 90/10 and dry H2 for the other test. The electrochemical durability was assessed recording voltage over time and via electrochemical impedance spectroscopy (EIS). The characterization of the cells was supplemented by post-mortem scanning electron microscopy studies. Key findings include: 1) The cells degraded to a significantly larger degree in SOEC mode than in SOFC mode during the initial galvanostatic testing sequences. 2) At the given conditions and applying a current density of -1.2 A/cm2, the cells operate at a cell voltage above thermoneutral potential during the constant SOEC operation while this is not the case for the low-current density tests applying an electrolysis current of -0.6 A/cm2. 3) Load cycling does not increase the rate of degradation; rather the cell voltage stabilized for parts of the load cycling test sequences. 4) The major contribution to the resistance increase during long-term operation can be attributed to the fuel electrode and for the test, having the highest fuel electrode resistance increase (high current density test) a ~60% increase of the ohmic resistance was also observed during SOEC testing. 5) For the harshest tested (high current density) cells, Ni migration have been observed in parts of the fuel electrodes. 6) Nano-scaled detachment between Ni and YSZ particles were observed in several locations in several cells together with non-percolating Ni particles ie. where Ni-Ni connection has been lost; and a general trend was that the higher the fuel electrode overpotential during testing, the more pronounced was the effect of Ni-network changes upon microscopy post-mortem analysis in line with previous studies (2). 7) Inclusions of silica-containing impurities were found in the Ni/YSZ fuel electrodes as also observed in previous work (3, 4). Based on the single cell tests the recommendation for stack and system operation will, at 700 °C be, to operate in the current density regime of 0.4 A/cm2 and -0.6 A/cm2 rather than 0.6 A/cm2 and -1.2 A/cm2 to ensure long-term durability. Single cell test results were supplemented and trends confirmed by long-term rSOC test on 5-cell stacks at similar conditions. Figure 1: Cell voltage curves over time for two long-term rSOC test at 700°C. Inlet gas compositions of p(H2O)/p(H2):50/50 (blue curve) for one test and p(H2O)/p(H2):90/10 and dry H2 for the other test (red curve). Current densities of 0.4 A/cm2 and -0.6 A/cm2, respectively and a H2O (SOEC) and H2 (SOFC), utilization of 40%. References A. Ploner, A. Hauch, S. Pylypko, S. D. Iorio, G. Cubizolles, J. Mougin, in ECS Transactions (2019), vol. 91, pp. 2517–2526. A. Hauch, K. Brodersen, M. Chen, M. B. Mogensen, Solid State Ionics. 293, 27–36 (2016). A. Hauch, S. H. Jensen, J. B. Bilde-Sørensen, M. Mogensen, J. Electrochem. Soc. 154 (2007), doi:10.1149/1.2733861. M. Riedel, M. P. Heddrich, K. A. Friedrich, Fuel Cells. 20, 592–607 (2020). Figure 1