Search

High Temperature Steam Electrolysis (HTSE), based on solid oxide electrolysis cells (SOE) is a promising way to produce massively hydrogen with high efficiencies. This technology also allows producing syngas (H 2 + CO) by co-electrolyzing a mix of steam and CO 2 . This syngas constitutes the basis to obtain further synthetic fuels. For both HTSE and co-electrolysis, durability and cost are still key points, but additionally for co-electrolysis the outlet H 2 /CO gas composition has to be tailored to fit with the targeted fuels. Previous works have been carried out to develop a stack design suitable for both applications. Here experiments at stack level in both electrolysis and co-electrolysis modes have been carried out. A 10-cell stack and a 25-cell stack have been tested in electrolysis mode, respectively producing 0.6 Nm 3 /h and 1.7 Nm 3 /h of hydrogen at 800 °C below the thermoneutral voltage (1.3 V) for all the cells and a steam conversion around 50%, with a small scattering between the different cells. Gas tightness of the stacks has also been evaluated. Moreover, the operation in co-electrolysis has been validated. Finally a cost analysis of this stack design has been performed and shows all the economical potentialities of this technology.

This work details an in-house 2D electrochemical and thermal model of solid oxide electrolysis cell (SOEC) dedicated to the simultaneous electrolysis of steam and carbon dioxide. The apparent exchange current densities for both H2O and CO2 electrochemical reductions have been fitted on single electrolysis polarization curves obtained on a standard cathode supported cell (CSC). By using these values, predictive simulations have been performed in a co-electrolysis mode. The results have been compared to experimental data obtained in the same conditions. A good agreement is found between the simulated points and the experimental polarization curves, meaning that the model stands to predict accurately the SOEC response in co-electrolysis mode. Simulations have been performed to investigate co-electrolysis mechanisms in terms of partial pressures, current densities, and overpotentials distributions along the cell. The relative influence of both water gas shift (WGS) reaction and CO2 electrolysis on the global CO production is also discussed.

International audience; High-temperature electrolysis (HTSE) is a quite recent topic where most of the studies are focused on performance measurements and degradation observations, mainly achieved by polarization curve. However, it mainly leads to the overall cell behaviour. To get more specific knowledge on the operation of the cell, Electrochemical Impedance Spectroscopy (EIS) is more appropriate. In this study, EIS and chronopotentiometry were combined in order to characterize the electrochemical performance and behaviour of a commercial electrode-supported cell of Ni-YSZ/YSZ/LSCF type. A two-electrode configuration was used while a three-electrode one is required to better separate each component behavior. Nevertheless, it allows applying EIS to any single cell mainly when no good location for a reference electrode is available. Experimental parameters such as current density, temperature or PH2O/PH2 ratio were analysed. Using electrical equivalent circuit (EEC) combined to the distribution of relaxation time (DRT) and the analysis of the difference in impedance spectra (ADIS) approaches allowed deconvoluting impedance diagrams into three or four arcs characterized by their specific capacitance and relaxation frequency. Each arc was ascribed to a phenomenon related to the electrochemical reactions. This work corresponds to an in situ diagnosis by EIS of solid oxide electrolyser cell reaction mechanisms.

Solid Oxide Electrolysis (SOE) technology is considered to be a highly efficient process to produce hydrogen. It is liable to produce hydrogen at low cost, thanks to an electrical consumption as low as 3.5 kWh per Nm3of hydrogen produced, considering a heat source available to produce steam around 150°C. A smart heat management into the SOE system needs to be found to heat this steam at 150°C up to the operating temperature of the SOE which is in the range of 700-800°C. Therefore it appears that the steam conversion rate is a key parameter in the overall efficiency of the system. Most generally, due to concentration overpotentials leading to a limiting current at the cell level, to which a non-homogeneous distribution of gases into the stack can be added, steam conversion rates of 50% or below are reported. Then large flows of steam need to be overheated up to the operating temperature, which impacts the overall efficiency of the system. The present paper presents the methodology undertaken to develop a stack design able to operate at high steam conversion rates (80% or above). Starting from a reference stack design made of 25 cathode supported cells and ferritic stainless steel interconnects, modifications of the design have been made to improve the gas distribution to the cells and its homogeneity in the stack. This improvement has first been validated at the scale of a SRU (single repeat unit, the cell active area being equal to 100 cm²). I-V curves have been recorded at 800°C, with a cathodic inlet flow constituted of 90%H2O/10%H2and an anodic flow made of air, the total cathodic flow rate varying between 9 and 30 NmL/min/cm². It has been shown that with these design modifications, i-V curves can be recorded up to a steam conversion rate of 100%. For an inlet flow rate very low, equal to 9 Nml/min/cm², the 100% of steam conversion rate is reached for a current density of 1.15 A/cm² at 1.23V (figure 1). For an inlet flow rate of 12 Nml/min/cm², the 100% of steam conversion rate is reached for a higher current density of 1.56 A/cm² at 1.40V. In both cases, a limiting current appeared only far on the curve, for a current density of 1.06 A/cm² in the first case and 1.45 A/cm² in the second case. I-V curves have been plotted with several other inlet flow rates, and it has been shown that the lower is the inlet flow rate, the higher is the maximum steam conversion rate. However, even for the highest flow rates tested (24 and 30 NmL/min/cm²), the maximum reachable steam conversion rates remain respectively at values of 90 and 80%. These results validate the design modification, which was then implemented at the scale of a 25-cell stack. The reference stack design, prior any modification, reaches, at 800°C, a current density of 0.85 A/cm² at 1.2V in average per cell, for a steam conversion rate of 55% and a corresponding inlet flow rate of 12 NmL/min/cm². For this value of steam conversion rate, a limiting current starts to appear. On the contrary, with the design modification, no limiting current appeared on the i-V curve. For the highest current density recorded, equal to 1.0 A/cm², corresponding to an voltage of 1.17 V/cell, the steam conversion rate was of 64% with a corresponding inlet flow rate of 12 NmL/min/cm². Therefore it can be concluded that thanks to the design modifications, the limiting current is shifted to values well above 65%, and that the stack can be operated at high steam conversion rate. Several 25-cell stacks have been produced with either the reference design or the modified design. With the reference stacks, the voltage scattering among the 25 cells of the stack is around 15%, which is acceptable. Thanks to the design modification, the scattering decreases to a value around 5%. It shows that the design modification not only allows to reach higher steam conversion rates, but also improves the gas distribution homogeneity into the stack. Finally, a 25-cell stack with the modified design has been integrated into a SOE system including all BoP components. It has been possible to operate this stack, at the system level, up to a value of steam conversion of 80%. Figure 1

Introduction The integration of renewable energies to the global energy market implies electricity storage. However, the available technologies for electricity storage have a low energy capacity and a short time-span. In this context, H2 or carbon molecules produced via water and/or water+CO2 electrolysis open up excellent perspectives for storing renewable electricity, especially the High Temperature Electrolysis (HTE). This technology, based on Solid Oxide Cell (SOC), consumes less electrical energy than low temperature electrolysis because of the energy demand is significantly lowered by the vaporization of water1. The required electrical power is also reduced at high temperature since the steam electrolysis is increasingly endothermic with temperature. Moreover HTE can also directly convert a mix of water and CO2 into syngas, (CO+H2) (co-electrolysis), which can be converted into methane or other hydrocarbon molecules. Moreover, this type of co-electrolyzer can operate under pressure which increases the energy efficiency of the system, reducing the compression costs of the downstream chemical processing unit and increasing cells performance.2 In addition, as nickel present in SOC is known to catalyze CO and CO2 (COx) methanation and water gas shift reactions3, a pressure increase in the electrolyser leads to higher rates of these reactions. HTE cells can also be used in fuel cell mode to generate electricity from methane and water4. In this case, nickel catalyzes steam reforming leading to syngas which can be used to produce electricity. However, the contribution of these catalytic reactions are not well-known in co-electrolysis and fuel cell modes. In this view, we propose to evaluate their impact in a SOC fed with carbonated gases by establishing a kinetic model necessary for a complete electrochemical SOC modelling. This work was carried out on two different SOCs to establish kinetic models for co-electrolysis and steam reforming at atmospheric pressure and then under pressure. Catalytic tests The cells studied come from two different suppliers (A and B). In order to study the catalytic reactions, induced by nickel in the hydrogen electrode, other layers have been removed by sandblasting. COX methanation catalytic tests were performed at 400°C with the cells reduced into powder with different inlet mixtures (CO and/or CO2 with H2 with COx/H2 molar ratio around 0.3). Tests were carried out with a total flow rate of 35NmL.min-1 (equivalent to 346NmL.min-1.cm-2 cell and 245NmL.min-1.cm-2 cell for cells A and B, respectively). These flowrates are far from those commonly used for electrochemical tests, but they allow to distinguish the catalytic activity of the two cells. The results showed that cells are not equivalent to each other regarding COx conversion whatever the conditions used (CO and/or CO2) (Figure 1). It also can be noticed that the CO2 hydrogenation activity is reduced after exposure to a CO/H2mixture (Figure. 2) due to carbon deposition blocking the active sites. Secondly, catalytic tests were achieved to demonstrate the influence of water at the inlet (up to 15mol.% of water). The temperature was varied from 400 to 800°C after initial stabilization at 400°C. These tests confirmed the negative effect of water due to a less favorable thermodynamic. Finally, methane steam reforming catalytic tests were carried out to simulate fuel cell feeding conditions (H2O/CH4 molar ratio around 2) with a total flow rate of 5.3NmL.min-1 (equivalent to 52NmL.min-1/cm-2 cell and 37NmL.min-1.cm-2 cell for cells A and B, respectively). These conditions make it possible to approach the operating conditions of an electrolyser. The temperature was varied from 600 to 800°C after initial stabilization at 600°C. It appears that the A cell is close to thermodynamic equilibrium up to 700°C followed by a decrease at higher temperature whereas the B cell does not exhibit this decrease (Figure 5). Conclusions and outlooks The cells exhibit similar catalytic activities with respect to COX methanation and methane reforming. Simulations carried out using the COMSOL software allowed to select three kinetic models for atmospheric SOC operation5 , 6. To discriminate these models and propose a quantitative kinetic model for the SOCs operating under pressure, catalytic and electrochemical tests will be carried out up to 15 bars. Simulations will then be performed (COSMOL) to simulate the functioning of a cell by coupling catalysis, electrochemistry, gaseous and thermal flows. 1. Ni M. et al. Int. J. Hydrogen Energy 33,2337–2354 (2008). 2. Bernadet L. et al. Int. J. Hydrogen Energy 40,12918–12928 (2015). 3. Gao J. et al. RSC Adv. 5,22759–22776 (2015). 4. Er-rbib H. et al. Energy 75,81–88 (2014). 5. Wang Y. et al. Int. J. Hydrogen Energy 34,3885–3893 (2009). 6. Peters R. et al. J. Power Sources 106, 238–244 (2002). Figure 1

International audience; The present study is focused on alternative structured oxygen electrodes for solid oxide electrolysis cells (SOEC). The Ln2NiO4+δ (Ln = La or Pr) nickelate oxides were selected as innovative electrode materials with respect to their mixed electronic and ionic conductivity. A thin interfacial ceria-based layer was added in between the electrode and the zirconia-based electrolyte to improve mechanical and electrochemical properties and to limit the chemical reactivity. These structured cells were characterized by electrochemical impedance spectroscopy on symmetrical cells, under zero dc conditions and anodic polarization. Low polarization resistance RP and improved anodic overpotential ηA versus current density curves were obtained for the Pr2NiO4+δ electrode with Ce0.8Y0.2O2-δ interlayer: RP is decreased down to 0.06 Ω cm2 at 800 °C, under air and zero dc conditions. Then, complete hydrogen electrode-supported cells including Pr2NiO4+δ as oxygen electrode were electrochemically characterized. At 800 °C, when the inlet gas composition is 90 vol.% H2O-10 vol.% H2 at the hydrogen electrode, air being swept at the oxygen electrode, the current density determined at 1.28 V reaches -0.9 A cm-2, the corresponding steam to hydrogen conversion ratio being 58%. These results are compared to those obtained with a reference cell including the oxygen deficient perovskite La0.6Sr0.4Fe0.8Co0.2O3-δ as oxygen electrode.

High Temperature Steam Electrolysis (HTSE) is one of the most promising ways for hydrogen production. If coupled to a CO2-free electricity and low cost heat sources, this process is liable to a high efficiency. The present study describes recent promising results obtained in terms of performance and durability in stack environment, thanks to the use of protective coatings on one hand, and of advanced cells on the other hand. As for Solid Oxide Fuel Cells, it has been demonstrated that the integration of protective coatings was mandatory to decrease the degradation rate in HTSE stacks, and that with optimized coatings, (CoMn)3O4 in the present case, the same durability as the one of the single cell tested in a ceramic housing could be reached. The type of cell was also shown to play a major role on the degradation rate. With advanced cells, degradations below 2%/kh could be reached. The higher is the current density, the higher is the degradation rate, with a mostly reversible effect. These degradation rates are close to the objectives, even if a bit higher than in SOFC mode. Finally a low-weight stack has been designed, targeting high performance and durability while reducing the cost by the use of thin interconnects. An electrochemical performance similar to the previous stack design has been obtained for a 3-cell stack (-1 A/cm2 at 1.3 V at 800 °C), with degradation rates below 3%/1000 h in the testing conditions. The thermal cyclability of stacks has been demonstrated, from 800 °C to 20 °C, as well as electrical load cycling. The results showed that the HTSE stacks considered in the present study can cycle very rapidly, and that the cycles considered do not induce any degradation. Therefore it can be concluded that these results makes HTSE technology getting closer to the objectives of performance, durability, thermal and electrical cyclability and cost, and that HTSE is a candidate to produce hydrogen as a mean to store renewable intermittent energies.