Application of Modified Thermographic Method in SOC Stack with Regenerated Fiber Bragg Gratings

ECS Meeting Abstracts, vol. MA2025-03, pp. 288

Abstract

Solid oxide cells (SOCs) offer multiple advantages, such as high efficiency, fuel flexibility and the readiness for reversible operation. The electrochemical performance of the SOCs is typically studied using direct and alternating current techniques, such as current voltage measurement and electrochemical impedance spectroscopy. These methods are powerful tools to investigate the performance and the underlying degradation mechanisms of SOCs. However, they only capture the response of the whole cell, providing very limited spatial information. In order to estimate the local behaviour of the cell using these methods, a segmented cell design is often required 1, 2 , which significantly increases the complexity of the experimental setup. Electrochemical reactions inherently involve thermal effect. Depending on the operating and thermoneutral voltage, the reaction can be exothermic, endothermic or thermoneutral. While the temperature profiles of an energy conversion device are influenced by various factors, analyzing the thermal characteristics can still provide valuable insights into the local cell activity. Infrared thermography has been used for detecting cell failure 3 and monitoring the operation of pellet type solid oxide cells 4 with high level of spatial and temporal resolution. Jeanmonod et al. 5 applied active thermographic methods to characterize the SOCs, demonstrating their added benefits as complementary sensing techniques. The active thermography involves an application of external stimulus, such as periodic (lock-in thermography) or pulse (pulse thermography) current perturbations, which greatly enhanced measurement sensitivity. Based on these advancements, we proposed a “modified thermographic method” to estimate the distribution of local current density in the solid oxide cell stacks. The heat production rate q of an electrochemical reaction at cell voltage U cell is given as: q = (E H -U cell )I where E H denotes the thermoneutral voltage, and I is the current. In a SOC stack, the high electrical conductivity of the metal interconnect ensures that the electrical potential difference between the fuel and air electrodes remains uniform across the entire active cell area. Since both E H and U cell are known, the only determining factor for heat generation is the current. For a finite area of the cell, the energy balance can be expressed as: c p m(dT/dτ) = q + q conductive + q convective + q radiative Where (dT/dτ) is the derivative of temperature T over time τ , c p is the specific heat capacity, m is the mass, q is the heat production rate mentioned earlier, q conductive, q convective , and q radiative represent the heat transfer contributed by heat conduction, heat convection and thermal radiation, respectively. During steady state operation, the inhomogeneous temperature distribution and the complex heat dissipation make it challenging to directly correlate the thermal response of the SOCs with local current density. Under open circuit conditions, the right-hand side of this equation sums to zero when the stack reaches thermal equilibrium. When a direct current pulse is applied, the gas composition and flow field quickly reestablish, whereas the temperature profile takes significantly longer to stabilize. Before any substantial temperature variation occurs, the conductive and thermal radiation terms should remain largely unchanged. The heat transfer coefficient is influenced by the changes in gas flow rate and composition, but this effect can be considered negligible, especially when the fuel and air conversions are relatively low. Furthermore, in the active cell area, the local heat capacity can be assumed to be positionally independent. Based on these assumptions, the local temporal thermal gradient (dT/dτ) should be linearly proportional to the local current density. Due to the intricate structure and gas tightness requirement of the stack, the conventional infrared imaging cannot access the active cell area without significant modification of the established stack and test rig design. As a result, thermographic methods are primarily limited to single cell testing. At the stack level, the temperature measurements are typically performed using thermocouples placed in the interconnects 6, 7 . However, achieving high spatial resolution with this approach is challenging due to the substantial cabling requirements, and the heat capacity of the interconnect prevents the thermocouples from locally capturing the intrinsic temporal thermal gradient in the active cell area. To overcome these limitations, the present study employed an optical fiber sensor, the regenerated fiber Bragg gratings (RFBG), for temperature measurement in the SOC stacks. This sensor functions in essence as a wavelength specific reflector with multiplex capability and well-defined sensitivity towards temperature 8, 9 . In the first stage of this work, a standard communication optical fiber inscribed with seven RFBG sensors was installed in an air channel to capture the local thermal response within the active cell area. The approximate positions of all RFBG sensors are schematically illustrated in Figure 1. The wavelengths of the RFBG sensors were sampled at 10 Hz using a commercial interrogator during the measurement. The five layer Jülich F10 stack was supplied with 0.44 L/min of H 2 , 1.33 L/min of CO 2 and 2.67 L/min of steam as reactants and 5 L/min of air in a counterflow configuration. All volumetric flows are referenced to the condition of 1.013 bar and 273.15 K. The stack was maintained at 800°C under open circuit condition until there was no visible change in stack temperature before applying the current pulse. Figure 2 presents the temporal thermal gradients recorded with a -20 A current pulse averaged over six consecutive measurements. The highest temperature change rate was recorded by RFBG 5, positioned between the middle and the air outlet side of the cell. This rate gradually decreased from RFBG 5 to 2, indicating a reduction in local current density. At both edges of the active cell area, the gradients measured by sensors 2 and 6 were relatively low. The error bars were calculated based on the standard deviation of these six measurements and the results were found to be highly reproducible. In this study, we demonstrated a modified active thermographic measurement using regenerated fiber Bragg grating sensors to analyze the spatial electrochemical performance of the SOC stack. The stack was tested under different operating conditions and the change in the local current density profile, as the

Authors 4

1. Shengzhi Liang Aachen Institute of Physical Chemistry

   Forschungszentrum Jülich · RWTH Aachen University

   Affiliation as printed

   Forschungszentrum Jülich, IET-1: Fundamental Electrochemistry

   RWTH Aachen University, Institute of Physical Chemistry

2. Dominik Schäfer

   Forschungszentrum Jülich

   Affiliation as printed

   Forschungszentrum Jülich, IET-1: Fundamental Electrochemistry

3. Remzi Can Samsun

   Forschungszentrum Jülich

   Affiliation as printed

   Forschungszentrum Jülich, IET-1: Fundamental Electrochemistry

4. Rüdiger‐A. Eichel Aachen Institute of Physical Chemistry

   Forschungszentrum Jülich · RWTH Aachen University

   Affiliation as printed

   Forschungszentrum Jülich, IET-1: Fundamental Electrochemistry

   RWTH Aachen University, Institute of Physical Chemistry