Condition monitoring of solid oxide fuel cells by fast electrochemical impedance spectroscopy: A case example of detecting deficiencies in fuel supply

• Published in: Journal of power sources 2021-03, Vol.489, p.229491, Article 229491
• Main Authors: Nusev, Gjorgji, Morel, Bertrand, Mougin, Julie, Juričić, Ðani, Boškoski, Pavle
• Format: Article
• Language: English
• Subjects: Chemical and Process Engineering; Electrochemical impedance spectroscopy; Engineering Sciences; Fuel utilisation; Monitoring; Non-linear optimisation; Signal and Image processing; Solid oxide fuel cell
• ISSN: 0378-7753; 1873-2755
• DOI: 10.1016/j.jpowsour.2021.229491

This paper deals with the monitoring of the internal condition of a 6-cell solid oxide fuel cell stack during a 3600-hour test with short fuel starvation intervals. The relationship between change in condition and change of impedance spectra as well as the parameters of the equivalent circuit models are analysed. Tentative features indicating reduced fuel supply are identified. The analysis uses discrete random binary sequence, a stochastic broadband excitation signal that allows much shorter perturbation times. More than 600 electrochemical impedance spectroscopy curves and the associated equivalent circuit model parameters were analysed. High fuel utilisation interval is detected in two ways: (i) by violating the Z-HIT test in the low-frequency part (below ≈ 1 Hz) and (ii) significant change of the area-specific resistance parameter. The degradation after fuel cut-offs is clearly reflected in the change in area-specific resistance and the serial resistance. The numerical implementation of the proposed algorithms is available at: https://repo.ijs.si/pboskoski/py_eis.git. •Electrochemical impedance spectroscopy using Discrete-Random Binary excitation.•Short measurement time with more than 50 impedance points per decade.•Evolution of ECM parameters from 600 EIS curves measured in more than 3600 h.•Accurately detection of high fuel utilisation at SOFC stacks using Z-HIT.•Computationally efficient algorithm for monitoring SOFC performances.