Coupling electrochemical impedance spectroscopy and model-based aging estimation for solid oxide fuel cell stacks lifetime prediction

Abstract

In this work, an online natural aging estimation algorithm is developed, coupled with an Electrochemical Impedance Spectroscopy (EIS)-based diagnostic algorithm, to refine detection features extraction during Solid Oxide Fuel Cell (SOFC) stack operation and to predict its Remaining Useful Life (RUL). A combination of a lumped dynamic model along with features extracted from real-time EIS measurements is herein proposed for on-line applications. An Equivalent Circuit Model (ECM) is considered to identify parameters, such as ohmic and total resistance, that are coupled with an Area Specific Resistance (ASR) approach within the lumped model. The information derived from the EIS spectrum allows to estimate the voltage degradation over time along with its nominal behaviour. Indeed, the time trend of the identified parameters is proportional to the aging of the cell if no other abnormal condition occurs. This guarantees an on-line RUL estimation and a more robust diagnostic algorithm for fault detection and isolation. The approach has been applied to a 6-cells anode supported short stack tested for about 5000 h, and the related RUL estimation identified a critical issue on the middle cell, affecting its neighbours.