DLR-Logo -> http://www.dlr.de
DLR Portal Home | Imprint | Privacy Policy | Contact | Deutsch
Fontsize: [-] Text [+]

Characterization and Quantification of Metallic Interconnect Degradation in Solid Oxide Fuel Cell Stacks

Linder, Markus (2016) Characterization and Quantification of Metallic Interconnect Degradation in Solid Oxide Fuel Cell Stacks. Dissertation, Universität Stuttgart.

[img] PDF - Registered users only


Stationary solid oxide fuel cell (SOFC) applications typically have targeted lifetimes of more than 40 000 h with a tolerable power degradation of less than 1% per 1000 h. To achieve these requirements powerful methods and strategies to characterize and quantify SOFC stack degradation are needed. The objective of this work was the Investigation and prediction of SOFC stack degradation based on in-situ measurements and post-test analyses. These results were combined with the findings from dedicated laboratory experiments for the investigation of single degradation effects combined with computer models to develop strategies to predict and subsequently reduce SOFC stack degradation. This requires the development of a model for the reliable quantification and thorough Interpretation of the degradation behavior of SOFC stacks under real, non-ideal conditions. The model extracts the time-dependent internal stack resistances from current-voltage data. As major advantage over assessing stack degradation simply from the slopes of current-voltage curves, these internal resistances can be directly compared with the sum of the resistances of different stack components as obtained from dedicated laboratory degradation experiments. Due to recent progress in improving the oxidation re-reduction (redox) stability of electrolyte supported SOFCs, the main contribution to the power degradation of stacks operated of more than 10 000 h is induced by the ohmic losses caused by the time-dependent and continuously formation of oxide scales on the metallic interconnects (MICs). Moreover, the oxide scale growth rate and the extrinsic electrical conductivity of the formed semi conductive Cr2O3 is influenced by various parameters such as scale morphology, temperature, gas atmospheres, impurities, reactive elements and interaction with adjacent components. For this reason oxide scale growth on chromium (CFY) and ferritic (Crofer) based alloys were investigated by scanning electron microscopy (SEM) including samples from SOFC stacks manufactured by the Hexis AG1, which were operated for up to 40 000 h. Comparison of the measured increase in ohmic resistance with mean scale growth rates obtained from SEM cross section images revealed a non-trivial, non-linear relation-ship. To understand the correlation between scale evolution and resulting ohmic losses, 2D finite element (FE) simulations of electrical current distributions were performed for a large number of scale morphologies. Oxide scale morphologies favor non-homogeneous electrical current distributions, where the main current flows over rather few “bridges”, i. e. local spots with relatively thin oxide scales. Combining electrical conductivity and SEM measurements with FE simulations revealed two further advantages: it allows a more reliable extrapolation of MIC degradation and it provides a new method to assess the effective electrical conductivity of thermally grown Cr2O3 scales under stack operation. On the anode side the degradation behavior of oxide scales is even more complex. For example, Ni particles released during thermal redox cycles from adjacent Ni containing components might be interspersed into the oxide scale. To study the influence of this interaction Cr2O3 pellets admixed with different amounts of Ni (up to 20 vol.%) were produced. The electrical conductivity was investigated in-situ in reducing forming gas (95% N2; 5% H2) atmosphere and air at 850. Furthermore microstructure and crystal structure are studied at different time steps with SEM and X-ray diffraction (XRD), respectively. Based on the applied methods it can be confirmed that during oxidation in air Ni forms a NiCr2O4 spinel phase. Exposure again in a reducing environment leads to an instantaneous decomposition of this spinel phase, which in turn leads to a fine dispersion of reduced Ni particles. This rearrangement of Ni by spinel decomposition improves the electrical conductivity of the Cr2O3-pellets. In conclusion, this work established new methods and models to more reliably assess of SOFC stack-degradation based on current-voltage data, which allow extracting internal stack resistances for direct comparisons with dedicated laboratory experiments of single degradation phenomena. The focus of attention is on the quantification and understanding of the complex degradation behavior of metallic interconnects (as major contribution to the overall SOFC stack degradation) caused by oxide scale growth under reducing and oxidizing conditions.

Item URL in elib:https://elib.dlr.de/109748/
Document Type:Thesis (Dissertation)
Title:Characterization and Quantification of Metallic Interconnect Degradation in Solid Oxide Fuel Cell Stacks
AuthorsInstitution or Email of AuthorsAuthors ORCID iD
Linder, MarkusMarkusLinder (at) icloud.comUNSPECIFIED
Date:October 2016
Refereed publication:Yes
Open Access:No
Gold Open Access:No
In ISI Web of Science:No
Number of Pages:211
Keywords:Interconnect ageing Scale Formation Modelling Performance curves SOFC
Institution:Universität Stuttgart
HGF - Research field:Energy
HGF - Program:Storage and Cross-linked Infrastructures
HGF - Program Themes:Fuel cells
DLR - Research area:Energy
DLR - Program:E EV - Energy process technology
DLR - Research theme (Project):E - Electrochemical Processes (Fuel Cells) (old)
Location: Stuttgart
Institutes and Institutions:Institute of Engineering Thermodynamics > Electrochemical Energy Technology
Deposited By: Friedrich, Prof.Dr. Kaspar Andreas
Deposited On:19 Dec 2016 16:03
Last Modified:19 Dec 2016 16:04

Repository Staff Only: item control page

Help & Contact
electronic library is running on EPrints 3.3.12
Copyright © 2008-2017 German Aerospace Center (DLR). All rights reserved.