The student Jon Perez Blanco obtained an EXCELLENT CUM LAUDE with mention INDUSTRIAL DOCTORATE
The student Jon Perez Blanco obtained an EXCELLENT CUM LAUDE with mention INDUSTRIAL DOCTORATE
The student Jon Perez Blanco obtained an EXCELLENT CUM LAUDE with mention INDUSTRIAL DOCTORATE
- Thesis title: Detection of aging and failure mechanisms in Li-ion Energy storage systems
Court:
- Presidency: Maitane Berecibar Uribe (Vrije Universiteit Brussel)
- Vocal: Henk Jan Bergveld (Eindhoven University of Technology)
- Secretary: Erik Garayalde Perez (Mondragon Unibertsitatea)
Abstract:
The increase of Electric Vehicles (EVs) in the market, and therefore, the increase of isolated Thermal-Runaway (TR) events have highlighted the need for measures to be taken in this matter, in response to the safety concerns among the industry and society. Thermal-Runaway events can happen over a certain period of time due to electrical or thermal abuse, or can happen suddenly due to mechanical abuse. Regulations such as UN ECE R100r3 or GTR-EVS have added safety norms by stating that a Thermal-Runaway event must be detected 5 minutes prior to the presence of a dangerous situation inside the passenger compartment.
This document presents the manuscript of the PhD thesis titled "Detection of aging and failure mechanisms in Li-ion Energy Storage Systems". The main objective of this doctoral thesis is to achieve an early detection of a Thermal-Runaway event and the mechanisms that may cause or eventually end in such a catastrophic event. This is achieved by studying the detection, prediction and prevention phases of a Thermal-Runaway event.
Firstly, Thermal-Runaway detection methods are investigated using both sensor-based and impedance-based approaches. Commercial gas and pressure sensors, along with techniques based on Electrochemical Impedance Spectroscopy (EIS) are experimentally evaluated under controlled abuse tests. An impedance-based detection methodology is proposed, developed, and validated across several use-cases. Early-warning times ranging from 6 to 12 minutes are achieved in fast-paced Thermal-Runaway events.
Secondly, Thermal-Runaway prediction is addressed through the early detection and quantification of Internal-Short-Circuits. Dedicated data generation frameworks are developed to support algorithm training and validation under a wide range of fault severities and operating conditions. Model-based and data-driven diagnosis methods are analyzed and compared. The proposed methods enable online detection during charging and more precise diagnosis during resting periods, detecting faults corresponding to self-discharge times of up to one week.
Thirdly, Thermal-Runaway prevention is investigated through the detection of Li-plating as a safety-critical aging mechanism. Voltage-based and impedance-based methods are analyzed using full cells, coin-cells, and three-electrode configurations, demonstrating that a combination of complementary techniques is required to detect both reversible and irreversible plating. An aging-aware operational strategy is proposed based on the strengths and limitations of each method.
Finally, the results obtained in this thesis are integrated into a multiphysical State-of-Safety framework based on Fuzzy Logic, combining indicators related to detection, prediction, and prevention into a single, interpretable safety metric. The framework is validated through multiple abuse tests across different degradation mechanisms, cell chemistries, and formats, demonstrating its suitability for real-time Battery Management System applications.