Kurzbeschreibung
(Englisch)
|
Lithium-based electrochemical batteries are gaining increased popularity with a global production capacity of 200 GWh in 2019. The vast majority of the annual production is reserved for mobility applications, where the batteries are used down to typically 80% of the initial capacity. To qualify such batteries for second-life applications it is necessary to determine the remaining capacity and inner resistance of each cell, which is achieved via a full C-D-C (charge-discharge-charge) cycle. In this project we intend to verify the validity of an accelerated testing approach that was developed at CSEM via a statistically representative number of different Li-ion cells (i.e. different underlying chemistries: NMC, NCA) from e-bike applications provided by Libattion.
|
Publikationen / Ergebnisse
(Deutsch)
|
Der starke Ausbau der Elektromobilität führt mittelfristig zu einem grossen Rücklauf gealterter Batterien aus Elektroautos. Diese Stromspeicher haben ihre ursprüngliche Leistungsfähigkeit eingebüsst, die Mehrheit der darin enthaltenen Batteriezellen ist aber weiterhin intakt. Diese können für den Bau sogenannter Second-Life-Batterien verwendet werden. Bevor Batteriezellen in einem zweiten Lebenszyklus genutzt werden können, muss ihr Zustand (‹State of Health›/SoH) bestimmt werden. Ein Forschungsteam aus Neuenburg hat in Zusammenarbeit mit einem Zürcher Industriepartner nach effizienten Messverfahren gesucht.
Zugehörige Dokumente
|
Publikationen / Ergebnisse
(Französisch)
|
Le fort développement de la mobilité électrique entraîne à moyen terme un important retour des batteries usagées des voitures électriques. Ces accumulateurs ont perdu leurs performances initiales, mais la majorité des cellules de batterie qu’ils contiennent sont toujours utiles. Ces dernières peuvent être utilisées pour la construction desdites batterie de seconde vie. Avant que les cellules de batterie puissent être utilisées dans un second cycle de vie, leur état doit être déterminé (‹ State of Health ›/ SoH). Une équipe de recherche de Neuchâtel, en collaboration avec un partenaire industriel zurichois, a cherché des méthodes de mesure efficaces.
Zugehörige Dokumente
|
Schlussbericht
(Englisch)
|
CSEM and Libattion have partnered in the BAT4SEL project to verify the validity of an accelerated testing procedure to estimate the SoH of second-life Li-ion cells (i.e., used Li-ion cells). The project objective was to give a statistical quantification of the trade-off between the SoH estimation precision and the required testing time to obtain this estimation; for this reason, a statistically representative number of samples from e-bike applications has been provided by Libattion to CSEM. The whole statistical analysis was based on the test of a high number of cells of four selected technologies to develop meaningful indicators which can be correlated with the SoH evolution. The testing methodology has been based onto two different sub-set of tests: (i) Diagnostic tests, where a representative sample of cells have been characterized to find existing correlations between indicators and SoH and (ii) validation tests, where a larger number of 5 samples have been tested to validate the protocols by comparison of the estimated and measured SoH values. Three main type of indicators have been applied: resistance-based indicators, efficiency-based indicators and EIS-based indicators. Moreover, the correlation assessment has been verified for three different testing conditions: (i) random voltage condition (no regulation of cell’s SoC); (ii) nominal voltage condition (regulation of cell’s SoC); (iii) random voltage filtered (regulating cells inside a specific voltage interval). The diagnostic tests showed that there is not an indicator prevailing on the others and that experiments at random voltage rarely provide a correlation. The results of validation tests showed that resistancebased indicators provided on average an estimation error below 2.5% and efficiency-based indicators showed a stable estimation error around 2%. EIS-based indicators showed slightly higher estimation errors (up to 3%) but lower standard deviation than in the two previous cases. Across all testing conditions all indicators bring a significant reduction on the testing time of at least 70%, with respect to the 5 hours standard discharge-charge-discharge cycle. In the case of tests at random voltage with filtering option, resistance-based indicators require 8 minutes (-97%), EE-based indicators require 19 minutes (-93%), while EIS-based indicators require 45 minutes (-85%). It should be noticed that uncertainties and larger errors could be introduced by the testing equipment; only high-end test equipment has been used along the diagnostic phase which ensured high accuracy and precision on the measurements and correlations. Moreover, initial bad conditions of the battery cells (e.g., deeply discharged) do introduce larger standard deviation; for this reason, the random voltage filtering approach provided the most reliable results In conclusion it can be observed that the efficiency-based indicators represent the best compromise for a precise and fast SoH estimation of second-life Li-ion cells. However, the choice of the most proper indicator to be used depends on which aspect is the most important for the final user: estimation accuracy, robustness (i.e., giving consistent correlations among cell technologies), testing time and machine applicability.
Zugehörige Dokumente
|
