Original title: From Laboratory Aging Studies to Field Predictions: Quantifying Uncertainty in Battery Storage Lifetime Predictions
Melina Graner, Jan Figgener, Andreas Jossen, Holger Hesse · September 2026
Predicting how long a battery system will last is vital for warranties and investments. Most existing models give only a single fixed guess and are rarely checked against real-world systems. Researchers tackled this by using an open-source framework that calculates degradation while measuring uncertainty. They bridged cell tests from labs up to full home storage systems using a cell-to-system approximation. Their model predicts cell health to within 0.4 % mean absolute error, cutting the error of earlier models for this dataset roughly in half. It matched all three real-world field capacity tests available. To capture differences between cells, they tested two bounding stress cases with small gaps in heat and electrical current. The average paths reached end-of-life 10 months apart, and the full predicted spread spanned approximately three years. This wide uncertainty comes from lab tests not matching field stress, alongside variations in the training data.
Imagine you are baking a cake and the recipe says to bake at 350°F for 30 minutes. But if your oven is off by just 5°, the cake might be done 10 minutes earlier or later. Now imagine that the recipe itself was developed in a perfect test kitchen, not your home oven. In this paper, the battery is like the cake, the lab tests are the test kitchen, and the small temperature differences in real use cause large uncertainty in how long the battery will last.
Battery storage systems are expensive, and their owners need to know how long they will last to plan warranties and maintenance. This work shows that current lab-based predictions can be off by years due to small differences in operating conditions. By quantifying that uncertainty, it helps investors and engineers make more informed decisions about battery deployment and replacement.
The framework was validated against only three system-level capacity measurements from the field, which is a small sample. The cell-to-system approximation uses only two bounding stress scenarios, which may not capture the full range of real-world variability. The model is specific to residential battery storage systems and may not generalize to other applications without further validation.
Ask anything about this paper. Answers come from its abstract, and say so when the abstract doesn't cover your question.
From Laboratory Aging Studies to Field Predictions: Quantifying Uncertainty in Battery Storage Lifetime Predictions — Melina Graner, Jan Figgener, Andreas Jossen et al.
Abstract licence: CC0
Explanation generated by Synthesis from the abstract — it is not the original paper. This is a preprint, so it hasn't necessarily been peer reviewed.