Battery Life Cycle Battery Life CycleIndependent research

Who

Dmitri Stepanov

Independent researcher — the ageing and degradation of batteries. No institutional affiliation.

ORCID 0009-0006-5549-0096 dvs@battery-lifecycle.net +358 40 351 8716 Helsinki, Finland

What I work on

Why published ageing measurements disagree

Measurements of how batteries age are almost always printed without the conditions they were made under — so two papers can disagree completely and both be right.

I collect those measurements together with their conditions, and work out which of them can be compared at all. The work runs under the name Battery Life Cycle.

What ageing does to a cell, measured

An impedance spectrum is one of the few measurements that shows ageing directly. Below: the same cell, measured when new and again after losing about 40 % of its capacity. Drag the slider to walk both spectra frequency by frequency.

Cell KIT_P018_1_S01_C06 · LG INR18650HG2 · NMC new · SOH 100 % aged · SOH 59.5 %
0.000.020.040.060.080.10-0.015-0.010-0.0050.0000.0050.0100.0150.020Z′ — real part, Ω−Z″ — imaginary part, negated, Ω
New · Z′
0.0179
New · Z″
0.0018
Aged · Z′
0.0428
Aged · Z″
0.0113
Cycled at 0 °C — charge 0.33C (1 A) CC-CV with a 0.3 A cut-off, discharge 1C (3 A), window 2.5–4.2 V. Frequency range 0.05 Hz – 14.7 kHz; each spectrum covers 29 frequencies, of which 28 are drawn: the dataset’s authors set the 14.7 kHz point to “not a number” as implausible, so it is empty in both spectra. The vertical axis is the negated imaginary part, the usual convention for cells and the one used in the source. Read plainly: cell resistance grows as the cell ages. No interpretation of the individual arcs is offered here.

Data: Luh M., Blank T., Karlsruher Institut für Technologie. Scientific Data, DOI 10.1038/s41597-024-03831-x; dataset RADAR4KIT, DOI 10.35097/1969, version 2. Licence CC BY 4.0. Values converted from milliohms to ohms, and two curves selected out of 76 reference measurements of one cell.

What the work rests on

2,495cells with measured ageing records, across seven cathode chemistries — NMC, NCA, LFP, LCO, a mixed NMC–LCO, LMO and lithium titanate; for 21 of them the source does not name the chemistry
463electric-vehicle battery packs with charging records, plus module-level sets — the literature is overwhelmingly about single cells, while the pack as a built object is described far less
2,108papers in the index: 1,285 on the subject itself, 803 adjacent, and 20 with no copy available
2,344verification records kept against the ASME V&V 10 and V&V 20 standards

Figures as of 17 September 2026. The verification count is a running journal and grows with each check; the rest stand until a new dataset or paper is added.

Measured data are lithium-ion throughout. The literature is read more widely — sodium-sulfur, sodium-ion, zinc-manganese, anode-free zinc, potassium-ion and flow systems are followed as adjacent work, with no measurements of my own on any of them.

Alongside the measurements, an animated illustration of a cell was built for this work — one lithium ion traced through a full cycle, with live controls. Open the animation → Two longer written explainers can be read there as well.

How the numbers are collected, and what is still missing →

Dmitri Stepanov · Independent researcher · Helsinki, Finland dvs@battery-lifecycle.net · ORCID 0009-0006-5549-0096

Last updated 17 September 2026