EV Battery Fires: A Timeline of the Dendrite That Started It All

Archi4 min read

evbatterysafety

Contents
  1. The short circuit that starts in microns
  2. 2010 to 2014: The separator as the weak link
  3. The plating condition
  4. 2015 to 2018: Fast charging raises the stakes
  5. The formation problem
  6. 2019 to 2021: Internal short circuits get measured
  7. What the data showed
  8. 2022 to 2023: Monitoring moves inside the cell
  9. The chemistry angle
  10. 2024 and after: What still has to change

The short circuit that starts in microns#

Most electric vehicle battery fires do not begin with a crash. They begin inside a cell, at a scale too small to see, where a needle of lithium metal pushes through a plastic membrane a few microns thick. That membrane, the separator, is the only thing standing between the anode and the cathode. When it fails, the cell shorts internally, and the chemistry does the rest.

This is a timeline of that failure, from the first stray lithium atom to the smoke that reaches the cabin.

Early lithium-ion cells for vehicles used polyolefin separators, typically 20 to 25 microns thick. They work well when the cell is new. The problem is mechanical: the separator must survive thousands of charge cycles, temperature swings, and the slow growth of lithium deposits on the anode surface.

Researchers had already documented lithium plating in the 1990s, mostly in lab cells. In vehicle packs, the issue reemerged because fast charging and cold weather both push lithium ions to deposit as metal instead of sliding into the anode's graphite layers. Once that metal forms, it does not sit still.

The plating condition#

Plating happens when the local current density at the anode exceeds what the graphite can absorb. Three factors drive it: charging at low temperature, charging at high rate, and a slightly uneven electrode surface. Automakers knew this. Early battery management systems limited charge current when pack temperature dropped below roughly 10 degrees Celsius. Those limits were conservative, and they worked, until owners wanted faster charging.

2015 to 2018: Fast charging raises the stakes#

As 50 kW and then 120 kW chargers spread, cells had to accept more current. Separator makers responded with ceramic-coated separators, adding a layer of aluminum oxide particles that resists puncture and shrinks less when hot.

The coating helped. It did not eliminate the root cause. Coating raises the puncture threshold, but a dendrite that grows long enough still finds a path. A 2018 study cycle-tested coated separators and found puncture resistance dropped after repeated plating and stripping, because the lithium metal roughens the anode surface over time.

The formation problem#

The first charge at the factory, called formation, sets the solid electrolyte interphase, a thin passivation layer on the anode. A poor formation protocol leaves weak spots. Dendrites prefer those spots. This is why two cells from the same production line can fail at very different ages.

2019 to 2021: Internal short circuits get measured#

By this period, test labs had a clearer picture of the failure sequence. A lithium dendrite grows through the separator, contacts the cathode, and creates a local short. Current concentrates at that point. Local temperature rises to several hundred degrees Celsius in milliseconds. The electrolyte, an organic solvent, vaporizes and decomposes. Gas pressure builds. The cell vents.

If the vented gas ignites, or if heat transfers to the next cell, the pack enters thermal runaway. A single 21700 cell holds roughly 12 to 18 watt-hours. A pack holds thousands of them, packed close together. Cell-to-cell propagation, not the first cell, is what turns a bad cell into a vehicle fire.

What the data showed#

Field data from that era suggested internal short circuits accounted for a meaningful share of EV fires, though fire-cause reporting is inconsistent across markets. Battery engineers started describing the problem in terms of probability per cell per year, a small number multiplied by millions of cells on the road.

2022 to 2023: Monitoring moves inside the cell#

Battery management systems historically read voltage and temperature at the module level. A short inside one cell can develop faster than a module sensor responds. Vendors began adding per-cell voltage sampling at higher rates and looking for the signature of a soft short: a slightly faster self-discharge on one cell compared with its neighbors.

Separately, separator suppliers pushed thinner, stronger films. Some moved to 12 micron base films with ceramic coating. Thinner separators improve energy density and lower internal resistance, which helps fast charging, but they reduce the mechanical margin. The engineering trade is real, and it does not disappear.

The chemistry angle#

Some cell makers shifted toward blends with silicon in the anode. Silicon holds more lithium, which raises capacity, but it swells during charging. That swelling stresses the separator and the electrode stack. It also changes where plating occurs.

2024 and after: What still has to change#

The root cause has not changed since the first lab cell. Lithium metal deposits on the anode, grows through the separator, and shorts the cell. Everything else, the coatings, the sensors, the cooling plates, the pack architecture, manages the consequences.

The next steps are concrete. Charging protocols that adapt to actual anode condition rather than pack temperature alone. Formation and aging steps that screen out cells with weak interphase spots before they ship. Separator designs tested against repeated plating cycles, not just a single puncture test. And pack layouts that accept that some cells will fail, and stop the heat from spreading.

A fire that starts in microns is not solved by a bigger extinguisher. It is solved by keeping the lithium where it belongs, and by assuming that on some cells, it will not stay there.

More articles