In short
Yes, in normal use. Hundreds of millions of lithium-ion batteries operate in the UK without incident, and the failure rate for reputable cells is very low. The reason they get attention is severity rather than frequency: when one does fail, the fire is fast, hot and hard to extinguish.
“Are lithium batteries safe?” is really two questions. How likely is a failure? And how bad is it when one happens? Conflating them produces either complacency or alarm, and neither helps somebody who has to decide where forty e-bike packs get charged.
How likely: rare, and not evenly distributed
The UK incident data available is dominated by one category, and it is not phones or laptops:
206
e-bike and e-scooter fires in London in 2025, an average of 17 a month. Around 83% involved an e-bike (171 fires). Two people died.
Source: London Fire Brigade, 28 January 2026.
211
e-bike and e-scooter fires recorded across the UK in 2024, with 86 injuries and 8 deaths. London Fire Brigade supplied 175 of those 211 notifications.
Source: Office for Product Safety and Standards, via the Department for Business and Trade, 21 July 2025.
Read those two figures carefully, because it is easy to misuse them. They cover different areas and different years, and the UK-wide total is built largely from London notifications — 175 of 211 — which means it reflects reporting practice as much as incident distribution. The safe conclusion is not “fires are concentrated in London” but “London measures this well and much of the rest of the UK does not yet”.
What the data does support is that the risk clusters. It clusters around e-bike and e-scooter packs, which are large, frequently charged, often modified and sometimes assembled from mixed components — rather than around the far larger population of phone, laptop and power tool batteries. Where a workplace has no e-mobility batteries, the profile is different and generally lower.
How bad: worse than the equivalent conventional fire
This is where lithium-ion differs. A cell in thermal runaway generates its own heat and releases its own oxidiser, so it does not respond to smothering. It vents flammable and toxic gas, usually before flame is visible. It can propagate to neighbouring cells, and it can reignite after appearing to be out. Premises fire precautions designed around conventional fire loads are not automatically adequate for this.
That asymmetry — low probability, high severity — is why the sensible response is neither “ban them” nor “carry on”. It is to reduce the probability where that is cheap (procurement, charging rules, inspection) and to limit the consequences where it is not (siting, separation, containment, detection).
What raises the risk
- Non-original, counterfeit or mismatched chargers — a recurring theme in UK reporting and in Electrical Safety First‘s campaign evidence.
- Conversion kits and self-assembled packs, where cells, BMS and charger were never designed to work together.
- Physical damage, including packs that were dropped and look undamaged.
- Charging unattended overnight, especially in escape routes or near sleeping accommodation.
- Storage in heat: uninsulated containers, vehicles, sunlit windows, next to heating plant.
- Keeping damaged or waste cells alongside healthy stock.
Every item on that list is a management decision rather than a property of the chemistry. That is the useful finding: the failure mode is fixed, but almost everything that leads to it is within a duty holder’s control.
Frequently asked questions
Are lithium batteries safe to leave charging overnight?
For a workplace, plan against it. Unattended charging removes the person who would notice the venting stage and act on it, and overnight is when premises are least able to respond. Where overnight charging is unavoidable — fleet operations, for example — the mitigation is location and detection: charge away from escape routes and sleeping accommodation, inside a tested enclosure where justified, with detection that alerts someone who can act.
Are e-bike batteries more dangerous than laptop batteries?
The chemistry is broadly the same; the exposure is not. An e-bike pack holds far more energy, is charged more often, lives a harder physical life, and is more likely to have been modified or fitted with a replacement charger. UK incident data reflects that difference clearly. It is a difference in how the battery is used and sourced, not in whether lithium-ion is inherently unsafe.
Should we ban e-bikes from our premises?
Some organisations do, and for premises with sleeping accommodation and no suitable external space it can be a defensible conclusion. It is worth recognising the trade-off: an outright ban often moves charging somewhere unsupervised rather than removing it. A permitted, supervised, well-sited charging arrangement usually manages the risk better than a prohibition that people work around. Whichever you choose, record the reasoning in your fire risk assessment.
How can I tell if a battery is about to fail?
Swelling, a change in shape, heat when not in use or on charge, unusual smell, discoloration, leaking, or a pack that suddenly loses capacity. Any of these means take it out of service and quarantine it — do not charge it and do not put it back in the store. There is no way to inspect for a developing internal fault, so visible signs are treated as decisive rather than as something to monitor.