What actually happens when a home battery goes into thermal runaway

By Declan Osei · August 31, 2026 · 7 min read
Electric vehicle charging cable connected to a car
Photo by Ernest Ojeh on Unsplash

Three years ago I sat through a manufacturer training session where the technician put up a slide of a battery module mid-failure, off-gassing visibly, and said "this is the thing we're all designing around." Not fires, specifically. Off-gassing. That stuck with me, because most homeowners asking about battery safety are actually asking the wrong question. They want to know "can it explode," when the more useful question is "what has to go wrong first, and how many separate failures does it take."

I've owned an EV for years and run a home battery through two full southern hemisphere summers, so this isn't academic for me. I've also had enough conversations with installers who deal with fault call-outs to know the difference between marketing-safe and engineering-safe. They're not the same thing.

What thermal runaway actually is, mechanically

Thermal runaway is a self-sustaining chemical reaction inside a lithium-ion cell where heat generation outpaces heat removal. It doesn't start as fire. It starts as an internal short, or a manufacturing defect, or physical damage, that raises a cell's internal temperature past a threshold. Past that threshold, the cell's own chemistry starts producing more heat than the surrounding structure can dissipate. That heat drives further decomposition of the electrolyte and electrode materials, which produces more heat, which decomposes more material. It's a feedback loop, not a spark.

The visible stages, roughly in order, are: gas venting (the cell releases flammable and sometimes toxic gases as pressure builds internally), then potentially ignition of those gases if there's an ignition source and enough oxygen, then propagation to adjacent cells if the heat transfers faster than the pack's design can arrest it. That last step, propagation, is the entire safety battle. A single cell failing in a well-designed pack is a contained, manageable event. A single cell failing in a poorly designed pack becomes every cell failing in sequence.

This is why cell-to-cell spacing, thermal barriers between modules, and battery management system (BMS) monitoring aren't decoration on a spec sheet. They're the actual safety engineering, and they matter more than the chemistry label on the box.

Why LFP genuinely behaves differently, not just marginally better

I've written before about the chemistry differences between LFP and NMC home battery chemistry, so I won't re-run that whole comparison here. But it's worth being precise about the thermal runaway angle specifically, because this is where the two chemistries diverge most sharply.

Lithium iron phosphate has a stronger phosphate-oxygen bond in its cathode structure than nickel-manganese-cobalt chemistries. Practically, that means LFP cells have a meaningfully higher onset temperature before runaway begins, and when they do vent, they release less oxygen to sustain a fire. NMC cathodes release oxygen more readily under thermal stress, which is part of why NMC failures tend to be more energetic and harder to arrest once started.

I want to be careful with the claim here, because I've seen it oversimplified into "LFP can't catch fire," which isn't true. LFP cells can and occasionally do enter thermal runaway. The difference is the onset threshold is higher and the failure, when it happens, tends to be slower and more containable. That's a real safety margin, not a marketing line, but it's a margin, not immunity.

Almost every mainstream residential battery sold in Australia and the US today, including the Powerwall 3 and the Battery-Box Premium HVS, uses LFP chemistry specifically because of this thermal margin. That's a fairly settled industry direction now, not a niche preference.

What the BMS is actually doing while you're not looking

The battery management system is the unglamorous part nobody asks about at the sales stage, and it's doing more real safety work than almost anything else in the box. A competent BMS is continuously tracking individual cell voltage, temperature at multiple points across the pack, and current flow, and it's set to disconnect or throttle the pack the moment any of those readings drift outside a safe band.

Overcharge protection, over-discharge protection, over-current protection and thermal cutoffs are the four jobs that matter most. A cell pushed past its upper voltage limit repeatedly is a cell whose internal structure is degrading in ways that raise runaway risk over time, not just in the moment. This is one reason I keep telling people that how you set your charge and discharge windows isn't just about bill optimisation. Hammering a pack to 100% every single day, or discharging it flat every night, adds cycling stress that a well-configured BMS will tolerate but that shortens the margin over years of daily abuse.

Where I'll push back a little on the industry: some cheaper packs sold through grey-market channels have BMS firmware that's noticeably less conservative than the tier-one brands, because tighter thresholds cost a little usable capacity. I think that trade-off is being made in the wrong direction for a device sitting on a garage wall. A percent or two of extra usable kWh isn't worth a thinner safety margin.

Why installation location is a safety control, not a preference

I've covered the garage-versus-wall-versus-outside question in detail elsewhere on the site, so I'll keep this section tight: where you install a home battery is itself part of the thermal runaway safety chain, not a separate topic from it.

The relevant standard in Australia is AS/NZS 5139, which governs battery system installation including clearances from doors, windows, egress paths and habitable rooms. In the US, the National Electrical Code (NEC) Article 706 and local fire codes cover similar ground, alongside UL 9540 and UL 9540A testing that assesses how a given battery product actually performs in a propagation test, not just what the chemistry claims to do. These aren't bureaucratic boxes to tick. They exist because a battery installed close to an egress path or under a bedroom window changes the outcome if venting does occur, regardless of how good the chemistry is.

I'd argue the clearance requirements in AS/NZS 5139 are sometimes treated by installers as the minimum to survive an inspection rather than the actual safety intent, and that's a mistake worth pushing back on if your installer proposes mounting a battery right beside a bedroom wall just because the numbers technically clear.

Sizing decisions that quietly affect thermal risk

Most people think about battery sizing purely in terms of self-consumption or backup runtime, which I've walked through in our sizing guide for those two different jobs. But there's a thermal dimension to sizing that gets skipped. A pack that's undersized for your daily cycling needs gets pushed harder, more often, closer to its charge and discharge limits, which is exactly the stress pattern that degrades cell chemistry fastest and narrows the safety margin over years of ownership.

Oversizing isn't automatically safer either. A bigger pack sitting mostly idle at high state of charge for long periods carries its own, smaller, stress profile, since lithium chemistries generally prefer not to sit at the very top of their charge range for extended stretches. The sweet spot is a pack sized to your actual daily throughput, cycling in a moderate band, rather than one chosen purely off a headline capacity number on a quote.

What homeowners can actually check before signing

You're not qualified to inspect a BMS firmware table, and you shouldn't need to be. But there are a handful of things worth asking a quoting installer directly, and the answers tell you a lot about whether you're dealing with someone who takes this seriously.

Ask whether the proposed battery has UL 9540A test data available (US) or complies with AS/NZS 5139 clearance requirements as installed, not just as a product spec (Australia). Ask what chemistry it uses and why they chose it. Ask where exactly it will be mounted relative to doors, windows and habitable rooms, and get that in writing on the plan, not just verbally. A properly itemised quote should show this detail rather than a single line item for "battery supply and install," and I've written separately about what a real battery quote should itemise. If an installer can't answer clearance and chemistry questions without checking, that's a signal worth weighing alongside price.

For the regulatory detail itself, the Clean Energy Council's battery storage guidelines and the US Fire Administration's residential energy storage safety material are both worth a direct read rather than relying on installer paraphrasing.

None of this is meant to make anyone nervous about buying a home battery. I run one, and I'd do it again. The risk profile of a well-installed, tier-one LFP battery, mounted to code, on a decent BMS, is genuinely low. The failures that make headlines almost always trace back to a skipped clearance, a cheap BMS, or a chemistry choice made for cost rather than margin. Ask the right questions and most of that risk gets designed out before the unit is ever switched on.

Tagged

Common questions

Can an LFP home battery still catch fire?
Yes, though the risk is meaningfully lower than older nickel-based chemistries. LFP cells have a higher thermal runaway onset temperature and release less oxygen when they do vent, which tends to make failures slower and more containable, but it isn't immunity.
Does the BMS matter more than the battery chemistry?
They work together, but a poorly configured battery management system can undermine even a safe chemistry by allowing cells to sit near voltage or thermal limits too often. Tier-one brands generally run more conservative BMS thresholds than budget imports.
Do installation clearance rules actually reduce fire risk?
They change the outcome if a failure does occur, by keeping the battery away from egress paths, windows and habitable rooms. AS/NZS 5139 in Australia and NEC Article 706 plus UL 9540A testing in the US both exist for this reason, not as paperwork.
Does sizing a battery too small increase safety risk?
It can, indirectly. A pack that's cycled hard near its charge and discharge limits every day faces more chemical stress over time than one sized comfortably for typical daily throughput.
About the author
DO
Declan Osei
Battery & EV charging · Brisbane, AU

Declan writes about home batteries, EV chargers and the emerging vehicle-to-grid category, drawing on his own multi-year EV ownership.

EV owner since 2019; home-charging enthusiast.

More from Declan Osei