Home battery chemistry: LFP vs NMC

What we set out to do, and what happened

The question everyone asks about home batteries is lithium iron phosphate against nickel manganese cobalt: which is safer, which lasts longer, which packs more energy into less space. Every buyer's guide answers it confidently.

We went to the manufacturers' own current data sheets to answer it properly, and the exercise failed in an informative way.

Two manufacturers state their chemistry plainly. Both identify their products as lithium iron phosphate.

One of the largest-selling home batteries does not state its chemistry on its data sheet at all. We looked for a chemistry row. There is not one. That is a verified absence rather than an oversight in our reading, and it means a buyer comparing that product against a stated-LFP competitor is comparing a known quantity against an unstated one.

And none of the documents publishes what a real comparison would need. Cell-level energy density, thermal runaway onset temperature, cycle life at defined depth of discharge: none of it is in the public data sheets. Those figures exist in the industry, and they are not in the documents a homeowner can read.

So we are not going to give you the comparison. We could compute system-level energy per kilogram from the published dimensions and weights, and we have, and it would be misleading to publish it as a chemistry comparison, because those enclosures contain different inverters alongside the cells. That number measures packaging, not chemistry.

The safety claims are worded differently, and the difference is real

The standard everyone cites for battery fire safety is a test method for thermal runaway fire propagation. It is worth understanding what it is before comparing claims made under it.

It is a test method, not a pass or fail certification. It produces data at several levels of assembly, and that data is then used by others, such as code officials, to decide what installation conditions are acceptable. There is no "passing" it in the way there is passing a safety listing.

Now compare two manufacturers' claims, quoted as they appear.

One says its product "Meets the unit level performance criteria" of the standard.

The other says its product is "Evaluated to [the standard] for thermal runaway fire propagation and reduced separation distance as required in" three named code provisions, which it lists by section.

Those are not the same statement. The first asserts performance against criteria at one level of the test. The second says the product was evaluated and ties the result to specific code requirements it is intended to satisfy, including reduced separation distance, which is the practical thing a code official cares about when deciding how far a battery must sit from other things.

We are not saying one product is safer. We have no basis for that and neither does anyone reading data sheets. We are saying that "meets the criteria of" and "evaluated to, for these code provisions" carry different information, and that a comparison treating them as equivalent has flattened a real difference. If a battery's placement in your house depends on separation distance, the second form of claim is the one that speaks to it.

The comparison that does work: warranties

Where the chemistry comparison fails, the warranty comparison is straightforward, published, and differs more than anything else we looked at.

Three dimensions matter and most people only look at the first. The term in years. The throughput or cycle guarantee, which caps how much you may actually use it. And the end-of-term capacity retention, which is what the manufacturer promises the thing will still hold.

One product offers fifteen years, six thousand cycles, and greater than 60% retention. Another line offers twelve years against 43 megawatt-hours of throughput, with a newer model at fifteen years and 60 megawatt-hours.

And one of the best-known products offers ten years, with no cycle figure, no throughput figure and no retention figure published on the documents we could obtain.

That last one deserves a second look, because of an internal comparison within the same manufacturer's own range. Its battery carries ten years. Its rapid shutdown device, a far simpler piece of equipment, carries twenty-five. We draw no conclusion from that beyond the obvious one: warranty length is a commercial decision about expected claims, and a manufacturer offering ten years on one product and twenty-five on another has made two different judgements.

One further asymmetry is worth noticing about how these warranties are written. A term expressed in years is a promise the manufacturer controls least, since it depends on how hard you use the product. A term expressed in throughput is a promise about energy, which the manufacturer can price precisely. Where a manufacturer publishes both, it has bounded its exposure on two axes and told you exactly what it is offering. Where it publishes only years, it has either taken a broader risk or left itself room to argue about what fair use looked like, and a buyer cannot tell which from the outside.

The throughput figure is the one to insist on if it is missing. A fifteen-year term with a six thousand cycle limit is a different product from a fifteen-year term with no limit, and if you intend to cycle a battery daily for arbitrage rather than holding it for occasional backup, the cycle or throughput cap may bind long before the years do.

Where you are allowed to put it, which differs by manufacturer

Permitted installation location is the dimension buyers discover last and it can determine whether a product works for your house at all.

One manufacturer's installation guide carries a boxed warning banning installation in habitable space. It specifies garages and sheds, or locations more than a stated distance from dwelling units, and sets out clearances: a margin around the unit, a larger clearance in front, a minimum gap between multiple units, a height requirement where the installation is in a vehicle path, and a structural requirement for blocked studs given the unit's weight.

Other manufacturers' data sheets say only that the product is indoor and outdoor rated.

Silence is not permission. A data sheet that does not restrict indoor installation is not a document authorising it, because the governing restrictions may live in an installation manual we could not obtain, in the fire code your jurisdiction has adopted, or in what a local official will approve. We could not obtain the installation manuals for two of the products, and we say so rather than reading their brevity as latitude.

The practical consequence is that this is a question for your installer and your building department, not for a comparison table. Ask where the specific product may be placed in your specific house, and ask which document says so.