Report 012 · Energy Storage
Grid storage quietly standardized on a battery that stores less energy than its rivals, and it wasn't a compromise. It was the right call, and the reason goes all the way down to a single chemical bond. Here's the materials read on why lithium iron phosphate runs the grid.
Here's a fact that should be more surprising than it is. The battery chemistry that now dominates grid storage is not the one with the highest energy density, the most range, or the flashiest spec sheet. BloombergNEF reported that energy storage additions "reached 112 gigawatts in 2025, up 48% from 2024," and that "in 2025, LFP accounted for more than 90% of annual additions." Lithium iron phosphate, LFP, the humble, heavy, unglamorous chemistry, took nine of every ten new batteries going onto the grid. I help design the AI systems for an energy-storage integrator that builds with exactly this material, so let me explain why the boring battery won, because the reason is genuinely elegant.
Every lithium-ion battery has a cathode, and the cathode is where the safety story lives. In the nickel-rich chemistries built for maximum energy density, NMC and NCA, the cathode is a layered metal oxide. Those layers hold oxygen, and here's the problem: when the cell overheats, that structure can start to break down and let the oxygen go. A peer-reviewed review in the journal Materials describes the mechanism plainly: in these layered oxides "phase transition often occurs in the surface layer of the particles, accompanied by the release of oxygen atoms," and "the released oxygen may react with the electrolyte to produce a large amount of heat." That's the engine of a battery fire. A hot cell releases oxygen, the oxygen feeds combustion inside the cell, the heat releases more oxygen. It's a fire that brings its own air supply, which is exactly why you can't smother it.
LFP is built differently, and the difference is atomic. Its cathode isn't a layered oxide; it's an olivine-structured phosphate, and the oxygen in it is locked to a phosphorus atom by one of the more stubborn bonds in this kind of chemistry. The same review puts it precisely: "the P-O chemical bond is strongly stable, and thus the oxygen in the lattice is not easily lost." So LFP "does not show significant weight loss at temperatures below 230 °C and has better thermal stability, which is attributed to the fact that Fe-P-O bonds in olivine-structure LFP are much stronger than Ni-O, Co-O and Mn-O bonds." Read that as the whole safety case in one sentence. The reason an LFP battery is so much harder to send into thermal runaway is that its oxygen doesn't want to leave. Take away the internal oxygen supply and you take away the self-feeding fire. It's not that LFP is fireproof; it's that you have to work much harder to get it there, and even then it has far less to give.
(I covered the real-world payoff of this in an earlier report on why grid battery fires keep dropping. This is the atomic-level "why" underneath that trend.)
Safety would be enough, but LFP has a second advantage that matters just as much for the grid: longevity. The same structural stability that keeps the oxygen in place also means the crystal doesn't tear itself apart with use. Nickel-rich cathodes swell and contract and undergo phase changes as they cycle, and that mechanical stress slowly damages them. The rigid olivine framework of LFP largely doesn't, which is why the Materials review credits LFP with a "cycle life reaching more than 2000 times" while pairing it with "high safety," and why commercial LFP cells are routinely rated for many thousands of full cycles. On the grid, that's not a footnote. A stationary battery might cycle once or twice every single day for fifteen or twenty years. A chemistry whose thousandth cycle still looks like its first is a chemistry whose economics actually close. Longevity is the quiet half of why LFP is cheaper: not just less to build, but far more use out of what you built.
There's a supply-chain argument too, and it's not small. LFP's ingredients are iron and phosphate, both cheap and abundant. It uses no cobalt and no nickel, which spares it the two most expensive, most geographically concentrated, and most ethically fraught inputs in the battery world. Cobalt in particular carries a genuinely troubling mining supply chain. A grid-scale battery built on iron and phosphate sidesteps all of that. It's cheaper to source, less exposed to price shocks, and cleaner to stand behind. When you're buying batteries by the gigawatt-hour, those advantages compound fast.
Energy density, and only energy density. Because LFP stores less energy per kilogram, an LFP pack is bigger and heavier than a nickel-rich pack of the same capacity. In an electric car, where every kilogram costs range, that penalty is close to disqualifying, which is why premium EVs still lean on NMC. But on a concrete pad next to a substation, weight and floor space are cheap. Nobody carries the battery. So stationary storage is the one application where you can hand over the single thing LFP is bad at and keep everything it's good at. That's the whole trade, and once you see it, the 90% market share stops looking surprising and starts looking obvious.
Disclosure: I help design the AI battery-cycling systems for a veteran-owned (HUBZone) energy-storage integrator that builds with LFP. I don't own the company and earn nothing from this link; it's disclosed because this is a material I work with, not just write about. And it's worth the honest caveat: a safer chemistry is a higher floor, not a free pass. The controls still matter, watching each string, keeping cells inside their envelope, because the record shows most storage failures trace to how a system is run, not to the cell itself. LFP gives you a battery whose worst day is much less bad. Good engineering is what keeps you from having that day at all. Full policy here.
LFP didn't win the grid by being the most advanced battery. It won by being the right one, and there's a lesson in that worth carrying past batteries. The market didn't reward the highest number on the spec sheet; it rewarded the chemistry that was safest, longest-lived, and cheapest to source, because those are the virtues grid storage actually needs. The flashy metric, energy density, turned out to be the one thing the grid could most afford to give up. When you next see two technologies compared on a single headline number, ask what that number costs and what the real job rewards. Sometimes the winning move is the boring material with the stubborn bond that simply refuses to let go.
U.S. Army combat veteran (Counter-IED / Electronic Warfare), peer-reviewed researcher in microwave spectroscopy, and founder & CEO of Shroombiosis. Consults on laboratory operations, AI, and supplement formulation.