Report 030 · Energy Storage
Most of the work in lithium-battery safety goes into stopping a fire once it starts: better sensors, spacing, suppression, smarter controls. A zinc-bromide battery takes a different bet. Its electrolyte is mostly water, so there's no fire to stop. In June 2026 an independent lab tried to make one burn and couldn't. That's a real advantage. It also isn't free.
Every battery is a controlled way to store a lot of energy in a small space, and the fear with lithium-ion is that the "controlled" part fails. When a lithium cell goes into thermal runaway, its own chemistry keeps feeding the fire, the flammable electrolyte ignites, and heat drives the neighboring cells to do the same. The whole grid-storage safety industry, standards like NFPA 855, module spacing, gas detection, deluge systems, exists to interrupt that chain. And the trend line makes it harder, not easier: the industry keeps building bigger cells to cut cost, and a bigger cell is simply more stored energy waiting in one place if it ever lets go.
There's another way to approach the problem, and it's almost embarrassingly direct: build the battery out of something that doesn't burn. That's the pitch behind zinc-bromide chemistry, and in 2026 it stopped being a lab curiosity and started raising real money. It's worth understanding what that buys, because the honest answer has two sides.
A conventional lithium cell uses an organic, flammable solvent as its electrolyte. A zinc-bromide battery uses an aqueous one, water with salts dissolved in it. It stores energy by plating solid zinc metal onto an electrode as it charges and dissolving it back as it discharges, with a bromide species doing the matching chemistry on the other side. The reason this matters for safety is almost too simple to trust: water does not burn. There is no flammable solvent to ignite, so the self-sustaining fire that defines a lithium thermal runaway has nothing to run on.
That claim is easy to make and, until recently, easy to wave off as marketing. So the useful development in June 2026 was an independent one. The Energy Safety Response Group, an outside testing firm, ran large-scale fire-test-equivalent abuse on modules of one commercial zinc-based system (Eos Energy's Z3) under two of the nastiest conditions you can impose: direct flame held against a module, and forced overcharge. The reported result was no thermal runaway, no sustained fire, and no propagation to the neighboring live modules, with no ignition of the vented off-gas. That's the difference between "we believe it's safe" and "someone who doesn't work here tried to set it on fire and failed." The distinction matters because it's exactly the kind of testing the 2026 NFPA 855 rules are pushing installers to demand.
Here's where a materials person has to be even-handed, because intrinsic safety is a genuine advantage and it is not a free one. Zinc-bromide belongs to the same family as the other stationary chemistries I've written about: it wins on safety, cost, and durability, and it loses on energy density. It is big and heavy for the energy it holds. Nobody is putting one in a phone or a car; like iron-air and vanadium flow, it's a chemistry for a fixed pad next to a substation, where weight is free and the job is holding a lot of energy for a long time, safely.
The chemistry also carries real, well-documented headaches, and I'd rather you hear them from the peer-reviewed literature than from a spec sheet. Zinc batteries are prone to dendrites, needle-like zinc growths from repeated plating and stripping that can bridge the electrodes and short the cell. The bromine side wants to drift across the cell and self-discharge, and elemental bromine is corrosive and needs careful handling. Side reactions like hydrogen evolution nibble at efficiency. A 2023 review of zinc-bromine batteries put practical energy efficiencies in the range of roughly 75 to 82 percent in the studies it surveyed, noticeably below lithium's high-eighties-to-ninety. The manufacturer of the system tested above says its particular design, a static bath rather than a pumped flow cell, with additives tuned to suppress dendrites, pushes round-trip efficiency up toward 90 percent and holds more than 97 percent of capacity over a 25-year life. Those are the maker's numbers, on the maker's design, and they're a claim to watch as fleets rack up real years, not a settled fact. The general point stands: you are trading some efficiency and a lot of energy density to buy a battery that cannot burn.
The reason this is worth a report now, and not two years ago, is that the financing caught up to the chemistry. In May 2026, Eos and the investment firm Cerberus announced Frontier Power USA, a venture to build and operate long-duration storage projects on this zinc-bromide technology. The structure is the interesting part: a 2 GWh manufacturing-capacity reservation, a $100 million equity commitment, and, tellingly, a 15-year technology-performance insurance policy with up to roughly $1.5 billion in capacity, written to reassure the lenders financing the projects. When an insurer is willing to underwrite fifteen years of a battery's performance, that's a market signal about perceived risk that no press release can fake. It doesn't prove the chemistry wins. It proves serious money now treats it as bankable.
Intrinsic safety is not the same as "no management required," and that's the part that keeps this interesting to me. A battery that can't run away still has to be cycled well to be worth its 25-year promise: how deeply you discharge it, how you manage the plating and the bromine over thousands of cycles, when you rest it, all decide whether it actually reaches year twenty or quietly dies at year eight. The chemistry sets the safety floor; the control system sets the economics. I help design the AI battery-cycling systems for a veteran-owned (HUBZone) energy-storage integrator, the layer that decides how a pack charges, discharges, and gets watched for drift, so a chemistry that removes the fire risk but still lives or dies on operating discipline is squarely the kind of problem I find worth solving. (Disclosure: I help design that company's AI; I don't own it and earn nothing from this link. Full policy here.)
There are two roads to a safe grid battery. One is to take the energy-dense chemistry that can burn and wrap it in enough engineering, spacing, sensing, suppression, and control, that it rarely does; that road is real, and it's why grid-battery fires keep getting rarer even as we build more. The other is to pick a chemistry that can't burn in the first place and accept that it'll be bigger and a little less efficient. Neither is "the answer." They're answers to different questions. So when a battery advertises that it can't catch fire, the honest follow-ups are the same ones I ask of every storage pitch: safe under whose test, and what did you give up to get there? For zinc-bromide the answers are, respectively, an independent one, and energy density. That's a trade a lot of stationary projects should be glad to make. It's just not a free lunch, and anyone who tells you it is hasn't read the second half of the datasheet.
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.