Australia's First Grid NaS Battery Proves Long-Duration Storage Works, But NGK Is Shutting Down Production

Australia has officially commissioned its first grid-connected sodium-sulfur (NaS) battery into the National Electricity Market ^[1]. The 250 kW/1.45 MWh system, installed by Allset Energy at Lava Blue's PRiSM facility in Brisbane, sits alongside 200 kW of solar and is wired directly to Energex's distribution network ^[3]. It can supply up to 100 percent of the site's peak operating load, including energy-intensive mineral-processing equipment ^[3]. The milestone matters because sodium-sulfur chemistry, built on sodium metal and elemental sulfur, both abundant and geographically dispersed, offers what lithium-ion cannot: continuous discharge for six hours or more at rated capacity, without the supply-chain chokepoint or environmental extraction cost of lithium, nickel, and cobalt ^[7].

This is also a cautionary tale about market timing and technology betting. NGK Insulators, the Japanese manufacturer that dominates global NaS deployment (roughly 5 GWh installed worldwide over two decades of commercial operation ^[5]), announced it would stop taking orders and end production. The last system ships in January 2027 ^[4]. The reason: rising material costs and the relentless price competition from lithium-ion ^[4]. BASF, which partnered with NGK on NaS development, stepped back from the collaboration ^[4]. What looked like a proven, ready-to-scale chemistry for industrial and grid-scale storage has hit a commercialization wall not because the technology failed, but because the lithium industry scaled faster, prices fell harder, and the installed-base momentum favors the incumbent.

The strategic lesson runs deeper than one product line. Long-duration storage, four to twelve hours of continuous discharge, is not optional for grids running 60 percent or more wind and solar. Lithium-ion dominates the two-to-four-hour band because cost curves fell hardest there; the $/kWh-cycled for LFP has compressed to roughly $0.07, 0.10 at DIY scale and $0.15, 0.20 at installed rates ^[1]. But four to twelve hours demands a different chemistry. Iron-air, long-duration flow, compressed-air, and gravity systems are in the lab or early pilot phase. Zinc-bromine and vanadium redox flow are commercial but expensive per kWh (roughly $0.30, 0.50/kWh-cycled) ^[5]. NaS was supposed to be the bridge: cheaper than flow, thermally stable at scale, and proven over two decades. NGK's exit suggests the bridge was built too narrow, or built too late.

Australia did exactly what a public research partnership should: it deployed the proven option, monitored real industrial load shifting and solar integration through QUT's Queensland Energy Storage Technology (QUEST) Hub, and generated operational data for the next round of chemistry development ^[2], ^[5]. The PRiSM facility was chosen as a testing ground precisely because it offers variable processing loads, grid demand signals, and local solar generation to validate the system under real stress ^[3]. That is the public value, not the widget, but the evidence. The risk is that NGK's exit signals to other investors and manufacturers that long-duration chemistries live or die by how fast lithium-ion prices fall, not by their own merit or fitness for the duty cycle.

For household and community storage, the lesson is grimmer but clarifying: if you are betting on a non-lithium chemistry to ship within five years and solve your long-duration problem at scale, you are betting against the speed of incremental lithium refinement and against the installed base. Sodium-ion (distinct from sodium-sulfur) is in earlier commercial deployment and has cost advantages for very high cycle counts and thermal tolerance, but it trades discharge power density for cycle life ^[6]. A household or small business modelling resilience and self-consumption under export-hostile tariffs (the coming regime in California and elsewhere) cannot wait for NaS or sodium-ion to scale; they must size and price systems around what is available and certified today, LFP with six to ten hours of usable storage, priced by $/kWh-cycled, including balance-of-system and installer cost. The Lava Blue deployment validates that long-duration storage works and earns its keep in industrial duty; NGK's exit validates that the path to mass adoption runs through scaling what already prices, not through waiting for the next chemistry.