CATL Launches TENER Sodium Energy Storage System; Robin Zeng Proposes EV Computing Power Vision on Same Day
Miles Bennett
CATL unveiled TENER Sodium, the first field-validated sodium-ion grid storage system, with China deliveries starting September 2026 and global rollout in June 2027; chairman Zeng Yuqun followed the next day with a proposal to turn 40 million idle EVs into distributed AI computing infrastructure — two paths that signal a battery maker redefining its own boundaries.
What problem does sodium-ion storage actually solve?
Sodium's core edge: widely available, abundant reserves, tolerates extreme temperatures, long cycle life — no dependence on lithium supply chains, purpose-built for large-scale, long-duration grid storage.
CATL CTO Wu Dianfeng framed the positioning explicitly: sodium is not a lithium replacement but the second pillar of a "lithium-sodium parallel" strategy. This means → CATL is betting on two chemistries running side by side, not pitting sodium against its own lithium business.
In plain terms = lithium is the main engine; sodium is a second engine — running both widens the supply-chain resilience and the range of deployable scenarios.
What do TENER Sodium's specs actually mean?
A single system is rated above 30 MWh. Building a 1 GWh station requires only 32–34 units, with configurable duration from 1 to 8 hours. This means → deployment engineering is radically simplified — no need to stitch together hundreds of small modules.
The system shares platform architecture, footprint, and interface standards with CATL's existing lithium storage products — customers can switch chemistries without redesigning projects or re-certifying. In plain terms = for an existing CATL lithium-storage customer, swapping to sodium is like changing a cartridge — the machine stays the same.
China delivery starts September 2026, targeting 1 GWh shipped by year-end; global commercial delivery begins June 2027. This reflects a "validate domestically first, then roll out globally" cadence designed to capture first-mover scale in sodium-ion storage.
How does it compare to lithium on performance and safety?
Cycle life: over 15,000 cycles at 25 °C, over 10,000 at 45 °C, and above 92% capacity retention at −20 °C. This means → stable operation in extreme climates, suitable for grid deployments in the Middle East, Northern Europe, and other high-temperature-swing regions.
Safety benchmarks versus LFP — lithium iron phosphate, the dominant chemistry in grid storage today: cell expansion force down 40%, thermal-runaway surface temperature roughly 60 °C lower, gas generation reduced 35%, critical overcharge state of charge raised to 140%. In plain terms = when a cell fails, sodium's "blast intensity" is much lower than lithium's, giving safety systems more room to react.
On efficiency: a purpose-built Bi-DC bidirectional voltage control system — a technology that cuts conversion losses during charge and discharge — lifts round-trip efficiency by nearly 2 percentage points, auxiliary power consumption drops to 1%, and heat dissipation falls nearly 30%. For a 1 GWh station, these gains translate to millions of additional kilowatt-hours of output per year.
350-millisecond self-healing — what is a grid station's biggest fear?
The top operational pain point for large storage stations: one module faults, the entire station is affected. TENER Sodium deploys a millisecond-level self-healing system — 200 ms to detect and isolate the fault, another 150 ms to restore power in unaffected zones, full closed-loop recovery in 350 ms.
Faulty modules can be isolated and replaced independently without shutting down the rest. This means → lower maintenance costs, and the "never go dark" capability that grid operators care about most now has hardware-level backing.
This reflects a competitive shift in storage from "cell-spec benchmarking" to "system-level reliability" — whoever makes grid operators sleep easier wins the order.
Is Zeng Yuqun's "computing factory" vision realistic?
On June 23 at the Dalian World Economic Forum, Zeng proposed that China's 40 million-plus idle EVs and plug-in hybrids — equipped with batteries, AI chips, and computing power — could be converted into distributed AI computing infrastructure. In plain terms = your car sits parked most of the day; Zeng wants to network those "parked computers" into a virtual AI computing factory.
The concept echoes Elon Musk's 2024 proposal to link Tesla vehicles into a distributed server network, though Musk's version drew skepticism over unclear electricity-compensation and revenue-sharing mechanisms. A more mature adjacent path is vehicle-to-grid (V2G) — using EV fleets to feed power back to the grid during peak demand — which China is currently piloting.
This reflects strategic ambition that has moved well beyond "selling batteries" — CATL wants to become an energy-and-compute infrastructure company. But between concept and viable business model, questions around electricity compensation, data security, and owner incentives remain unanswered.
Two paths unveiled at once — what should investors watch?
CATL's revenue still comes primarily from EV battery sales, but storage-system demand is growing fast. Whether TENER Sodium hits its 1 GWh shipment target by the end of 2026 is the first hard benchmark for sodium-ion storage at production scale.
The EV-computing vision remains at the forum-speech stage — no product roadmap, no named partners, no business-model specifics. This means → in the near term it is aspiration, not an earnings driver.
The shared signal from both paths: CATL is transitioning from "battery manufacturer" to "energy-infrastructure platform." The market will watch sodium-storage delivery data first, then decide whether to price in the more distant computing vision.
Content is for reference only, not financial advice.