China has likely resumed making weapon-grade uranium at Emeishan

Bulletin of the Atomic Scientists | Hui Zhang

China is expanding the Emeishan uranium enrichment complex in Sichuan Province, completing a fourth centrifuge plant that will increase site capacity to 5.4 million separative work units annually by late 2026. This rapid scaling indicates Beijing has likely resumed producing weapon-grade highly enriched uranium for its accelerating nuclear warhead expansion, ending a nearly four-decade hiatus.

While commercial imports and output from Lanzhou and Hanzhong comfortably meet domestic civilian power reactor requirements, Emeishan operates outside National Nuclear Safety Administration oversight and International Atomic Energy Agency safeguards. China's legacy stockpile cannot support this growth. Reaching the US Department of Defense projection of over 1,000 warheads by 2030 requires approximately seven additional metric tons of highly enriched uranium for thermonuclear secondaries. At full projected capacity, Emeishan can generate 25 metric tons of weapon-grade material annually, forcing a strategic trade-off where nuclear enrichment infrastructure prioritises rapid military breakout capability over transparent civilian fuel cycle integration.

Comment

The rapid commissioning of CEP4 at the Emeishan site reveals how China National Nuclear Corporation leverages modular centrifuge architecture to obscure military fissile production within a civilian enrichment network. By standardising 1.6 million SWU plant modules across Lanzhou, Hanzhong, and Emeishan, Beijing creates an elastic industrial buffer capable of pivoting between low-enriched reactor fuel and weapon-grade HEU without visible structural redesigns. This dual-capable infrastructure forces a strategic trade-off, trading international oversight and civilian fuel transparency for unannounced rapid breakout potential. Plant 814's unannounced expansion signals a transition from fixed fissile stockpiles to just-in-time industrial capacity for warhead manufacturing.

This operational flexibility depends directly on second-generation domestic centrifuges, which yield a 40 percent efficiency gain over earlier cascades and reduce the physical footprint required per separative work unit. Consequently, CEP4 can achieve high-enrichment cascades within compact hall footprints that remain indistinguishable from commercial facilities under satellite observation. The integration of these advanced cascades allows the PLA Rocket Force to bypass traditional fissile material bottlenecks without expanding dedicated, single-purpose military enrichment sites.

Strategic Question for Discussion

If Plant 814 standardises dual-use centrifuge cascades across Emeishan, how can external intelligence agencies reliably distinguish between routine commercial LEU production and covert HEU breakout?

The available evidence points toward significant monitoring friction, as commercial satellite imagery alone cannot verify cascade interconnections or isotope enrichment levels inside active centrifuge halls. My assessment is that intelligence tracking will increasingly rely on upstream precursor movement, specialized valve procurement, and power consumption anomalies rather than structural site signatures. However, without direct IAEA inspections or environmental sampling at Emeishan, identifying the exact moment of weapon-grade HEU diversion remains exceptionally difficult.

Share your assessment in the comments below.

Read the original article