The generic name for the reactor that comes after ITER — every major fusion nation is designing one, and none has started building yet.
DEMO is not a single machine. It is a concept class — the label the international fusion community uses for the first generation of demonstration fusion power plants intended to follow ITER and prove that fusion can deliver electricity to a grid. The term has been in circulation since at least the 1990s, but the urgency around DEMO design has intensified sharply since the 2020s as ITER construction advanced and private fusion companies began setting their own timelines.
ITER is designed to demonstrate sustained burning-plasma conditions and achieve a fusion energy gain factor Q = 10, but it will not generate electricity. Its thermal output will be radiated away, not converted. DEMO is the bridge between that physics demonstration and a commercial power plant. A credible DEMO must solve problems ITER was never asked to address: continuous tritium self-sufficiency, high-duty-cycle operation, materials that survive years of 14.1 MeV neutron bombardment, and a balance-of-plant that converts fusion heat into saleable power at an acceptable cost.[1]
The most mature DEMO design effort sits within EUROfusion, the European Consortium for the Development of Fusion Energy. EU-DEMO targets 300–500 MW of net electrical output from approximately 2 GW of fusion power, using a conventional-aspect-ratio tokamak with superconducting magnets. The current pre-conceptual design assumes a major radius around 9 m, a toroidal field on axis of roughly 5.5–6 T, and a pulsed operating scenario with burn durations of two hours or more. EUROfusion’s roadmap targets a DEMO construction decision around 2040, with first operations in the late 2040s or early 2050s.[2]
Europe is not alone. China has CFETR, Japan has explored a Slim-CS-based DEMO concept, South Korea is pursuing K-DEMO, and India has outlined a DEMO-relevant roadmap linked to its SST-2 program. The United States has historically underinvested in DEMO planning relative to its ITER obligations, though the 2022 Bold Decadal Vision report and subsequent DOE community planning exercises have called for an American fusion pilot plant by the 2030s or 2040s — effectively an accelerated DEMO equivalent driven partly by competition from the private sector.[3]
The technical challenges facing any DEMO are formidable and largely unsolved. Tritium breeding blankets must achieve a tritium breeding ratio (TBR) above 1.0 — and realistically above 1.05–1.15 to account for losses — but no integrated blanket has been tested under reactor-relevant neutron fluence. Structural materials such as EUROFER-97 reduced-activation ferritic-martensitic steel must withstand 50–150 displacements per atom (dpa) of neutron damage. The divertor must handle steady-state heat fluxes exceeding 10 MW/m². Remote maintenance systems must disassemble and replace activated components on timescales compatible with high plant availability.[4]
Perhaps the deepest uncertainty is economic. A DEMO that costs more per kilowatt-hour than renewables-plus-storage would struggle to justify the decades of R&D behind it. Reconciling physics performance, engineering feasibility, and commercial competitiveness within a single design is the defining challenge of post-ITER fusion.[5]