New fusion-program review puts award design beside engineering progress
A September 23 paper examines the mechanics of federal fusion partnerships, while recent company milestones show what a design review does and does not establish.
News analysis | Sources reviewed September 28, 2026.
A paper posted September 23, 2026 offers a closer look at the federal program supporting private fusion developers, including a problem less visible than a new magnet or reactor building: the time and negotiation required to make a public-private agreement work.
The retrospective by Scott C. Hsu, Samuel E. Wurzel and Narayan S. Subramanian examines the Department of Energy’s Milestone-Based Fusion Development Program. The authors describe work performed while they were at DOE, but say the paper represents their views, not an official government position. It was submitted for publication; the arXiv posting should not be treated as a completed peer-review process.
The authors identify award negotiations, intellectual-property terms and agreements with national laboratories among the implementation issues. Their account describes a program intended to support engineering progress while fitting the circumstances of privately financed companies.
What a milestone actually establishes
DOE’s program description identifies viable pilot-plant designs and technology roadmaps as near-term objectives. Technology Investment Agreements provide tailored terms, and progress reviews help determine movement into subsequent phases. That is support for development, not a declaration that commercial fusion electricity has been delivered.
A recent example comes from Commonwealth Fusion Systems. In an August 5 statement, the company reported completing milestones for an ARC preconceptual design and technology roadmap. It said an independent panel assembled by DOE reviewed the work.
CFS distinguishes ARC, its planned power plant, from SPARC, its demonstration machine. The August milestones concerned design and a development path. They were not an operating license, a completed plant or evidence of electricity supplied to customers.
The project-delivery implication
The new paper and the company example illuminate different sides of the same process. Technical progress needs a defined test. Public support needs an agreement specifying what evidence earns a payment and what rights or obligations follow.
For infrastructure readers, the practical discipline is to identify the object of each announcement. Has a team completed a design review, demonstrated a component, secured a site approval or operated an integrated system? Each can be meaningful without answering all the remaining questions.
A preconceptual review can expose inconsistent assumptions before detailed engineering proceeds. It cannot establish the future availability, cost or maintenance performance of a plant that has not yet operated. Those require later evidence.
The September paper adds an administrative perspective to the fusion-development story. Its lessons concern how agreements can support or delay progress. They do not remove the engineering work still required to turn a promising concept into dependable power infrastructure.
The photograph shows the historical NSTX research device at Princeton. It does not depict SPARC, ARC or a commercial power plant.
National Spherical Torus Experiment at Princeton Plasma Physics Laboratory; historical research photograph. Photo: U.S. Department of Energy / Public domain (U.S. government work). No editorial alterations; WordPress may resize or crop for display.
.jpg){kind=link}