Reactor Atlas - Every nuclear reactor on Earth, mapped since 1954

by
301 nuclear sites and 779 reactor units worldwide on an interactive globe, with a timeline running from 1954 to today. Every unit has a technical record built from IAEA PRIS, Global Energy Monitor and Wikidata, cross-checked and corrected against primary sources. There's an animated wiki explaining how each reactor family works and a live signals engine that ingests public feeds to detect outages, power changes and regulatory events. Built by a nuclear engineer. Free, bilingual, no signup.

Add a comment

Replies

Best
Hi everyone, I'm Fede. I'm a nuclear engineer, and I built Reactor Atlas because the data about my own industry is a mess. There is no single place where you can see every nuclear reactor in the world with its real status. The IAEA publishes authoritative records inside PDFs. Global Energy Monitor publishes a spreadsheet. Wikipedia has a third version. They contradict each other constantly. While reconciling the main sources I found 31 units whose operating status was wrong in widely used datasets, including two reactors in my own country listed as suspended while they were actually running at around 68% capacity factor. So this is really a data reconciliation project with a map on top: 301 sites, 779 units, a timeline from 1954, a sourced technical record per unit, an animated wiki on how each reactor family works, and a layer of research reactors. The part I'm most excited about is newer. It's a signals engine that ingests public feeds (the NRC daily status report, seismic data, grid operators) and derives events by comparing each reading against the previous one. Nobody publishes "unit X tripped". They publish numbers. The event is in the difference. Free, bilingual, no signup, no ads. If you work in energy or data: what would make this genuinely useful to you, rather than just interesting to look at?
What became possible in your product with Astra that was not practical before?
Astra made it practical to iterate on Reactor Atlas at a much higher speed. It helped me resolve conflicts across multiple data sources, structure and normalize complex nuclear data, and rapidly design and redesign the UX based on feedback. It also made large refactors much more practical, allowing me to continuously simplify the codebase, improve performance, and ship new features without accumulating as much technical debt.