Research
We publish what we learn so that projects, and especially countries building their first plant, do not repeat expensive mistakes.
01 / Multi-case study
Seven projects, including Olkiluoto 3, Flamanville 3, Hanhikivi 1 and V.C. Summer, together with Korea's 2012–2013 certificate scandal, and what their documentation failures have in common. It ends with a red-flag checklist for your own licence file.
Publishing soon
Research paper
A physics-informed framework for stakeholder dynamics and supply-chain resilience in the nuclear ecosystem. It sorts project information into layers from public to classified, scores how evenly information is spread with an entropy measure, and models stakeholders as a multiplex network of material, financial, regulatory, information and political links.
Preprint in preparation
02
Published on LinkedIn, in English.
03
Our reproducible environment for nuclear simulation. A source scanner detects how each code builds, a recipe generator writes repeatable build recipes, and a metadata tracker records exactly what was built and with which inputs. It supports open solvers such as OpenMC, PETSc, SLEPc and Trilinos. Licensed codes such as MCNP and SCALE run only under the user's own licence and are never redistributed.
Automated execution, post-processing and uncertainty reporting for Monte Carlo runs.
Best-estimate-plus-uncertainty methods as a technical basis for risk-informed licensing decisions.
04
Countries starting a nuclear programme face a regulatory gap, a standardisation paradox and the need to build trust from scratch. We are preparing open guidance and training material on these challenges.
The companion code for our research and parts of Djed Lab will be released as open source, free to use. Anyone who finds it useful will be able to sponsor the work.
How sponsorship works
Send a short description. We reply within two working days and sign your NDA before any document changes hands.
