Project
CHASM: An Evidence-Graded Atlas of Carbon Hydrogen-Storage Claims
A fully specified, pre-registered synthesis that grades the field's storage claims and bounds defensible capacities, using my own samples as the case study.
Motivation
CHASM is the capstone the other independent projects exist to feed: an evidence-graded synthesis of the field’s carbon hydrogen-storage claims, connecting spectroscopic evidence to defensible capacity bounds, with my own experimental samples as the case study. Given the field’s history of unreproduced capacity claims, grading is the right operation rather than aggregating. The question is which published numbers survive scrutiny and what bounds the trustworthy subset actually defends.
Design
The project is fully specified in an execution manual written before any execution: eight modules, pre-registered questions, an evidence-grading rubric, a claim translator, and a fixed evidence window, with hard gates and scheduled decision points. Two modules consume the other projects as pinned, pip-installable dependencies at tagged commits. The atlas module ingests HyCAN-DB’s schema, validators, and converters and adds 35 to 60 freshly extracted records under CHASM’s own extraction protocol. The system-projection module imports H2STAR’s validated equation-of-state and isotherm layers to translate graded material claims into system-level capacity, with a disclosed parametric fallback (system-level derating factors of 0.4 to 0.7) if the tank layer misses its gate. The design deliberately converts both upstream projects into bounded deliverables with deadlines rather than open-ended efforts.
Current status
Specification complete, execution deliberately deferred. In July 2026 I decided to finish the near-term HyCAN-DB and H2STAR milestones before starting CHASM’s clock, which is why this project is listed as paused rather than active: the manual, gates, and schedule exist, and day one has intentionally been held.
Future work
Execution opens with the pre-registration, anchor, rubric, translator, and evidence-window days, which touch neither dependency, then proceeds to the atlas and the system projection as the upstream gates go green. The deliverables are the graded atlas and a capacity-bounds report tying the literature, my own samples, and DOE system targets into one defensible document.