H₂

Acceptance testing of 700 bar dispensers to EN 17127

Delivered as engineer at a national H₂ network operator.

RESULT

Evaluation went from hours of manual chart work to minutes per test, in one defensible format.

Multi-axis acceptance test chart of a 70 MPa fill with pressure corridor, temperatures, flow and state of charge — FIG. 01 — 70 MPA FILL: PRESSURE CORRIDOR, MAT30, SOC, FLOW — ONE TEST, 20+ SIGNALS

THE CHALLENGE

Before a station goes live, its dispensers must prove in FAT/SAT tests that every fill stays inside the EN 17127 / SAE J2601 pressure and temperature corridors. One test produces 20+ simultaneous sensor traces, and the evidence must hold up before manufacturer, operator and notified body at once.

Evaluating those traces by hand, test after test, was slow, error-prone and inconsistent: exactly what acceptance disputes feed on.

WHAT I BUILT

I built an automated analysis pipeline in Python: it ingests the raw test logs, reconstructs the target pressure ramp (APRR), overlays corridors, MAT30, state of charge and flow, and renders a standardized chart plus pass/fail per criterion.

The same tooling derived PLC limits from thermodynamic boundary conditions and pinpointed which subsystem (precooling, compressor staging, communication) caused a corridor violation.

WHAT CHANGED

Violations became diagnosable: the chart shows which subsystem pushed the fill out of bounds.

Dozens of acceptance tests supported across multiple stations and dispenser generations.

Open dispenser cabinet with measurement laptops during a site acceptance test — FIG. 02 — SITE ACCEPTANCE TEST IN PROGRESS