Skip to content

Refueling simulator

What happens between nozzle click and full tank? Watch a 700 bar fill run by the real dispenser control law, the SAE J2601 MC formula — pressure ramp, tank heating, state of charge, and what happens when precooling can't keep up.

  1. CONNECT
  2. ·2 PULSE / LEAK CHECK
  3. ·3 MAIN FILL⏸ BANK SWITCH
  4. ·4 TOP-OFF
  5. ·5 COMPLETE

0 s

0.00 kg

20 °C

CONNECT · 0.0 g/s

70 MPA FILL
STATION PRESSURETANK PRESSURETANK GAS TEMPFUEL DELIVERY TEMP

SAE J2601 (MAY2020) LIGHT-DUTY MC FORMULA · SELF-DERIVED COEFFICIENTS · VALIDATED AGAINST A REAL REFUELING (ANONYMIZED)

Advanced — vessel, communication, station

TABLE 8 CATEGORIES A–D · 49.7–248.6 L

WITHOUT IR DATA THE END TARGET IS LOWER (EQ. J83)

LOG-TRUE: THE REAL FILL SHOWS THREE (§9.3.2.1)

SLOW = OUT OF SPEC AT T30/T20 (EQ. J93 GATE)

COLD-CASE BOUND −40 °CHOT-CASE BOUND −33 °C

Table 7 window of the T40 category. The self-derived coefficients assume App. A's hot case — a compliant station delivers at this temperature within 30 s of mass flow.

Simulation of the SAE J2601 (MAY2020) light-duty MC Formula with self-derived tfinal coefficients (App. H method) — validated against a real refueling (anonymized), but not a dispenser control system and no substitute for the standard.

How it works — physics, data sources, scope

The dispenser side runs the SAE J2601 (MAY2020) light-duty MC Formula — the control law real 700 bar dispensers use. Every second it updates the mass-average fuel temperature (MATC), looks up the allowed fill time, and derives the pressure ramp rate, the corridor the station must stay inside, and the moving end-of-fill pressure target. The vehicle side is a lumped-parameter tank model on NIST real-gas data, giving the temperature rise and state of charge.

The standard's fill-time coefficient tables are licensed content and do not ship. The coefficients here are self-derived with the standard's own Appendix H method: constant-ramp worst-case fills of the boundary vessels, pushed to the 85 °C / 97 % SOC limit. The detail view shows the live corridor, MATC and ramp-rate values the control law is computing.

The model is validated against a real 700 bar refueling (anonymized): fill time, dispensed mass, temperature peak, bank-switch pauses and end-of-fill target all reproduce the logged fill within a few percent.

The failure modes are the point: a warm day with mild precooling is not an abort — the protocol trades speed for safety and the fill slows to a crawl. A precooler that misses its temperature gate is an abort, exactly where the standard prescribes. Both are one click away above.

How long does refueling a hydrogen car take?

Under normal conditions — 700 bar system, −40 °C precooling, 20 °C ambient — this simulation completes the fill in 3 min 34 s, consistent with the few minutes real SAE J2601 dispensers need. Warm weather with weak precooling does not abort the fill; the protocol trades speed for safety and the fill slows down instead.

Why does a hydrogen tank heat up during refueling?

The incoming gas carries compression work as enthalpy, so a fast fill heats the tank toward the 85 °C material limit. That limit, together with the 97 % state-of-charge cap, is exactly what the SAE J2601 MC formula manages — it is why dispensers precool the gas to −40 °C and control the pressure ramp rate.

What is the SAE J2601 MC formula?

It is the control law real 700 bar dispensers run: every second it updates the mass-average fuel temperature, looks up the allowed fill time, and derives from it the pressure ramp rate, the corridor the station must stay inside, and the end-of-fill pressure target. This simulator implements the MAY2020 light-duty version with self-derived coefficients, since the standard’s own tables are licensed content.

A data or systems problem in hydrogen or e-mobility?

Let's discuss it in a free 30-minute call — no obligation. In English, German or Dutch.