Fill plan
Ideal-gas estimateCatalina S80
No bleed-down needed
Pressure checkpoints
Stop at, gauge PSIRead addition pressures after the gas settles at 80°F.
Keep the existing gas
500 PSI21% oxygen, measured at 80°F.
Add 36% nitrox to
2,711 PSIAdd 2,211 PSI equivalent at 80°F.
Top with 21% air to
3,000 PSIAdd 289 PSI equivalent at 80°F.
Allow for cooling
10 min continuous fill. Uncalibrated heat-transfer estimate.
Estimated peak pressure: 3,218 PSI, above the 3,000 PSI rating. Do not fill to this pressure. Allow cooling and follow approved top-up procedures.
Verify this ideal-gas estimate and analyze the finished mix.
Calculations and limits
Oxygen balance
The calculation accounts for oxygen in the remaining gas and both fill gases. It retains the most existing gas that allows you to reach the target mix and pressure.
Cylinder volume cancels out for a single rigid tank. Check the selected pressure preset against your cylinder’s stamped rating.
Have a qualified gas-blending professional review this prototype before using it for a real fill. Verify the cylinder’s rating and equipment compatibility. Stay within rated pressure during filling, including when the gas is hot. Follow approved filling procedures and analyze the finished mix.
Fill temperature
The estimate balances energy carried in by the gas with heat absorbed by the cylinder and lost to its surroundings. It uses fill duration, inlet temperature, water capacity, cylinder mass and material, and editable heat-transfer coefficients.
It assumes a continuous fill at constant molar flow, with the gas and wall starting at the current gas temperature after any bleed-down. Cooling pauses between gases are not simulated. Checkpoints use settled readings unless cooling correction is enabled with entered gas temperatures for each addition stage. Bleed readings always use the starting temperature.
Heat-transfer values are uncalibrated assumptions. The model uses ideal diatomic gas properties and excludes real-gas compressibility, source-bank limits, and temperature gradients. Neither a model estimate nor a cylinder-surface reading establishes the actual bulk gas temperature.
If the gas were at 119°F when the gauge reached 3,000 PSI, the fixed-quantity ideal-gas correction gives 2,797 PSI after cooling to 80°F. This does not simulate stopping early at the rating.
For fixed gas quantity, pressure correction uses PSI + 14.7 and °F + 459.67. Zero gauge pressure still contains atmospheric gas.
Model reference: pressure-vessel energy balances. The equations apply here; hydrogen-specific coefficients from that study are not used.