Qualifying a Pressure-Decay Test Procedure for a Small Vessel
A procedure for pressure-decay leak testing of a 2-liter aluminum vessel specifies pressurizing to 150 psig with dry nitrogen, stabilizing for 5 minutes, then measuring pressure drop over 10 minutes with a gauge having 0.1 psi resolution. The acceptance criterion is a pressure drop of no more than 0.5 psi. The vessel's specified maximum leak rate is 1×10⁻³ atm·cc/s. The procedure was written by a Level II and approved by a Level III. A Level III reviewing the procedure notes that the test temperature is not controlled and ambient temperature varies by ±5 °C during the test.
1. Evaluate whether the stated pressure-drop acceptance criterion is technically adequate for the specified leak rate, and identify the dominant uncertainty.
Read the worked answer
Convert the leak rate to an equivalent pressure drop. For a 2 L (2000 cc) vessel at ~10 atm absolute (150 psig ≈ 10.2 atm), 1×10⁻³ atm·cc/s corresponds to a molar leak rate; the pressure drop per unit time is ΔP/Δt = Q/V (in consistent units). Q = 1×10⁻³ atm·cc/s; V = 2000 cc, so ΔP/Δt = 5×10⁻⁷ atm/s ≈ 3.8×10⁻⁴ torr/s, or about 7.6×10⁻⁶ psi/s - over 600 s that is ~0.0046 psi, far below the 0.1 psi gauge resolution and the 0.5 psi criterion. The criterion is therefore not sensitive to the specified leak rate; it would pass gross leaks only. The dominant uncertainty, however, is temperature: a ±5 °C swing on a fixed-volume gas changes pressure by roughly ΔP/P ≈ ΔT/T ≈ 5/300 ≈ 1.7%, or ~0.017×150 ≈ 2.5 psi - larger than the 0.5 psi criterion. The test as written cannot distinguish a leak from thermal drift.
2. What corrections would you require before approving the procedure?
Read the worked answer
Require: (1) temperature stabilization and monitoring - the vessel and gas must be at thermal equilibrium, with temperature recorded during the test and drift limited so that thermal pressure change is small compared with the criterion; (2) a gauge or transducer with resolution and accuracy adequate for the intended leak rate (a 0.1 psi gauge is inadequate for 1×10⁻³ atm·cc/s; a differential transducer or mass-flow measurement is needed); (3) a defined stabilization time verified by a no-leak blank run; (4) a demonstrated correlation between pressure drop and leak rate, or replacement of the pressure-decay method with a more sensitive method (e.g., helium leak detection) if the specified rate must be verified. The Level III should not approve until these are addressed and the procedure is re-qualified.