Hydrostatic and pneumatic test pressure

A pressure test proves the integrity of a completed pressure boundary by taking it above its working pressure under controlled conditions. The margin above MAWP is deliberate: it demonstrates that the joint efficiency assumed in design is real, exposes gross workmanship defects and leak paths, and produces a small amount of local yielding at stress concentrations that redistributes residual stress.

The stress ratio is the part most often missed. Codes require the test factor to be multiplied by the ratio of the allowable stress at test temperature to the allowable stress at design temperature. A vessel designed for 400 °C has a much lower allowable stress at temperature than at the ambient temperature it is tested at, so a bare 1.3 × MAWP test would not develop the intended fraction of the design margin. ASME VIII-1 UG-99(b) requires 1.3 × MAWP × (S_test/S_design) using the lowest ratio for any material in the vessel. ASME B31.3 §345.4.2 uses 1.5 × design pressure with the same ratio, capped at 6.5.

Pneumatic testing is different and more dangerous. Water is nearly incompressible, so a hydrostatic failure releases very little stored energy — the pressure collapses as soon as the boundary opens. A gas-filled vessel stores enormous energy and fails explosively. Because of that, codes set a lower factor for pneumatic tests (1.1 in ASME VIII-1 UG-100 and B31.3 §345.5) and require additional precautions: exclusion zones, remote monitoring, a stepped pressurisation with holds, and a documented justification for why a hydrostatic test cannot be used.

The test pressure must also be checked against yield. A test that puts the shell close to its yield strength risks gross distortion, particularly at nozzles and other stress raisers, and B31.3 permits the test pressure to be reduced where it would produce a stress in excess of yield at test temperature. Finally, remember that the pressure at the bottom of a tall vertical vessel is the gauge test pressure plus the static head of the test water — around 0.098 bar per metre — and it is the bottom course, not the gauge, that sees the highest stress.

Stress ratio = S at test temperature / S at design temperature
ASME VIII-1 hydrostatic (UG-99b):  P_T = 1.3 × MAWP × ratio
ASME VIII-1 pneumatic  (UG-100):   P_T = 1.1 × MAWP × ratio
ASME VIII-2 hydrostatic (8.2.1):   P_T = greater of 1.43 × MAWP and 1.25 × MAWP × ratio
ASME VIII-2 pneumatic  (8.3.1):    P_T = 1.15 × MAWP × ratio
ASME B31.3 hydrostatic (345.4.2):  P_T = 1.5 × P_design × ratio, ratio ≤ 6.5
ASME B31.3 pneumatic   (345.5.4):  P_T = 1.1 × P_design
Hoop stress at test:  σ = P_T·(R + 0.6t) / (t·E)