Cold and Hot Testing / Calibration of PSVs

This is an excerpt from ASME Section VIII, Division 1, 1992 Edition, Pressure Relief Devices, Section UG-136, Subsection (d):

(4) Each valve shall be tested to demonstrate its popping or set pressure. Valves marked for steam service or having special internal parts for steam service shall be tested with steam, except that valves beyond the capability of the production steam test facility either because of size or set pressure may be tested on air. Necessary corrections for differentials in popping pressure between steam and air shall be established by the manufacturer and applied to the popping point on air. Valves marked for gas or vapor may be tested with air. Valves marked for liquid service shall be tested with water or other suitable liquid. Test fixtures and test drums where applicable shall be of adequate size and capacity to ensure that valve action is consistent with the stamped set pressure within the tolerances required by UG-134(d).

(5) A seat tightness test shall be conducted at a maximum expected operating pressure, but at a pressure not exceeding the reseating pressure of the valve. When testing with either water or steam, a valve exhibiting no visible signs of leakage shall be considered adequately tight. Leakage tests conducted with air shall be in accordance with industry accepted standards.

(6) Testing time on steam valves shall be sufficient, depending on size and design, to insure that test results are repeatable and representative of field performance.

I have performed quality control surveillance, verification, and inspection activities for set point testing in nuclear power plants where this function is taken very seriously. I don't believe there is a test method that allows the use of noncompressible fluid to certify set pressure for safety valve. I'd check with the manufacturer for assistance.