Before the industry slowdown of 2008, some end users performed third-party pressure testing of valves for use in critical service according to the acceptance criteria of ASME B16.34, API-598, API-6D, etc.
However, since that time, requests for independent valve pressure testing have increased greatly. This increase has been noted on non-destructive evaluations (NDE) such as radiographic testing (RT), magnetic particle inspection, liquid penetrant inspection (Figure 1) and visual inspection of valve body castings to meet customer specific requirements or industry standards such as ASME B16.34 Section 8 for special class services.
Performing NDE such as RT on a valve body/bonnet casting requires the full disassembly of the valve to gain unobstructed access to critical areas of the casting. Once an RT is complete and the castings verified as defect-free or the castings weld-repaired to meet industry codes, the valves can then be assembled, bolting torqued to OEM specification and the valve fully pressure-tested to ensure no seat or shell (atmospheric) leaks. The increased volume of RT inspections thus means an increased volume of valve pressure testing.
The world has become more environmentally conscious and the additional valve NDE can be seen as a result of this. This increase in NDE volume has meant addition of personnel and equipment; however, this is a positive sign for testers because it means increased work load.
TESTING BUTTWELD END VALVES
To perform a pressure test of a buttweld end valve on site requires the valve to be welded into a piping system or caps welded to the valve buttweld ends. Seat leaks detected on a valve body once a valve is installed in a piping system can be very expensive since the valve needs to be cut from the line and sent to a repair facility. Because of this, many end users request that buttweld valves for critical service have an independent pressure test performed to ensure the valves meet respective design pressure test requirements.
Valve manufacturers and repair/modification facilities use hydrostatic test stands that can seal buttweld end valves by using test plate adapters to seal the buttweld ends through a retained O-ring seal (Figure 2). These test plates are energized by means of a hydraulic ram. As the internal hydrostatic valve pressure is increased, hydraulic ram pressure also is increased on the test plates to equalize the external force against the valve buttweld ends with the internal valve hydrostatic force. This means an independent pressure test can be performed on a buttweld-type valve to ensure no defects without having the expense of finding a leak after it has been welded in a piping system.
SHELL PRESSURE TESTING
An aspect of pressure testing that has only become apparent in the past few years is the requirement to perform a 10-minute shell pressure test only on valves and fittings. The odd part of such a test is that the client only seeks a shell pressure test, not seat leakage testing. This would seem to imply that valves will be going into a piping system and that the client wants to meet the 10-minute pressure test requirements of ASTM B31.3 for process piping. Some clients have even requested testing forged flanges and fittings with a 10-minute hold.
A misconception of this test is that the fabricator still needs to perform a full shell pressure test on the piping system as a whole to ensure joints and connections are examined for leaks. Valves manufactured to meet ASME B16.34 will have a factory pressure test done that ensures the valve itself is leak tight. Most American Society of Mechanical Engineers (ASME), American Petroleum Institute (API) and Manufacturers Standardization Society documents that apply to valve pressure testing have much shorter test duration than process piping standard B31.3. In some cases though, depending on the test standard the valve manufacturer follows, the valve may meet or exceed the 10-minute shell pressure test duration. An example would be valves 12 inches in diameter and larger that are factory pressure tested per CSA [Canadian Standards Association] Z245.15.
The problem rests in the interpretation of the ASME B31.3 code—in particular, in the following areas:
The second paragraph within ASME B31.3, 326.3 states “The design, materials, fabrication, assembly, examination, inspection, and testing requirements of this Code are not applicable to components manufactured in accordance with the documents listed in Table 326.1, unless specifically stated in this Code, or listed document.” This poses the question: would a valve manufactured to meet the requirements of ASME B16.34 and pressure-tested at factory to API-598 criteria be exempt from a pressure test required by ASME B31.3? The answer is “yes,” in the context that the valve would have been tested to ensure no body seat leaks and no atmospheric shell leaks at the factory to meet the design criteria requirements of B16.34. However, the answer is “no,” in the context that paragraph 345.1 of B31.3 states that “each piping system shall be tested to ensure tightness.” Furthermore, Paragraph 345.2.2 (a) states “Examination for Leaks. A leak test shall be maintained for at least 10 minutes, and all joints and connections shall be examined for leaks.” The only way to satisfy this requirement would be to have the spool pressure-tested by the fabricator with the individual valves, fittings, pipe, etc. assembled as a complete unit and then pressure-tested as a whole to prove that the component joints (connection ends) are free from atmospheric leaks.
To conclude, a customer that wants to do an additional 10-minute pressure test on a valve to ensure no atmospheric leaks at an extended duration time compared to what the factory would have done is no problem. An extended duration test, however, does not relieve the fabricator from the requirements of ASME B31.3 to prove that the final joints and end connections that connect the valve to the piping system are to be examined for leaks. This growing trend towards such testing shows that users are increasing efforts to prove functional assurance of components provided through global supply, not to imply replacement of final testing required by ASME B31.3.