Stress Analysis for existing PSV systems

[DEEPBLUE76]

We have number of PSV systems in an 1970's refinery where the client is adding isolation valves on inlet and outlet lines of the PSV's. We want to know if we require to do stress analysis for these systems where isolation valves are added. if yes then is there any clause or section in ASME 31.3 or API which states any requirement. If no stress analysis is required we need to convince the client with some concrete reasons. Help reqd on the above issue.

 

[EnergyMix]

You don't say where you're located or why you feel stress analysis is required. This sounds like something I'm going to hear about on the nightly news one evening -- after a huge explosion or fire.

Want to know the do's and don'ts of Eng-Tips? Read FAQ731-376.
English not your native language? Looking for some help in getting your question across to others or understanding their answers? Go to forum1529.

 

[SNORGY]

Check your PSV inlet line lengths and tailpipe lengths. If the length of a block valve is significant relative to the pipe lengths, stress analysis would be prudent, since valves are treated as rigid elements in most analyses. The point masses might also contribute to dynamic effects.

 

[DSB123]

The additional valves may be weighty (dependant upon size). How is the weight being supported such that it does not hang onto the PSV? Normally PSV's should be installed where the connected pipe weights are kept minimal as any significant weight could affect the performance of the PSV itself due to bending moments. I would seek the advice of the PSV supplier if the additional valve weights are significant.

 

[LittleInch]

In retrofit engineering, you still need to design to either the code as it exists today, or the code as it existed when the plant was built, both of which requre analysis of the system to ensure it is capable of accepting the loads present. Which one is really down to agreement between you and the client and the differences between now and then as it is often difficult to "ring fence" changes in code within a plant. I am surprised that the refinery doesn't have policies and standards / specifications for modifications as there will probably have been onfgoing since it started work some 40 years ago....

Sections in B 31.3 are part 5, sections 319 to 322.

The actual impact is dependant on your actual design and piping configuration, but you're potentially adding some significant weight and mods to the piping in order to squeeze two more valves in which will probably have an impact on the systems, including the original process pipework and the vent pipework. Sounds like you don't want to do the analysis but without doing at least one or two "worst cases" it is very diffiuclt to prove a negative. If you pick two or three out which look like they could have the biggest impact, study / analyses those and then make your decision, at least you have some proper data to use. Probably not what you wanted to hear, but many incidents have been caused by retrofit engineering occurring piecemeal and everyone being able to justify not doing things until the cumulative effect overwhelms the, probably quite corroded, strength of the system.

My motto: Learn something new every day

Also: There's usually a good reason why everyone does it that way