Valve design for shell integrity

[joeswoes]

The specifications (ANSI B16.5, MSS SP72) I've seen pertaining to housing/flange shell integrity require a 50% safety factor over the cold temperature rated working pressure. Is this to mean that, in a pressure surge of more than 50% over max. working pressure, one can expect catastrophic failure ? Anyone familiar with additional shell safety factor requirements ?

 

[peaseman]

The final safety factor for the valve shell(body & bonnet), in regards to internal pressure, is in the hands of the valve designer. For steel valves, ASME B16.34 and API 600 or API 602 (and others) stipulate minimum wall thickness requirements based on valve size and pressure Class. These minimum requirements assure a margin of safety based on many years of usage and experience. I personally have not investigated the safety factor used to establish the min. wall requirements listed in the standards. API 600 assures a greater margin of safety than ASME B16.34 as API 600 minimum wall thicknesses are somewhere between 35% and 50% greater. The additional wall stated in API 600 was added for a corrosion allowance as this specification is for valves in refinery and chemical plant applications, the added wall was not intended for an additional margin of safety but it does inherently add a margin of safety.
Regarding your reference to ASME B16.5 and MSS-SP-72 and your comment of 50% safety factor, I'm not sure I understand where you got this number. Neither of theses standards address or discuss margins of safety.
Perhaps you are refering to the requirement of valves being shell pressure tested at 1.5 times the rated cold working pressure. Please note that these are required production tests to verify leak tightness and are not for the purpose of determinng a margin of safety.

 

[unsy]

If you are using B16-5 or B16-34 it is not helpful to think in terms of over pressure. The maximum pressure a vessel or flange rated to these standards can be used for is based on the relevant rating table for the material. You must not forget temperature.

As peaseman said B16-34 is conservative in its approach when calculating wall sections and as you alluded to will be tested at 1.5 times the cold rating. If you calculated a shell thickness based on say B31.3 you would find a significant difference in sections.

Look at Annex F1.3 and the 1.5 factor shown in the calculation for Wall Thickness.

 

[Celebur]

Are you perhaps thinking of the hydrostatic shell test requirements?

ANSI requires that a shell-integrity test be done at 1.5X the rated cold working pressure. Note that API is somewhat different--they require 2X the cold working pressure for 5K and less, and 1.5X at 10K and up.

Your design needs to consider both the rated working pressure AND the shell test requirement. The relationships are as follows:

At working pressure: Sm = (2/3) Sy

At test pressure: St = 0.83 Sy

Your selected material must have sufficient strength to satisfy these stress limits.