B31.8 Hoop Stress Calculations with MSS SP-75 / ASTM A860 fittings or MSS SP-44 / ASTM A694 flanges

[gwalkerb]

I have two questions related to interpretation of certain B31.8 sections.

We build compressor packages, usually to B31.3, and so we use a lot of ASME B16.9 fittings and B16.5 flanges.

We've been building to B31.8 for some packages these days, and there's a minor ambiguity I want to confirm. DOT 49 CFR 192 (pipeline regulations) doesn't reference SP-75 or A860, although B31.8 does have some references to SP-75. Both documents do reference MSS SP-44 for flanges. So we specify all four of those ASME and MSS specs as appropriate (materials for the ASME fittings from ASTM A106/A234/A105). However, it's a slightly ambiguous as to how this affects my pressure design calculations. The relevant reference from B31.8 is below.

B31.8 831.3.1 (b)
Steel buttwelding fittings shall comply with either ASME B16.9 or MSS SP-75 and shall have pressure and temperature ratings based on stresses for pipe of the same or equivalent material.

While B31.8 does explicitly state the SMYS values for API 5L pipe, it doesn't for SP-75 fittings (or SP-44 flanges). I've taken the 'equivalent material' to be the WPHY grade that matches the API 5L grade (e.g. we would use MSS SP-75 WPHY60 fittings to connect to API 5L X60 pipe). However, the published SMYS for the SP-75 fittings are slightly lower than those for the equivalent pipe. For the aforementioned example, 5L X60 has a SMYS of 60,200 psi, while SP-75 WPHY60 has an SMYS of 60,000 psi.

Therefore, I've used 60,000 psi in my calculations to 841.1.1 (a) in B31.8, since it's all part of the same piping system, the lowest yield material would govern. This is all straightforward, but it's when I've been looking at the requirements of reinforcement (B31.8 831.4) that I'm not completely sure what stress value to use. Table 831.4.2-1 notes different requirements depending on the 'ratio of design hoop stress to minimum specified yield strength in the header'.

So my first question is this: given that we would only ever be making openings that could require reinforcement in piping (i.e. the header), and not fittings or flanges, is it acceptable to use the higher SMYS of the pipe for this hoop stress ratio calculation, rather than the limiting SMYS of the piping system based on the fittings or flanges? It's close enough that this won't make a difference in the majority of cases, but that doesn't cover every possible scenario.

My second question is also related to section 831.4.2 of B31.8.

831.4.2 Special Requirements. In addition to the requirements of para. 831.4.1, branch connections must meet the special requirements of the following paragraphs as given in Table 831.4.2-1:

My interpretation of this clause is that any of the special requirements of 831.4.2 are only applicable when that sub-section is specifically referenced in Table 831.4.2-1.

831.4.2 has sub-sections (a) through (k). However, Table 831.4.2-1 only references (a) through (j). (k) provides rules for using MSS SP-97 fittings (olets). Does this mean that (k) is always applicable, or does it mean that it is never applicable? 831.4.2 (k) doesn't appear to be referenced anywhere else in B31.8. (k) was just added in B31.8-2020, and doesn't exist in previous revisions, so it seems odd to me they would add it without any proper references to it's applicability.

 

[Rubje]

These are some interesting questions.

From what i understand, it is acceptable to use the higher SMYS of the pipe for the hoop stress ratio calculation in the reinforcement requirements of B31.8 831.4, as long as the pipe material itself is able to meet the requirements for the highest pressure and temperature conditions the piping system will be subjected to. The fittings and flanges are used to connect the pipes together, and the reinforcement requirements are specific to the pipe material, not the fittings or flanges.

 

[GBTorpenhow]

API 5L X60 should have a SMYS of 60,000 psi. API 5L 2004 is the version I am looking at, in this version table 3A/3B is in freedom units so the value you state of 60,200 sounds might be a back and forth metric conversion error?

In any case, the "in the header" implies that the SMYS they are referring to is the header pipe. Going by pure memory here but if you were using a grade B header you would need proportionally more material at the lower SMYS to replace the lost material in the X60 header, vs an X60 branch material. Logically, the header SMYS governs how much 'strength' needs to be replaced by the reinforcement area.

 

[gwalkerb]

Thanks to both of you.

GBT: I did go back and look at some of the historical versions of 5L and B31.8 to get some context. The latest version is edition 46 (2018), which is a little ambiguous, as it states that the SMYS would match the numerical portion of the material grade (i.e. X60 = 60,000 psi / L450 = 450 MPa). But in the minimum strength table (Table 6), the SMYS is noted as 60,200 psi (or 415 MPa). So you might be correct, the SI to USC conversion may have had an effect. But this is the specific number that API has published in the last couple editions.

With that being said, B31.8 has carried that same number through, and therefore within the boundaries of that code, the correct number to use is 60,200 psi. And DOT CFR 192 references API 5L 45th edition (2013), which has the same table as in the 46th edition (at least for those specific grades). So I'm comfortable with the slightly higher number, since API 5L pipe manufactured to either of those versions would need to meet the 60,200 psi minimum as documented.

 

[GBTorpenhow]

It's been a while since I have been on the 192 side of things, I stand corrected. Looks like API tweaked the SMYS numbers across the board to make it a little easier on the mm and MPa gang.