Dear all,

I have been said to justify all the welded attachments in a HP steam system of a Combined Cycle Plant (B31.1) by means of Code Cases N-318/N-392. I am facing some problems I am not very sure how to resolve and would be grateful for any help:

- The code cases are intended to use only for attachments to straight pipe but I have some trunnions welded to elbows. Do you think that the method is conservative for elbows or on the contrary if I use it I may understimate the stresses? For example MW Kellogg method is usually used for trunnions on elbows but I think it was created for straight pipes only...
- The Code Cases do not consider any reinforcing pad. As I have pads in some cases. I thought of using a pipe of "equivalent" thickness equal to pipe real thickness+pad thickness. Is there any recommendation about the minimum pad dimensions in order to assure that the local stresses calculated with this "equivalent" thickness will be correct? (I suppose that the pad must cover the whole area affected by the local stresses, which is that area?)


Thank you and best regards,

My opinion is ;
Both methods are conservative for elbow trunnion because the contacted/welded line is longer than straight pipe.
Based on my experience and practice, it is not a big deal to use the total thickness(pad + pipe thick.) because both N-318/392 and Kellogg are thumb approach and conservative comparing to FEA.
More important things are to use both methods properly, that is, to select proper load cases and allowable values for primary and secondary stress evaluation in accordance with individual piping codes.

I don’t have these papers in my hand but I used these code cases for a while.
My understanding is Code case N-318/392 is provided for the strength evaluation of welded attachment on pipe in nuclear power plant. However, I think, in these days…these papers are not used that much because simplified tools with finite element analysis(i.e. FE/Pipe) is more popular for critical case check. Also, they are not well known to stress peoples in oil & gas industry(B31.3).

Major differences;
Kellogg method is just simple calculation
- stress check only at the junction
- not subjected to weld size..it counts just as line

Meanwhile; ASME Code Case N3-318/N-392
-check the stress of attachment and base anchor as well local stress at junction
-weld size and weld type(concave, convex, etc.)
I would like to say Code Case N3-318 and/392 are more detailed formular and standardization. The rules and limitation are almost similar conception as pressure vessel and piping codes.

Dear Carletes,

I've used the both calculation approaches.

In my understanding, the both methods were developed for "straight pipe - pipe trunnion" type junction.

However, in my opinion, ASME Case 392 method yields more conservative results than Kellog approach. Besides, ASME Case 392 regards the local stress states developed both in the pipe and in the trunnion wall, while Kellog method is applied (after my knowledge) only for the main pipe stress checking.

So, unless otherwise required by the Project specifications/procedures, I generally use ASME Case N-392 method, both for straight pipe and pipe elbow trunnions/dummies. This aaproach appears to be conservative enough for the both cases.

However, if overloading/overstress problems occur and the "pipe/elbow-trunnion" reinforcement possibilities are limited, it would be necessary to involve Finite Element Analysis approach (such as FE Pipe software) for an accurate analysis purpose.

If you still prefer to use Kellogg approach, be careful to superpose the local stresses developed by the attachment local loads over the initial stresses induced by the main pipe internal forces and moments.

In the past I saw some versions of Kellog method (developed as mandatory Project procedure) where the initial stresses developed by the main pipe internal forces and moments were ignored.
From this point of view, it should be noted that ASME Case N-392 takes into account explicitly the initial stress states of the main pipe.

Best regards,