Thanks to both of you for responses.

ASME B31.3 par. 319.2.1 in combination with 302.3.5 tells me that the max stress range is L8 = L6+L7. And the first part of Dave’s response confirms the same:
“…Any change in stress can cause fatigue damage. But by using a simplified approach, (most) Codes would include those displacement-based components be they thermal strain of the pipe or any sort of support movement….”

But then Dave said:
”… I would argue with your L8. Instead, I would use L8=L1-L3. There may be other ranges to check if the maximum range is not obvious. If you get more than one stress range that is "large", then you should be accumulating damage ala Eqn (1d)…”

L1-L3 apart from delta T (max) and Ds would include delta P and delta W! These delta P and delta W are not the full range so additional P2 and W2-WNC would have to be added…, but is this the Code requirement???

Operation of my piping system is cyclic and 2-3 times daily goes from empty to W1, P1, ambientT1, D1 condition to W2, P2, D2, T2 condition and back to empty. So it does the full cycle 2.5 times a day.

To summarize my understanding of the Code requirements in this case and in general:

1. Sustained Stress Check, as per L2, L4, L5 and any additional “hot sustained” case when non-linear supports / boundaries present

2. Full Expansion Stress Range, as per L8 = L6+L7 (319.2.1), but including stress range factor as per (1c) and (1d) considering all stress changes (due to P, T, W, D,…) to protect against fatigue. Also, in general, as Dave said:”… other ranges to check if the maximum range is not obvious…”

L9 and any other combined and / or individual FAT analysis are optional (nice to have in Caesar)

Ranka