ASME B31.3 302.2.4 Temperature Variation and piping flexibility

[DBreyer]

We have an existing pipe for which a new, rare set of process conditions was identified, Exceeding design temperature in the case of power outage (well below the creep range).



For the circumferential stresses we can assess this with the provisions in 302.2.4 as a temperature variation.

Sustained longitudinal stresses also shouldn’t be significantly changed, as this would only be the case if the pipe lifts of supports.

Now we ask ourselves, how to we assess the increased elongation due to this and the resultant secondary stresses? 302.2.4 only mentions primary stresses and makes no mention of stress ranges and paragraph 319 makes no mention of temperature variations.

My take on it would be, since the flexibility assessment is based on a fatigue analyses, this is covered by the provision in 302.2.4 that no more than 1000 load cycles are allowed for the variation.

This significant reduction from the typical 7000 cycles would also allow a higher allowable stress range.

Since the higher temperature will also only result in secondary stresses, which are self limiting, some yielding is expected but no catastrophical failure.



Do you agree or would you see the need for the temperature exceedance to be assessed as the new design temperature (including flexibility analysis)?

 

[RVAmeche]

I would say the proper thing to do, especially if this is a newer system where you have the original pipe stress analysis, is to update the pipe stress analysis and confirm its okay, support loads are still within acceptable ranges, etc. I would guess you're realistically correct that a very rare event probably won't cause a massive failure, but its hard to say in this hypothetical. If its a 40 year old system and you're going from 400F to 900F that may break some things.

If this is an older system that was only designed with flexibility analysis and not pipe stress, I think it'd be prudent to develop an analysis.

 

[Snickster]

I do not agree that flexibility analysis is covered by fatigue analysis.

I believe by 302.3.5.d if it is expected that the temperature of concern only happens one time in the life of a piping system it still needs to be considered based on the stress range factor "f", as far as allowable displacement stress range is concerned.

f = (N)^-0.2 < or = to "fm" (eq. 302.3.5.d.1c)

where maximum value of fm = 1.2 and "N" is number of displacement cycles.

Therefore when "N" is 1 then f=1.2 a maximum, so even one cycle at temperature of concern is considered.

 

[GD2]

Typically, Power Outage is considered in a PHA during the design stage. Can you provide what is the current (designed) design condition and the new operating conditions?

 

[KevinNZ]

The rules in 302.2.4 are for pressure and temperature variations above design but the criteria for allowing this is for pressure very little/nothing is said about the upper limit for temperature stress range variation. As noted above, therefore we are not allowed to exceed temperature stress range limit SA. If the design temperature range has maxed SA then you can not exceed design temperature. (Note I am not referring to latest B31.3, could have changed)

Interestingly B3.1 uses "ambient to normal operating temperature range" which could be less than min (winter cold shutdown) temperature to design temperature range.

B31.1 defines

displacement stress: a stress developed by the selfconstraint
of the structure. It shall satisfy an imposed
strain pattern rather than being in equilibrium with an
external load. The basic characteristic of a displacement
stress is that it is self-limiting. Local yielding and minor
distortions can satisfy the displacement or expansion
conditions that cause the stress to occur. Failure from
one application of the stress is not to be expected.
Further, the displacement stresses calculated in this
Code are “effective” stresses and are generally lower
than those predicted by theory or measured in straingage
tests.1

1 Normally, the most significant displacement stress is encountered in
the thermal expansion stress range from ambient to the normal operating
condition. This stress range is also the stress range usually considered
in a flexibility analysis. However, if other significant stress ranges
occur, whether they are displacement stress ranges (such as from other
thermal expansion or contraction events, or differential support point
movements) or sustained stress ranges (such as from cyclic pressure,
steam hammer, or earthquake inertia forces), paras. 102.3.2(b) and
104.8.3 may be used to evaluate their effect on fatigue life.

1023.2 (b) & 104.8.3 takes you back the same old SA stress range limit.

What even say highest range is to the normal operating temperature when design temperature will be higher?