Thermal stratification on pressure vessels

[LesT]

1. Does the code set limitations as to the maximum temperature differences between two points in the vessel wall? other than for heattreatment?
2. And any limitations on the heating or cooling rate? (ie Thermal Shock)


Any good reference or guideline in this regard?

 

[arto]

look in Sec.VIII Div 2 AD-160 & appendix 5-130 for "thermal ratcheting"

 

[unclesyd]

I don't know how the newer editions of the code handle this now that thermal fatigue is addressed. Maybe someone more familiar with the newer codes can answer your question better.

We are operating Code designed and fabricated vessels under the following conditions. I hope this gives you some information.

We operated numerous fairly large S/S (304/304L) pressure vessels built to ASME codes under severe cyclic loading. We were suffering thermal fatigue or differential thermal fatigue long before it was considered by the codes.

We have 5 vessels that cycle on twenty minute intervals with a temperature differential of approximately 200°C along with a pressure differential of 250 psi. These vessels operate 24/7/365.
These vessels are approaching 50 years old and the repair weld metal has exceeded the original weight of the vessel. Numerous modifications have enabled the inspection interval to be increased to 2 years, but expect a lot of repairs.

There are also 12 S/S-C/S jacketed vessels that undergo cycling mainly form the jacket side. Vaporized heating oils, using pressure control, are use to heat the vessel to 550°F on a 1 hr cycle. When the hot vapor is initially condensed and collecting in the lower part of the jacket the initial thermal stresses created are very high. These vessels to undergoes frequent repairs.

Another set of all S/S jacketed vessels with a set rotating internals, again on the initial heat up (80°F-550°F) with vaporized oil, using temperature control, we have severe distortion of the whole vessel due variable heating rates of the shell and jacket. It makes a banana. This again is caused by the oil condensate accumulating in the lower sections. This vessel also undergoes the water boiling in glass pot syndrome preventing a better control of the heat up. The vessels are between 16' and 21' long and designed for FV and 70 psi @ 700°F

We usually didn’t see any problems with other S/S vessels that have a big temperature differential if it stays constant.

We also had problems with pressure vessels and pipe where we had poor mixing of two liquids (hot and cold) at the pressure boundary of same. This has caused some failures. This has occurred with temperatures differentials under 100°F.

Addressing these problems has been an evolutionary affair. All the modifications were within the frame work of the code. There were some that would have made a big improvement but were not allowed.

We have never seen any problems with C/S vessels operating under 450°F.

 

[LesT]

Thank you all, most helpfull.

The thermal fatigue as required by the code was calculated in our case, but does not seem to address the rate of heating or cooling. Or have I missed it? How significant is it?

UCS-56 d,2 eg states a heating rate of 400deg F per hour per inch thickness, for above 800degF. This however relates to heattreament. Would this apply to normal operation as well? Does this mean that below 800 degF there is no problem?

 

[prex]

The thermal fatigue must of course account for both steady state and transient temperature distributions. As you seem having considered steady state only, your analysis is not complete.
Thermal shocks may be grouped under two main categories:
- the true thermal shock, that is the rapid change in temperature of the fluid in the vessel, causing a thermal transient in the wall thickness. This phenomenon is unlikely to be relevant for normal fluids (unless you use liquid metals), but may be relevant for particular locations in the vessel (flanges, skirt connection) and anyway will add up to other thermal stresses.
- the transient temperature differential created under different parts of the vessel during an operational cycle. If you already evaluated the maximum temperature differential in all operating conditions, this may be already included in your analysis.
The heating rate for PWHT has nothing to do with all this and you can't base any type of reasoning on that. BTW if your operational transients are so soft, you really shouldn't worry for them.

prex

http://www.xcalcs.com

Online tools for structural design

 

[kyong]

LesT,

Please pay attention on this fact : If cooling or heating rate is high, temperature gradients will bigger and amplitude of cycle bigger, which means fatigue will come sooner. If you can evaluate this temperature gradient in terms of heating or cooling rate, you will be able to set an acceptable level of the heating or cooling rate.

kyong