I am on a project, where the site ambient temperature can be as low as 10^C and can be as high as 85^C. The project directive is to use 10^C as the ambient temperature (by changing the default 21^C in Caesar) when the line is hot and use 85^C as the ambient temperature (again by changing the Caesar default value of 21^C) when the line is cold. I understand the intent to consider the greatest temperature range, but is it really correct to do so, i.e. to change the value of the ambient temperature.
My next question is: is it true that piping are really installed at those extreme temperatures ?
Lastly: If the dimensions in the isometric are fabricated at 21^C, and the installation temperature is different, what are normally done when these situation occurs?

Thanks in advance,

Hello Homer,

It is important to remember that when calculating thermal stress RANGE, you are not calculating the stress when the system is extremely cold (10 degrees C) or calculating the stress when the system is extremely hot (85 degrees C). Rather you are calculating the range of stress between the two extreme temperatures. As you say, the intent of the project directive is to design for the range.

If the lowest metal temperature that the piping system will experience is 10 degrees C and the highest metal temperature that the piping system will experience is 85 degrees C, your design temperature range is (plus or minus) 75 degrees C. When calculating the thermal expansion stress range, it really will not matter much what the temperature was when the piping system was installed as long as it is somewhere between the two extremes (the temperature at which the last "closure weld" is made will determine how much inadvertent "cold spring" will be "installed" in the system). You are doing a linear elastic analysis. You can set the ambient temperature at 10 degrees C and use a "delta T" of positive 75 degrees C (i.e., make the design temperature be 85 degrees C), or you can set the ambient temperature at 85 degrees C and use a "delta T" of minus 75 degrees C (i.e., make the design temperature be 10 degrees C). The (unsigned) calculated stress range will be the same.

If you use 21 degrees C as the ambient temperature in your analysis, then you can do the thermal stress range calculation using a design temperature of (the "delta T" is positive 75 degrees C) 96 degrees C. Or, you can do the thermal stress range calculation using a design temperature of (the "delta T" is negative 75 degrees C) minus 54 degrees C (if I got my sums right). The range will be 75 degrees C either way. Just be careful of the difference in allowable stress at temperature; you want the allowable stress at the higher temperature.

Piping systems are not normally installed at extreme temperatures. The temperature inside the pipe fabrication shop (when the pipe spools are fabricated) will usually be close to the (favored for human comfort) 21 degrees C. If there are great differences in temperature during construction, the pipefitters will find the piping in a different position in the morning when they report to work than the position they left it in when they left the site for the day (they learn to adjust themselves to that). There will be less of that if the system is erected (installed) at a temperature close to the temperature at which it was fabricated.

Regards, John.

Dear Mr. Breen,

The given conditions are these:

Line design temperature: 300^C
Ambient temperature : 10^C ~ 85^C

Approach 1: (project directive)

Go to Kaux, set Ambient temp to 10^C
set T1 = 300^C
then the load cases will be:

L1) W+P1+T1 (OPE)
L2) W+P1 (SUS)
L3) L1-L2 (EXP)

Approach 2: (my personal opinion)

Use default Amb temp = 21^C
set T1 = 300^C
set T2 = 0^C

The load cases will be:

L1) W+P+T1 (OPE)
L2) W+P+T2 (OPE)
L3) W+P (SUS)
L4) L1-L3 (EXP) > from 21 to 300
L5) L2-L3 (EXP) > from 21 to 0
L6) L1-L2 (EXP) > from 0 to 300

If it is most likely that the piping will be fabricated at 21^C (following the dimensions on the isometric which we input in CAESAR), it is best to assume that the piping will be installed at 21^C instead of one of the extreme temperatures.
Also, I am worried about the effects of contraction (hitting adjacent pipe / lift-off)that was not considered on the first approach.
Am I on the right track ?
Lastly, do I still have to consider the possibility that the pipe will be installed at 85^C(having the same dimensions on the isometric, i.e. neglecting the change in dimensions if the piping were fabricated at 21^C))then shutdown at 10^C (I think this will show the worst case displacement due to contraction)?

Hoping for your understanding,
delpilar

You are right to use load case combinations instead of changing the ambient temperature. This calculates the intended range of stress, avoids confusion on appropriate allowable stresses, and gives you a more realistic operating case to help address your displacement concerns.

In your problem, any ambient temperature piping will have at least 75^C thermal expansion.

For such piping, while laid straight on a pipe rack, will you allow for expansion in changes of direction in piping layout or allow for 3D expansion loop (each loop made with extra 6 LR elbows & some pressure drop & additional drainage points etc) for every 150/200 m straight run to keep axial movement of piping resonable with respect to standard pipe shoe length ?

regards,

sam

Delpilar,
I think the authors of your project directives meant to say degrees F and not degrees C.
85C is 185F.
The hottest recorded outdoor ambient temperature on Earth is reported to have been 56.7C (134F) in
Death Valley, California in 1917.
If the authors of the project directive truly mean 85C, I would question that number.

The site may be in a very warm spot next to a volcano or hot springs.... or in hell.....

The other possibilty is the 85C is trying to account for solar gain...

Its hard to say but the range is the range.... According to the code it must be accounted for one way or another...

(although I'm willing to bet it often times has not been properly accounted for)

After giving this more thought, I looked into the topic and have read from a seemingly reliable source that in equatorial regions, painted metal temperatures can reach 80C (light colored paint) to as high as 120C (dark colored paint).

Recalling that I've nearly burned my hands on my car on hot summer days (nowhere near the equator unfortunately), I hereby humbly retract my previous statement.

eh thats fine your locale is a bit north of the equatorial lattitudes...

but range is really the topic here and I stand by my comments its usually missed....

After investigation, I have out that the 85^C (which the project wants us to use as ambient temperature) is really the maximum surface temperature under the sun.
Actually I dont question the range. What I question is the proper way to take the range into account. Mr.Howard gets my point and I am happy to know that I am not alone.

Regards,

hm ?
If you have temperature on a pipe under sun, then you have in shadow site lower temperature.
And you have additional thermal bowing effect.
Best Regards
Ohliger

Try this:

T1=300C
T2=10C
T3=85C
(default ambient=21.1)

L1:operating loads with T1
L2:operating loads with T1-T2 (as in W+P1+T1-T2)
L3:operating loads with T1-T3

L1 will produce results for ambient of 21C,
L2 will produce results for ambient of 10C,
L3 will produce results for ambient of 85C

I get the same strain in L2 & L3 as when I manually change the ambient while running just T1.

Confirm your own results...

Thanks Mr. Diehl. I already confirmed the results. Even without changing the ambient temperature using Kaux, we can check the displacements from different ambient temperatures just by using the proper load combinations.

Regards,
delpilar

I am working on a project where min ambient 10 deg C & max. ambient temp 52 deg. C. We are considering 10 deg as ambient for hot lines. & 52 deg temperature for cryogenic lines to cover maximum stress range.

For all rotating equipment it is define installation temp. as 27 deg C.
Can I take any advantage of that to reduce nozzle load of rotating equipments.

Regards

Habib

Hi Mr. Diehl,

Just encountered this old post...

Is this concept (subtracting temperatures as in "W+P1+T1-T2") still true after having changes in the softwares, ASME codes, analytical attitudes, etc?

In a very quick/rough study I found it not as much precise via looking at displacements, element forces, stresses and so forth, and there seem to be some discrepancies. Am I right?

any comment on the subject?

CAESAR II still functions in the same manner in this regard.