Pre-stressing Expansion Loops During Construction

[Mechville]

I have designed an aboveground pipeline with an expansion loop to account for thermal expansion from ambient temperature. The client has requested information on prestressing the piping/loop based on the temperature at the time of construction. I'm generally aware that this is possible from previous conversations with engineers and construction teams, but I cannot find any guidance to provide to the client. Can someone point me in the right direction? Is it as simple as installing a bit more or less length on the straight piping spans, using a linear relationship between my expected length of expansion as a function of temperature?

E.g. if I've calculated 10 inches of expansion for a 100 degree temperature range, do they install 1 more or less inch of piping for every 10 degrees off the median temperature?

 

[RVAmeche]

I'm not sure what the client is asking for? Was the loop designed assuming 70F to process temps and it's being installed in winter so they want to account for that difference? Or are they asking you to cold spring an expansion loop? Which I've never heard of.

 

[Mechville]

"Was the loop designed assuming 70F to process temps and it's being installed in winter so they want to account for that difference?"

Yes, exactly.

 

[RVAmeche]

That's not something I as an engineer have ever had to provide, but I believe it's common practice for the construction team to take that difference into account when making their cuts if it's significant (ie being built in 20F winter).

 

[TugboatEng]

This sounds dangerous. I'd think you would want the pipe to be relatively neutral at ambient temperature when being constructed or serviced to minimize stored energy around the workers.

 

[EdStainless]

I have seen this issue with large (long) heat exchangers and SS tubes.
(length change for 80' SS tubes between mill at 110F and field at 20F is 0.75")
As long as everything was measured with a tape that has the same expansion and is at the same temperature the lengths will be correct.
However, at installation you may end up with gaps between flanges because the end points are fixed and pipe has contracted.
I have seen electrical heat tape used to expand tubes to fit.
Not sure how you heat piping.

= = = = = = = = = = = = = = = = = = = =
P.E. Metallurgy, consulting work welcomed

 

[KevinNZ]

I would assume the piping is drawn at the ambient temperature the designer has assumed for when the pipe is being built. The design will/should have allowed for expansion from ambient to the design hot temperature and for contraction from ambient to min temperature the pipe could see. If the pipe is being built at temperature away from the designer's ambient then this may need to be compensated with a change in build location of the pipe at the supports.

This can be a problem for lines with large expansion range and pipe needs to be correctly located between the side tabs gaps on guide supports.

For example a pipe with guide side gaps of 3 and 200mm at ambient will pull back on to the 3mm gap when the pipe is colder than ambient. If the pipe is built colder than ambient (with 3/200 gaps) it could expand pass the 200mm gap.

 

[Dhurjati Sen]

Check if cold pull of the piping is necessary.
Design of proper support with load calibration needs to be established.

DHURJATI SEN
Kolkata, India

 

[LittleInch]

Is it as simple as installing a bit more or less length on the straight piping spans,

More or less yes. The real difficulty is making it happen. If your installation temp is lower than your design basis temperature then you somehow need to expand your expansion loop. That will need some sort of hydraulic jack to be able to weld the pipe. Reverse for the temperature being hotter.

They do this for hot pipelines, especially in Canada for their bitumen oil lines. They run hot water through the pipeline, let them expand then weld and back fill before they contact again. I've also seen it for heat traced lines where they turn the heating on and do the same thing to lock in some thermal expansion.

Remember - More details = better answers
Also: If you get a response it's polite to respond to it.

 

[1503-44]

If its an expansion loop, its usually flexible enough that you can use the D9 method.

--Einstein gave the same test to students every year. When asked why he would do something like that, "Because the answers had changed."

 

[Mechville]

If its an expansion loop, its usually flexible enough that you can use the D9 method.

I'm not familiar with the D9 method?

If the pipe is being built at temperature away from the designer's ambient then this may need to be compensated with a change in build location of the pipe at the supports.

Right, I understand that and I'm trying to determine how to compensate for the ambient temperature at the time of install. Can you elaborate on the proper way to compensate for the change?

Thanks.

 

[LittleInch]

I'm not familiar with the D9 method?

Other bulldozer manufacturers are available....

Remember - More details = better answers
Also: If you get a response it's polite to respond to it.

 

[Snickster]

Right, I understand that and I'm trying to determine how to compensate for the ambient temperature at the time of install. Can you elaborate on the proper way to compensate for the change?

I believe you can just change the 70 F default installation temperature in the computer program such as Caesar.

Also you could just adjust the design temperature that you input into the program so that the overall cold and hot delta T's are still the same. For instance if the actual design temperatures are 10 F and 120 F with the default installation temperature 70 F but an actual installation temperature of 50 F. Then you could input the cold design temperature as 30 F and the hot design temperature as 140 F to get the same delta T's between installation and operation. Although at very high temperatures the program calculated allowables may be sligthy in error based on temperature of metal stress values.

 

[1503-44]

You don't actually have to do any prestressing, unless your additional pipe is at a different temperature then the pipe you are adding it to. Just make sure all pipe is at the same temperatures before cutting to the tie in length and prestressing is avoided.

But ... what you may want to do, is fabricate all pipe such that that you wind up with the dimensions shown on the drawings when the ambient and pipe temperature is equal to the defined installation temperature, even though the current fabrication temperature is different.

Meaning ...

Assume that pipe dimensions on the drawing are given at the installation temperature, say 70°F and alpha is 0.0000065 in/in/F°. If you then actually install at 30°F, a 100 ft length between flange facings on the drawing will be 100 x12 - (100 x 12 x(30-70) x 0.0000065) = 1200 inches -0.312 inches = 1199.688 inches, or 99.974 ft length at 30°F pipe temperature.

Use a laser measure, not subject to measuring temperature contracted dimensions.
When the pipe is at 30°F, cut your pipe to a length of 99.974 ft - (2 x flange neck dimensions).
Then locate the tie-in flanges to the flanges on your pipe, or visa versa. If they do not fit, then you know that the pipes are not at the same temperature and some force will be required. (That's the prestress). Otherwise make all pipe temperatures the same; they will fit together at any temperature without prestress, if you use this method.
Lastly, bolt it up.

When the temperature rises to 70°F, the distance between flanges will be 100 ft.

If you are installing an expansion loop, cut all 70°F dimensions on the drawings to their respective corrected dimensions at 30°F actual installation temp.

If your actual installation temp is higher than 70F, then all cut dimensions will be longer than what appears on the drawings. When it cools to 70, your dimensions will match the drawing.

When fabricating a tall steel offshore platform jacket over a period of 6 months, temperature must be taken at the time every cut is made and the temperature corrected dimension is cut. That way the fit up is maintained without warping as the seasonal temperatures vary from morning to afternoon and January to July.

--Einstein gave the same test to students every year. When asked why he would do something like that, "Because the answers had changed."

 

[Snickster]

KISS

 

[Snickster]

Pre-stressing an expansion loop is the most hair brain brain fart that you can ever do in engineering. An expansion loop is supposed to, and able to, take up all thermal expansion that exists. Why in the world would you cold spring an expansion loop. You are trying to make up for your inability to calculate an abstract thing with a physical compensation that don't make sense. Just put in the correct size expansion loop based on physical realities which are based on sound calculations that you need to be able to do.

 

[1503-44]

The only time I've ever done it (D9 method) is when cutting into a pipeline to weld in a tee for a new connection, or to replace a valve, or short piece of pipe, or when pulling in a connection to an offshore riser, when the pipeline is not yet buried in the mud. When cutting into an existing line, residual stresses sometimes can cause a lateral offset in alignment that isn't great enough to cause you to want to install a pup with a couple of trimmed elbows, or if you can make it up with a misalignment of less than 3°. In those cases, pushing or pulling the pipe into realignment can often be done, without introducing too much "prestress", if you have some bending flexibility there. Correcting axial offsets usually is difficult because of the typically large amount of stress such a correction adds to the pipe, and if its only an axial misfit, just cut an accurate pup length and you're done.

Prestressing, as it exists in structural engineering, to artificially increase compression stress during installation, only to reduce, or reverse it when loads are applied later, or visa versa, is not a concept that is favored in the piping world. The greatest stress in piping is the tension from hoop stress and its usually impractical to apply pressure outside of the pipe in an attempt to reduce the maximum hoop stress later under operational loads. But there is one notable exception to that, which is an offshore pipeline in deep water. In that case, water pressure outside the pipe provides a compressive prestress to the hoop stress that is later reduced and usually reversed (but not always) into tension as pressurised products are introduced into the pipeline. Be careful if the pipeline enters shallower waters.

--Einstein gave the same test to students every year. When asked why he would do something like that, "Because the answers had changed."

 

[LittleInch]

I disagree with snickster here.

If the installation temp is lower than the figure used in design, but you weld it all up in a neutral stress condition, then it will expand more than predicted and could overstress the pipe or fall off a support.

However compensating for this is not easy as you then need to press stress your expansion loop, I.e make it act as it will in operation if your tie in temperature was 70F not 30F. Also your calculation or analysis has a margin of error.

Best thing to do is heat the pipe up to the designed neutral install temperature just before you tie in to the expansion loop. Running hot air through is sometimes used. Physically expanding the expansion loop can require considerable force which needs to be held in place whilst the tie in weld is made.

Remember - More details = better answers
Also: If you get a response it's polite to respond to it.

 

[KevinNZ]

The OP has an above ground pipeline stetting on supports. The cold/fabrication set up just needs to consider where the pipe is located at each support. The design support gaps and slipper/shoe location may be for a different ambient/construction temperature. Corrected setup gaps can be calculated.

Piping design under B31.1 (or B31.3?) does not require or give credit for cold pull (prestressing). These codes are look at thermal stress range only. -0% to 100% is the same as -50% to +50% and-100% to 0%.

But the supports and things the piping is attached to may have design rules that do not use thermal range so 0%-100% may be 2x higher loads that -50% to +50%. In this case some cold pull might help with the design of these elements. I have only seen cold pull used to reduce to nozzle loads on vessels. The amount of cold pull was petty small and seemed to an academic exercise.

 

[1503-44]

Codes do not address prestress, outside of the external pressure aspects, but neither do they prohibit it.
A prestress situation simply becomes another temperature load case at some equivalent temperature as soon as the pipe is welded up. If it only involved axial stress, the equivalent temperature could be easily calculated and it would become the "actual installation temperature" condition, which could be different than the actual ambient temperature at the time of welding. If there were other than axial stresses introduced, the equivalent temperature could still be calculated, but at a cost of significantly more "CPU time".

--Einstein gave the same test to students every year. When asked why he would do something like that, "Because the answers had changed."