Pipeline Thermal Expansion

[Pmatherne]

When running aboveground pipeline (30" diameter, not insulated)what is the best way to account for thermal expansion? Basically I am trying to find a way to absorb the expansion on runs over 1,000 feet with only 2 bends (one 45 degree and one around 70 degrees).

My first instinct was to use an expansion loop, which I tried calculating using the method in the Tube Turns manual, but the loop was 30 x 60 ft. Since there is very little horizontal room for the pipe corridor and the client does not want to install anything vertical this option got shot down.

My next idea was to use a slide bearing support on the corners and use the guided cantilever equation to let the pipe absorb some of the energy before using a guide to prevent lateral movement.

Is there any other option I am missing

Future PE Engineer
Pet project I am working on to help other engineers, not much yet hoping to get it grow as I learn more

http://www.peexamquestions.com

 

[BigInch]

You will usually see that hot lines steam lines in refineries, etc. have as many as 1 loop (4 90deg bends) per 100 ft of pipe. Use more bends. Make them 90deg. Generally for above ground locations 20 x 20 ft vertical loops are preferred in order to not block future parallel runs at the same elevation level. There is little choice in this method, other than reduce the operating temperatures. If you want to try expansion bellows .... well.... more loops will usually be more user friendly in the long run.

 

[LittleInch]

Pmatherne,

"Is there any other option I am missing " - Not easy unless oyu give us a bit more of a picture here.

Data that will make a difference includes
Design pressure
Contents
Thermal range being considered
Sketch plan showing bend locations and anchor locations - (fixed at the end of not?)
what is your corridor width?

I have seen lines zig zag across the desert before, literally one 90, short section another 90 then a long section before one 90...

This can be 90s at different orientations one after the other (left, right, left...) or LRRLLRR...

As BI says bellows can be used, but you need to get this on a stress model and see what sort of friction pads you need on your supports. You can get lock up on long supports if you're not careful. Only when you can see what expansion you need to cater for can any reasonable judgements be made.

Just because the client doesn't like it doesn't make the problem go away.... Loops are very common and simple to install.

Not sure what this "guided cantilever equation" is, but bending stress sounds like a much better method to me....


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

 

[borobam182]

If you know the material and the design temperature, you would be able to calculate the thermal expansion per unit of length - that would help to assess how big and how many loops you would need. These are the main drivers behind thermal expansion values.

If I was you, and if you have access to stress analysis software, I'd create a model of the pipework system, apply all of the conditions and material properties, apply the approproate end conditions, and run some test case model - looking at support types/locations, and loop size and placement. Position and type of supports used will influence how the pipework behaves local to any loops you introduce.