I am new to pipelines. I have a problem of providing expansion loops on a long pipeline.

If I have the following:

1) Material (Hence allowable stress, coefficient of thermal exp), diameter & thickness of the pipe.
2) Temperature of the pipe.
3) Insulation information (Type, density, thickness).
4) the applicable code (B31.4 or B31.8)
How can I get the anchor to anchor length (between which I have to give one loop)using above information? Is there any equation / procedure? Or is it only some rule of thumb (each 50m or so)? Can anybody help?
_________________________
Shahid Rafiq Piping Engineer

If you just want a "ballpark" guess as to the number of loops required, there are many sources that give expected stress levels for expansion loops of particular dimensional ratios. Probably, though, you want to do some analysis to be a little more precise.

Typically, you want to anchor long pipelines between each loop, or each pair of loops. You choose an arbitrary distance between anchors (the longer the better) and find how big an expansion loop you need (the longer the anchor-to-anchor distance, the bigger the loop, but you want smaller loops). So you have to balance two conflicting goals - long spacing between anchors vs. small loops desired.

If you have minimal lateral loads due to wind, waves, seismic, etc. you might be able to accomplish both, but normally after one or two loops you have utterly destroyed the ability of the system to resist horizontal loadings. Don't forget to consider external occasional loads parallel to the long run of the pipe - if you have too many loops between anchors these loads could cause one or more pipe shoes to fall off their support steel.

After a few iterations on the anchor spacing and number of loops, you have a design where you have balanced the cost to construct, support, restrain, operate, and maintain the piping system. Usually you will need input from other disciplines (civil, structural, process, I&C, construction, operations, and maintenance) and from the owner in order to make this decision effectively. Note that process typically wants a few big loops (fewer fittings to produce head loss), while the other disciplines would rather have a lot of small loops.

Hello Shahid,

To expand a bit on what Craig has said above on getting your "ballpark estimate" and to give you some explicit guidance per your question please go here:

http://www.usace.army.mil/publications/eng-manuals/em1110-1-4008/toc.htm

Download Chapter 4 of this pdf file document and read page 4-15, et. al.

To find a trove of FREE manuals that you might find to be of some help in the future go here and look around:

http://www.usace.army.mil/publications/eng-manuals/em.htm

Regards, John.

For initial manual estimate, You can use equation given in Para 319.4.1 of ASME B31.3 to determine loop dimentions. Similar approach is explained in Peng's paper as well. Check the below link.

http://www.pipestress.com/papers/QuickFlex.pdf

Regards,

dear sir

you can not use loop in pipeline that ther is pigging on it .
you must use bend with the minimum radius of curvature 7D.
consider we have a pig with mass of M and velocity of V .in the rapid change of direction we have rapid change in direction of velocity vector and change to linier momentom creates large force that may deform the pipeline so for reducing the change in linear momentum we must use bends with large radius of curvature.

suppose that we are pipeline with 200 km lenght between refinery and our user.you can not analyze all of the root.so you must use iterative root (zig zag configuration)that you are assure that the stress and load are within the allowable.

ther are some procedure of design of pipe line in the IPS STANDARD
(iranian petrolume standard,you can download it from http://WWW.NIOC.COM)

best regards
a.sadeghabadi

Alireza - that equation will give you an incredibly oversized loop in most cases. You can design a much more economical installation with a couple of hours of work. If you have access to the old Grinnell "Piping Design and Engineering" book, that had a comprehensive collection of charts for sizing expansion loops that was very good.

I agree that pigging, if the product requires it, is a big issue with the design of long pipelines. Not only must the geometry be considered, but you must account for the transient forces created by the passing of the pig and/or the passing of a slug of fluid with a different density than the product.

I also agree that it is common to run long pipelines in a mild zig-zag route to create flexibility.

Ali Reza, This is not pipeline with pig traps at ends.

Dear All

can you enlighten me the significance of the intermediate anchor in the pipe rack for the multiple expansion loops. I understand that it is just only for seperating the each expansion loop and it functions independantly.
The axial load at the support point with or without intermediate anchor is same.Even expansion stress also same for the both cases.Assuming that it is a symmetrical loops.


regards

Sha..

Intermediate anchors ensure that in normal operation the loops stay in position and expansion is properly distibuted. In theory, when operating the axial load is zero on equal spans. However long pipe runs do not heat uniformly or instantly and during start-up in particular, very large axial loads can be generated in these locations.

To follow on from MoverZ - the practical effect of this is that, you are not going to report the rather small loads that you get out of Caesar to your Civil/Structural engineering group for design. Assuming a long bank of approximately equal loops, Caesar is going to report small loads on those intermediate anchors.

While they don't need to be designed for a full unbalanced load like an end anchor, good practice is to report a decent percentage of an end anchor load for to use for the design of the intermediate anchors.

I've seen it happen in the field where the C/S group didn't brace the intermediate anchor points well enough and the beam twisted over due to the torque as the upstream side of the run heated up faster than the downstream.