Pipeline Expansion Loops. 7

[Frank1344]

Hello Folks,

Why do we often use 45 degrees bends (3D, 5D, etc.) rather than a Standard 90 Degrees Elbow?

My understanding is that we will have less pressure drop and less stress.
Also it looks that bends can handle hammering, slug flow better that 90 degrees elbows.

I would really appreciate if I can have your comments on the subject.

Best Regards,
Frank
PS: Sorry if the subject has been discussed before.

 

[pennpiper]

Based on my experience I think there is a much simpler answer.

Consider the vertical dimension difference available between north/south runs in a pipeway and the crossing east/west runs.

Two 90 degree elbows may exceed the space available. Whereas a 45 degree ell and a rolled 90 degree ell (with or with-out a short pup piece may work just fine.

Sometimes the simplest answer is the right answer.

 

[dcasto]

smart pigging

 

[Frank1344]

Thanks pennpiper,
The loop I am asking is in one plane, there is no vertical legs.
I think the one you are talking about is usually on the pipe rack inside the plant.

Mine is on the Pipe Way in off site and pipe is laid out very close to the ground.

Or I misunderstood your response.

Thanks again,
Frank

 

[Frank1344]

Thanks dcasto.

There is not going to be any pig launch/receive station.

 

[Ziggypump]

45 expansion loops nest naturally, and don't need different sized pipes for a nested 90 loop. Also, if not all, most of the 45 loops are also seismic rated. A little thing, but it's there.

 

[BigInch]

It is because its easier to more or less parallel a road, railroad, or power line right-of-way, or a river valley and when you cross it back and forth a couple of thousand times, the pipe you save by using Pythagoras' theorem can pay your and a few other's salaries.

Plant piping on the other hand is either mostly up and down or horizontal anyway and, if you use 45s, you wind up blocking out a much bigger 3D block of space, which may inhibit adding more pipe to the piperacks and pipeways in the future, or interfering with the cable trays.

 

[dcasto]

your title says pipeline. That to me means a pie going a long distance. In the US, all pipelines must be piggable.

 

[BigInch]

Usually not a problem for 30[deg], 45[deg], 90[deg], or X[deg] field bends, as long as the radius will allow the pig to get around it and normally a 5D, if not 3D, will handle most instrumented pigs these days. They used to be much larger in the olden days. Most of the time the only problem we might have are those half-open valves and a tee or two with a relatively large outlet without any guide bars installed in it.

Hard solid content, dense slurries, and Roto-Router type cleanout requirements for some kinds of pipelines might also lead you to favor 45[deg] or X[deg] bends.

 

[waterpipe]

For large diameter pipes it's the matter of manufacturing the bend (in the factory or at site by using a bending machine). Think of a 1800 mm, 90 degree bend that might aim to be lesser than 3D. Just consider the dimensioning of the inner and outer surfaces and the stresses at the pipe wall.

If using bending machine the limitation could be the maximum allowable changes in the pipe wall thickness and if you use miters then it could be the welding and fitting.

Thrust forces are also important in water pipelines depending on pipe diameter, pressure and backfill height. Direct 90 degree bend could mean a big thrust block.

As you said, pressure loss can also be of concern all depending on your size and pipe flow velocity. In a biginch it could be not that important for bends less than 30 degree but you should do the calculations case by case. For a water project with 60" pipe and a normal velocity for a 90-degree break with 4 pieces each 22.5 degree (according to AWWA), the headloss is equivalent to 40 meter straight pipe while for a direct 90 degree bend it would be equal to 250-300 meter pipe.

Another concern is the more smooth flow streamlines when dividing the break (bend)to smaller pieces. Sudden 90-degree bend distracts the flow streamlines. This causes a rotating vortex right in downstream of the bend which blocks the pipe partially and makes the nearby passing flow to accelerate. This will bring localized high velocity which could be of concern and also stimulates flow-induced vibrations. Take a look to flow velocities in this CFD model (red and blue are high and low velocity magnitudes respectively) and observe the rotating vortices next to 90-degree bend.

BTW, pigging is not common in water pipelines.

 

[rconner]

While the title of this post is “Expansion Loops”, your original question sounds a little vague as most such pipe loops we see in at least general literature depict either large, graceful “bent round” loops, or instead basically square-cornered loops employing (not 45 but instead) 90 degree bends at the “corners” of such devices. It also appears most of the prior responses to this thread are directed to a less specialized question of what angle bends are best to accomplish arguably more generic pipeline off-set situations (i.e. not trying to satisfy thermal or other axial length variations in the system, as is a normal function of expansion loops), such as simply transitioning to a different line and/or grade.
While all have made good points, and I suspect any sorts of bends might work for some expansion loop circumstances, see however a fundamental design rule expressed for these specific devices e.g. at

http://books.google.com/books?id=4m...11vjevcQ6Ih8DeP6eg&hl=en#v=onepage&q&f=false,

“Design and installation of these devices should be such that maximum stresses will not exceed allowable levels established in codes, specifications, and practices appropriate to the area or facility involved.”
While I would not purport to be an expert in these specific devices, I suspect depending on the geometry and space constrictions you could be dealing with, all 45 or 90 cornered loops might not necessarily be equal in this regard (unless perhaps you put two very close together 45’s at each otherwise ell location?) More information and sketch of what you are really trying to do might elicit more specialized help.

 

[Frank1344]

Thanks a lot Folks.

Here are the line data:

24", Steam 354 C @ 17240 kPag,

Pipe Material Carbon Steel Grade 550 MPa Yield Strength

Wall Thickness 19.1 mm

Design Code :CSA-Z662, 2007

 

[waterpipe]

O.k, then this removes most of the concerns that I'd mentioned.

 

[IFRs]

The OP also mentioned the radius of the ells. From what I understand, the 3D and 5D are usually only required for pigs, otherwise long radius ells are used unless they don't fit in which case a short radius ell is (reluctantly) specified.