Bent Pipelines
Bent Pipelines
(OP)
An ILI found dents >6% deep near 3 of our underground guides in a double S-bend riser for an NPS 6 pipeline. The dents are located at the 6 o'clock position. Verification digs will be performed in a few months, but we are thinking it is either caused by rocks (not likely to have caused all 3 dents), the support falling off the guide or the support piling has moved.
Has anyone else come across a similar situation with a different cause for the damage? Corrosion does not appear to be a significant factor.
Has anyone else come across a similar situation with a different cause for the damage? Corrosion does not appear to be a significant factor.





RE: Bent Pipelines
RE: Bent Pipelines
Buried pipelines should be continuously supported by soil. Introduce hard points /point loads and you have huge trouble..
Can you sketch this or describe it better please.
Remember - More details = better answers
Also: If you get a response it's polite to respond to it.
RE: Bent Pipelines
RE: Bent Pipelines
RE: Bent Pipelines
The ILI data indicates that it is right where the support is, but we don't have verification of where exactly it is.
RE: Bent Pipelines
These should be removed forthwith and never installed again.
Remember - More details = better answers
Also: If you get a response it's polite to respond to it.
RE: Bent Pipelines
RE: Bent Pipelines
RE: Bent Pipelines
RE: Bent Pipelines
The risk for any point load or differential settlement is a shear failure causing a complete rupture.
I remain absolutely astonished that a few of you seem to think this is somehow "normal practice".
Whoever started this, which then somehow got moved onto the next project as something that is normally done needs to cease pipeline engineering and go back to piping where they belong. Is this a strange Canadian thing? Wherever it comes from it needs to stop - I though anchor blocks were bad enough, but "Underground guides" are far far worse. Think yourself fortunate that you only have a dent - it could easily be much worse.
Point made?
Remember - More details = better answers
Also: If you get a response it's polite to respond to it.
RE: Bent Pipelines
Differential settlement of a structure to which the pipe is rigidly connected can induce not only high bending moments, but also shearing forces. These forces and moments are set up when the structure and/or the pipe moves laterally with respect to the other. Quantitatively, these induced stresses are not easily evaluated. Effort should be made during design and during construction to see that differential settlement is eliminated or at least minimized. This can be accomplished by the proper preparation and compaction of foundation and bedding materials for both the structure and the connecting pipe.
A realistic design load for a pipe is the prism load, which is the weight of a vertical prism of soil over the pipe. The prism load will be different for the pipe support and the pipe because of the different cross-sectional area. Also, a true trench condition may or may not result in significant load reductions on the pipe since a reduction depends upon the direction of the frictional forces in the soil.
RE: Bent Pipelines
Since the original design information is unavailable and this line has seen service, I am hesitant to remove the support/guides only to have the loads transferred to the pipe as it comes aboveground or at the aboveground support.
RE: Bent Pipelines
There are no mysterious loads other than in the mind of demented piping engineers.
Design this properly and there are no issues. Don't be beholden to a piping engineers view of buried pipeline design. They really don't understand it, hence this monstrosity you've been landed with.
Dig up the entire thing, remove these point loads, compact the ground again then you won't have an issue ever again.
Remember - More details = better answers
Also: If you get a response it's polite to respond to it.
RE: Bent Pipelines
RE: Bent Pipelines
RE: Bent Pipelines
RE: Bent Pipelines
I like this thread ... it has "demented piping designers" and a strange "Canadian thing" ..... lot of potential for cross-border conflict !!!
MJCronin
Sr. Process Engineer
RE: Bent Pipelines
To have a proper isometric and something which some one has clearly thought about and still done is just mind boggling.
BTW, it would be Trans Atlantic conflict....
Remember - More details = better answers
Also: If you get a response it's polite to respond to it.
RE: Bent Pipelines
If we remove the supports, what can be done to ensure that the pipeline doesn't settle excessively and cause dents elsewhere?
RE: Bent Pipelines
RE: Bent Pipelines
RE: Bent Pipelines
RE: Bent Pipelines
Generally the only successful option is to excavate down to something quite hard, build it back with rocks / stone etc and then lay the pipe in or on that surface.
Not great ground normally but then I assume the AG section has some good foundations?
Differential settlement is not good news for any type of pipe, but putting supports there just makes it even worse, especially for small diameter, thin wall pipes like you have.
The key is to put in sufficient material to prevent differential settlement, monitor the situation and allow some flexibility in the incoming pipe.
Either that or extend the supports and make it an above ground pipe.
Remember - More details = better answers
Also: If you get a response it's polite to respond to it.
RE: Bent Pipelines
What do you mean by extending the supports to make it an aboveground pipe?
RE: Bent Pipelines
Remember - More details = better answers
Also: If you get a response it's polite to respond to it.
RE: Bent Pipelines
Does anyone have any experience with the idea of removing the supports but putting a plate (say 3/16"x2'x20') below the pipe to give the pipe a larger area to bend over and also restrict the settlement of the pipe?
RE: Bent Pipelines
Remember - More details = better answers
Also: If you get a response it's polite to respond to it.
RE: Bent Pipelines
I can see the reason the designer used the pipe supports. A better solution would be to install the pipe above grade on the supports.
RE: Bent Pipelines
Remember - More details = better answers
Also: If you get a response it's polite to respond to it.
RE: Bent Pipelines
In any case, pipe have of course been supported on intervals for hundreds of years, more commonly aboveground, but also in many cases over the years in unstable soil masses as well. However, the pipe, as well as contact points and the support itself, must be sufficiently strong to take the loads imposed (as well as stiff enough to prevent objectionable deformation). High localized stresses and/or deformation can occur at supports when thin or weak pipes are supported at relatively unyielding, flat supports (i.e. where the contact area may be near infinitely/VERY small) and particularly underground.
I believe all manners of ground and pipeline behaviors, in addition to vertical settlement of the soil and pipe mass (or flotation, if e.g. a line lighter in bulk density than that of the OP is strapped down in liquefaction incidents) of the pipe and soil mass, can contribute to localized loads/reactions. Examples of perhaps further complicating behaviors would be the effects of e.g. at least slight Bourdon (extension) movements of the piping due to thrust at bends near the supports, as well as thermal expansion/contraction etc. In the case of the figure configuration, it would appear even slight thermal expansion and outward and downward thrust forces due to internal pressure would (pivot the riser to a some longer vertical elevation difference and) perhaps press the pipe barrel even more tightly against the nearest support plate as the upbend at the bottom tried to move downward? [Also, what is the span of the piping both between the supports or guides as well as outside of the two supports depicted, that were to "limit the settling" in the muskeg, is there any traffic over the area, and could/do ice lenses form over the line between same and the surface in the winter?]
I think some background of and reference to pipe on supports and research involving same, particularly to calculate maximum localized stress once reaction loads are defined, can be found in many past threads, e.g. http://www.eng-tips.com/viewthread.cfm?qid=394902.