Underground Pipeline Expansion 3

[Inchtain]

Dear Engineers,

One of the recent issues that I faced during my last work trip is the bending up of a flow line that connects an oil well to a manifold of a length of about 1.7 km.

The working parameters are:

Temperature: 70 degrees Celsius.
Pressure: 720 psi
Diameter: 8 inch
Thickness: 8.18 mm
Fluid: Crude Oil with 11% water cut
Material: API 5L Gr. X52
Depth of the Trench: 1.6 m
Joints were fully gamma rayed.

I've attached the photos for your reference.

Could you possibly help in identifying the reasons? and how to avoid re-occurrence?

Is it recommended to replace the affected portion? or keep monitoring it?

Thank you very much.

 

[LittleInch]

Great photos!

That looks like a relatively rare example of lateral expansion coming from axial expansion of the long line to the left of the second photo.

70 C isn't that high but in sand is enough to generate some large forces.

This force results in movement which is concentrated at the bend and in cohesionless soils like sand the pipe can expand upwards and sideways if its not buried deep enough.

Then you might have experienced "ratcheting" of the pipe where if the flowline stops and cools down, the sand backfills under the pipe and it doesn't return to its original position. Then the next time it flows and heats up it pushes upwards and sideways some more as this is now the weak point in the system.

Some things to check.

1) was it really buried 1.6m deep? - Do a couple of trial holes on some other sections
2) Was it really operated at 70C or something higher?
3) Was it installed on a cold day? Or backfilled first thing in the morning (if anyone can remember)
4) How many times in its lifetime has is started and stopped?


Avoid re occurence? Either put in a more gentle bend or bury the bend deeper (2m cover maybe). Sometimes people build in an underground chamber to allow the pipe to move without touching the sand.

Lots of ways to avoid it, but get your data correct first.

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

 

[Inchtain]

Great answer LittleInch, Thanks

 

[1503-44]

Inch... (another "inch"??) I'm having a hard time seeing what happened there. It kind of looks like the pipe lifted up out of the ground and fell over. Or did it just plow sideways? How did the marker sign fall over? You may have lost some of the original sand cover due to wind blown loss.

Sand allows a lot of movement and 70C of that in the wrong place can do a lot of damage. You are lucky that it moved sideways, up and out and did not try to plow into the manifold, which it might have severely damaged if you had the wrong tie-in configuration.

I think we need to consider rethinking how that tie-in should approach the manifold, if you want to stop this from happening again. Some kind of controlled flexibility is needed, or you may have to provide some "drag anchors" or even a hard anchor, or cover the pipe with a protection that will not be removed by the wind.

I'd like to know more about what that looked like before it moved. Can you provide an as-built sketch of the plan view of that connection? Include 500m of flow line if you can.

 

[saplanti]

Is the pipeline raute straight? No change of direction?
What was the installation temperature?

As said the pipeline had a larger thermal differential than it can handle, and the pipe came out of the soil in the weakest location of the cover. I guess virtual anchors are in a long distance, the soil does not have resistance against thermal expansion.

I suggest you check the stress analysis if available with the soil factors used against the soil that you have. The entire pipe should be checked with the both ends’ connectivity, including their movements. In case the pipeline is straight the analysis will show this motion in some degree between virtual anchors.

Sometimes engineers do ignore/do not pay attention the pipe stress analysis since the temperature deviation looks small, but they forget the soil behaviour unfortunately when they make this decision.

As it is, it has already changed shape plastically. Check the thickness at elbow shaped location, especially outside the bend. Check the available thickness, and compare with extrados required thickness. I trust you are not running the line with the full design pressure. Check it. If you do the bend thickness might be critical. In case the thickness is less than required you might need to alter the pipe in this localised uncovered area. Otherwise, if the external temperature is not a problem and uncovered portion does no cause a safety problem you may leave as it is but you need to monitor the movement for further evaluation.

 

[LittleInch]

It's certainly moved more than I thought it would, but appears to be buried too shallow. The bend is actually not as sharp as the first photo makes it look. If you don't bury pipe bends deep enough they can have a high vertical force on thermal expansion which results in them coming out of the ground. The sand shape to me says the pipe cam out at a shallow angle as a near horizontal buckle. Normally in sand like that you get a vertical buckle, but it is what it is.

I think that is a pre formed bend, but without an alignment sheet we can't tell - Inchtain, can you supply a route drawing of the entire flowline please?

If the pipeline was straight before but is now as bent as this then you were very lucky it didn't buckle. Can't see it myself, but you never know.

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

 

[1503-44]

It would actually difficult to get that much bend without a buckle, so I assumed its a bend, but pictures don't always tell the whole truth.

 

[1503-44]

It may also be necessary to avoid overheads. Keep the pipe nice and level there as at 70C overbends can be initiated by as little as a 6-7° vertically sloped overbend.

 

[HTURKAK]

i= Apparently, The reason is poor design .. the designer did not take into consideration the temperature and pressure variations .. just write ( surface laid pipelines ) and search the web.. In KSA, ( could be more stringent conditions) similar pipelines are laid on the surface of the loose desert sand without any restraints. If buried, the pl should be restrained with someway ..( deeper trench, thrust blocks etc..)
ii= In order to get valuable recommendation , pls provide more info.( pl setting out ( plan showing the bends and bend angles ..), pressure variations, temperature variation ( daily, seasonal, steady state temperature of the content etc..)

PS. I guess you know what to do if your car starts skating and sinks in the loose sand.. drive the car back and forth.. at every step some amount of sand fill under the tire and eventually the car will get free..

 

[Inchtain]

The oil well was shut down and after 3 days of its operation the pipe experienced the shown deformation. The sign post was approximately in the middle of the affected area and it was lifted up by the pipe.

I agree with you with regards to sand back fill erosion due to badly 2-day sand storms.

I forgot to inform you that about 30 meters of the pipe were replaced during day when the weather temperature was around 33 degrees Celsius, and a portion of the new pipe is our case.

The pipe has a slight bend at this location.

The flow line is block anchored at the well and at the manifold and no stress analysis was performed.

As long as the pipe is new, can the occurred bend affect the welding joints?

By the way, it was very good example Mr. HTURKAK

Regards,

 

[1503-44]

Hard to tell if the welds are ok. I can't see anything there.
So is that bend an elbow, or a bent pipe?

I guess the tie-in to the manifold is OK. The anchor stoppd the damage there but forced pipe expansion and uplift to occur at the pictured location.

Without knowing further details, I'll suggest that you do this,
Get the pipe back down on the ground as flat as possible.
Cover it with 2 meters of sand berm, using a 1 vertical to 2 horizontal slope on each side of the pipe.
Place a rock covering of some type on top of the berm. I would suggest a gabion mattresses.

 

[Inchtain]

Good suggestions there.

It is a bent pipe.

Can you elaborate more on how to get 1 vertical to 2 horizontal slope?

Thanks

 

[1503-44]

Slope of the berm 1m vertical rise in every 2m of horizontal distance.
A berm 2m high over the centerline of the pipeline would have a slope extending 4m to each side of the pipeline.