Determination of the Virtual Anchor Location. 1

[Davidkay]

Thread378-112705
With reference to the above thread, has anybody here modelled underground pipeline with Autopipe. If yes, how can the need for Anchor Block be obtained from the Autopipe simulation result? Can the location of the Vitual Anchor be determined using Autopipe. If no, how can one effectively determine the location of the virtual anchor?

 

[BigInch]

I don't know Autopipe's capability.

Virtual anchor lengths are typically assumed to be the located at a distance where the equivalent force from stopping thermal and pressure expansion equals the friction provided by the soil. It is the length obtained by taking the equivalent thermal-pressure expansion force and dividing by the soil's resisting friction per unit length of pipe.

If it ain't broke, don't fix it. If it's not safe ... make it that way.

 

[Davidkay]

Thanks BigInch for your post. I appreciate it.
I have a spreadsheet that I normally use to calculate the thermal Expansion of Pipelines. The Spreadsheet gives me the Expansion in the Hot and Cold ends and also the Length of the Virtual Anchor from the Hot(Launcher) and Cold(Receiver) ends of the Pipeline. For instance, if the Pipeline is 100km long and the Virtual Anchor occurs at say 2km from the Hot end, then the distance to the cold end from the Virtual Anchor is 98km. But 2 issues confuse me when I am carrying out a stress analysis calculation:
1. When I model my pipeline for stress analysis(Using Autopipe), I start from the Pig Launcher to the Virtual Anchor position. But for the Cold (Pig Receiver) end I don't know where to normaly end the model since the Virtual Anchor Location is normally too long(like 98km as in the pipeline above)
2. If it is true that I have to model from the Pig Receiver to the Virtual Anchor position, then my question is; why should it be so since the pipeline is assumed to be fully restained after the Virtual anchor.

Please I will appreciate every response to these questions. Thanks

 

[BigInch]

No. You should do another calculation at the "cold" end where you will have another virtual anchor and pipe to the pig receiver. On the hot end you are partially restrained from pig launcher to virtual anchor. Once you pass the virtual anchor, all straight pipe is assumed to be completely restrained until you get to the virtual anchor at the cold end, then it becomes partially restrained again to the pig receiver... except for...

You can also have unrestrained portions of pipe between the "hot and cold" or pipeline end virtual anchors which form up at horizontal bends, vertical bends, closed valves, vault penetrations, etc. where axial forces within the pipe do not line up and cancel each other as they normally do with directly opposed joints. Horizontal bend angles, overbends, underbends, anywhere that might have unbalanced axial force vectors can create movement and deflections and bending moment from those axial eccentricities. Wherever you have partially restrained locations and pipe movements, interior virtual anchors will form up on each side of the partially restrained point. You may have to analyze those areas separately.

You should not have to model fully restrained, straight sections of pipeline between virtual anchors, as they will only have axial forces, no moments, which are easily checked for the local crushing buckling mode, or uplift buckling, probably without resorting to using packaged pipe stress programs.

If it ain't broke, don't fix it. If it's not safe ... make it that way.

 

[Davidkay]

BigInchThanks once again for your Spot on reply. It answered quite some doubts.One more thing please.
If we are going to do another calculation for the Cold end, what temperatures are we going to use as the departure and arrival temperatures and Pipeline Length. The spreadsheet (here attached) that I use has these inputs.
Secondly, when a Bend or a "Tee" occurs before the Virtual Anchor position, how do we deal with it.

Thanks once again in anticipation.

 

[BigInch]

Pipeline operational outlet temperature appears to be 25C. Installation temperature appears to be at ambient I guess, so the [Δ]T is 5C.

A 12"x 23 km pipeline at these temperatures will generally be fully constrained with a virtual anchor within 200-500 m of each end, even if in poor soil, or even if laid on the mud line and friction is only between bottom of pipe and mud, at least if you have enough negative buoyancy concrete coating to provide stability.

A branch near one of the ends of the pipeline will provide some additional anchorage and shorten the usual distance from pipeline end to virtual anchor on the mainline, and more than likely put a good deal of bending at the outlet. Beware of that.

A branch off the interior of the pipeline, say 2000 meters upstream of the pipeline's end, would cause virtual anchors to form on the mainline at roughly 100 meters upstream and and another 100 meters or so downstream of the branch. (I'm just estimating that distance for a relatively cool 12" pipeline, so that 100 meters is just a hopefully good guess). If the line was straight at that location, no bending would be imparted in the mainline from misalignment of the mainline's axial loads, as the axial loads on each side of the tee would directly oppose each other and no eccentricities and additinal bending moment would be produced. If there was some deviation from straight alignment, axial force x sin of the misalignment angle would produce bending moments in the mainline, some of which will be transferred to the branch. You should try to keep branches placed on straight segments of mainline. From the tie-in run out perpendicular to the mainline for 5 to 10 meters, then turn 90 deg and run parallel to the mainline for 10 meters or so, then head off in the direction you need to go.

The effect of the branch line restrained expansion and its resulting axial compression load would also have to be considered on itself and, as that would tend to expand into the mainline; the mainline tends to act as an anchor for the branch too. This compression stress in the branch from it being constrained, or partly so, so you will find another virtual anchor 100 meters or so down the branchline. From there the branch will try to expand back into the mainline. That will introduce bending into the mainline, similar to how a point load does on a beam, except probably to a lesser extend, because the mud on the opposite side of the outlet tends to support the mainline against that point load, but none-the-less, some bending will more than likely be introduced into the mainline from the branch's axial compressive load.

If it ain't broke, don't fix it. If it's not safe ... make it that way.

 

[Davidkay]

Hello BigInch,
The Operational temperature is 30degC (which is the departure temperature), the arrival (at the Cold End) is 25C. How do we calculate the Virtual Anchor Location on the Cold End of the Pipeline. That is, do we use the Inlet (Operating) temperatures or the discharge (outlet) temperature?

 

[BigInch]

The differential temperature at the outlet = outlet temperature - installation temperature.

If it ain't broke, don't fix it. If it's not safe ... make it that way.