road crossing design

[pipeliner3000]

We have a road embankment planned to go over our pipeline which would increase depth of cover over the pipeline by approx. 8m bringing total cover to 10m. Typically we would use API 1102 but have heard this is limited to shallow depths only i.e. <3m. Is anyone aware if API 1102 is still usefull when at 10m depths..? What alternative method could be used..?

Many Thanks

 

[LittleInch]

Why don't you look it up. It works to any reasonable depth.

However usually the issue in these situations is what is the impact on ground settlement when this huge extra load compresses the ground. This can literally shear your pipeline if the ground settlement is anything close to significant and you need to do a full stress analysis and not just circumferential stress which is what 1102 looks at.

Unless you're already in bedrock, you may well need a concrete bridge erected within the embankment to prevent this compaction or divert the line.

My motto: Learn something new every day

Also: There's usually a good reason why everyone does it that way

 

[BigInch]

API 1102 is for rather shallow depths really. Not much more than 6 ft of cover or so. Traffic loads are not really applicable after that depth anyway. As L" says, you may want to assume a settlement profile and analyze for combined axial, pressure and bending stress.

Independent events are seldomly independent.

 

[pipeliner3000]

Thanks,
It looks like some of the curves are limited to 3.0m depth but these could be interpolated to lower valves for deeper pipes. 1102 doesn't appear to be limited to circumfrencial stress only. It does include an effective stress check with the longitudinal expand stress. There is no bending stress though - but in our case there is no bending.

Thanks

 

[BigInch]

You cannot extend the 1102 curves to deeper depths.

Independent events are seldomly independent.

 

[LittleInch]

Read the reply, how do you know there is no bending? If you add hundreds of tonnes of weight on top of your pipeline, it will settle in that section and then bend relative to the section where there is no settlement. This is what breaks pipelines.

Its up to you, but as the incumbent resident of the land, you can make the road builder do lots of things to protect your pipeline....

My motto: Learn something new every day

Also: There's usually a good reason why everyone does it that way

 

[pipeliner3000]

Agree, What I meant was there is no bending at the moment but this could change if settlement occurs, hence 1102 not suitable as it does not include bending. Regards the curves some are limited to 3.0m depth - why would interpolating based on curve trending for lower depth not work - obviously there would be inaccuracies but would these huge?. Are these based on empirical data?

 

[LittleInch]

Which version of 1102 have you got. I've had a look at mine which is 7th edition and can't quite see where your limit of 3 m comes from.

In terms of bending then you will need to plug this into a combined stress analysis. Also look at the shear values, especially if the embankment will be quite steep.

Usually the difficulty is getting a geo tech engineer to tell you how much the soil is going to compact.....

My motto: Learn something new every day

Also: There's usually a good reason why everyone does it that way

 

[pipeliner3000]

For Highways figure 15 & 17 limited to 3.0m max depth of curves.

 

[BigInch]

I believe it is because static transverse bending (in the hoop direction) becomes significantly higher with the increased soil load, so ovaling deformation should be checked independently.

Independent events are seldomly independent.