Residual pipe stresses due to a pipeline installed in an open trench 1

[SKD_BW]

Hi all,

I recently submitted a surcharge load assessment for a high pressure gas main and although the assessment was approved by by the asset owner, I was informed by there lead engineer that the assessment did not make allowances for the residual stresses caused due to the pipe being installed in an open trench. He remarked that the stresses where only minor (approx 8MPa in this case), and would not cause the pipe to be over utilised so he was happy to approve the design submission.

I am now looking at another HP gas crossing which has a higher utilisation, hence I want to include this as part of the Von Mises stress calculation to ensure it is not over utilised, however I am having trouble following the direction given in the specification and I was wondering if anyone had any experience with this? The specification used thus far was NEN3650-1:2003 (Requirements for pipeline systems - Part 1). The spec deals with this specifically in section D3.4 - "stresses due to external load with pipelines installed in a trench". The specification states that stresses due to settlement differences should be taken into account. Can anyone elaborate on this?

I am assuming this is due to the fact that the pipe could be laid on rock, which would then cause differential settlement in the pipe and consequently residual stresses. I am also assuming that there will be some reduction in horizontal pipe support with the pipe being installed in a trench (as opposed to trenchless installation), which could also cause residual stresses. Are either of these assumptions valid, and if so how do I calculate the resultant stresses?

Apologies for rambling and thanks in advance for your help, this has been puzzling me for a couple of weeks now and I am sure there is something glaringly obvious that I am missing but I can't for the life of me get to the bottom of it. I am happy to read any book / specification on the subject if anyone can recommend a good source on the topic.

Much appreciated!!

 

[LittleInch]

The NEN code is designed for the Netherlands (and Israel for strange historical reasons) and due to the particular ground condition encountered there has quite a lot to say about ground movements.

In the Netherlands you often get significant settlement of fields relative to "hard" points such as roads, railways, berms for the zillion drainage channels that they have and hence the code is quite hot on that subject.

The issue is that in most cases you simply have no reliable data bout any sub terraenean ground movement. In theory, the pipeline is lad flat or level with no residual stress. Only if the pipe is elastically laid over a change in elevation or a change has been detected either by geo pig or measurement or digging the bloody thing up and somehow comparing this to the as laid condition will you get any useful data.

I have no idea where 8MPa came from.

differential settlement due to a high surface load is a big issue as pipeline don't like shear or point loads. There are many other ways to calculate the impact of superimposed loads, but generally, so long as you are > 1.2m deep it shouldn't be an issue.


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

 

[SKD_BW]

Firstly thanks for the response.

I used NEN3650 as it was referenced in the National Grid Specification (T/SP/GM/1), primarily as I found the NG spec lacked the required level of detail. I did think it strange that a UK spec refereed me to a Dutch Standard. Do you recommend any other specifications which you think will be more applicable / useful?

The pipelines I am required to assess tend to be in operation (primarily in the UK), and and in most cases I struggle to find any useful as-laid information. Hence, making an assessment based on actual pipe settlements or deviations will not be possible. A geo-pig would be useful and I have used these in the past when installing offshore pipes, but for this application, trial holes are probably the most extensive explanatory inspection that we will be able to conduct, and these will usually be conducted at construction stage, and the assessments are required prior to the commencements of any works and this will not account for settlement due to future surcharge loads (which I need to account for).

The pipelines do tend to be quite shallow (1 to 1.5m GBL),but I think you are right and issue of differential settlement must be what there engineer was eluding to. I think I will try and calculate the maximum deflection expected due to surcharge loading and ground settlement, and then assume that one end of the pipe is fixed (due to existence of rock for instance) and use this differential settlement to estimate some residual pipe stress - does this sound a sensible approach?

 

[LittleInch]

What are you trying to do exactly?

Sounds like an assessment for dumping loads more soil on top of the pipeline? Like you're extending a motorway embankment or something like it?

The NEN code is one of the better ones at providing good design methods for soil stresses and loads due to their strange settlement issues.

The issue is mainly about differential settlement and that needs good geotech assessment to work out if the pipeline will actually sink a little bit due to having a load more earth or other load dumped on top of it. Then its more a matter of some sort of FEA analysis nowadays to see what happens, but it is not a straightforward calculation if indeed there is settlement and defining the boundary between settlement and non settlement is difficult. It gets much worse if there is a sleeve involved as these tend to be quite rigid elements and can induce significant shear and bending loads if there is differential settlement.

1 to 1.5m cover to me is "normal" not shallow....



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

 

[SKD_BW]

The assessments are exactly that; surcharge load calculations for highways diversions, car parks, private roads and temporary construction haul roads, assessing the surcharge load from the road cross-section and vehicle loadings.

I have had difficulty with modelling pipe-soil interaction with FEA in the past (again for offshore applications), so I try and avoid this as it is out of my capabilities if I am honest, hence I try and use hand calculations where possible, but as you say this may be too complicated in this instance.

Overall, I am reasonably happy with the assessment, this is the only comment I have received from the asset owner's engineer so if I can't find a way of assessing these residual stresses I may have to go back to there engineer and ask for guidance. I wanted to make sure I wasn't missing anything glaringly obvious but my the sounds of it I don't think I am?

 

[dhengr]

SKD_BW:
Given what you are indicating is your situation and the pipelines you are evaluating, maybe the best thing you can do is some generalized study and understanding of the various possible stressers. For example…, given several different possible max. truck axle loads as surcharges on the pipeline; you have pipe dia., thickness, mechanical properties, etc., depth of cover over the pipe, soil properties above, around and below the pipe, compacted, loose, virgin soil, etc., which pretty much define the problem. There are some old std. longhand methods for this problem, along with highway dept. stds. and pipeline and railroad co. stds., which could be put into spreadsheet form so you can study the changes in any of the variables, and plot these, for your own edification, and for your future use. Another example…, given a pipe dia., thickness and mechanical properties, etc., look at pipe stress due to differential settlement. This is a beam on elastic foundation problem, with considerable theoretical history and longhand methods. But, only a good approximation, unless the client is willing to spend time and money on soils testing, FEA, etc. This problem too, is amenable to a spreadsheet solution. Then, you can study the changes in the significant variable; given a specific pipe, change the relative defection over a 10m, 20m, 30m length, etc., and vary the deflection 50mm, 100mm, etc.

These are highly indeterminate problems, so we really can’t give definitive, absolute answers. FEA is a wonderful analysis tool, but it is only as good as your model and as good as your guesses at basic data and variables. We are talking about ranges of stress values, to be superimposed on the other pipe stresses, which are a somewhat more determinate pipe stress problem. But, you have to become more familiar with these general analysis and design methods, so you can speak with some confidence about their effect on the entire system. I’ll bet that if your report had had a short paragraph about these types of secondary stresses (I’m not real sure I would call them residual stresses, that phrase has a slightly different meaning to me) the reviewer would not have mentioned that issue either. My comments would have been general order of magnitude stresses vs. deflection over 30m length, etc. I would not be at all reluctant to ask my client, in the first place, what secondary effects do you want me to consider? This is a good time discuss the indeterminate nature of the problem, unless they will spend the time and money for a finer analysis, or unless they have a specific example problem, with good soils data, etc. You are not shirking your responsibility as an engineer, but just reminding them, if they are engineers, of the complexity of the problem, with a bunch of unknowns, which they are unwilling to pay to refine.