Interesting that AS 2885.1 explicitly states to ignore roping stresses (Clause 5.7.5) 1

[auzie5]

Pipelines are typically constructed using a combination of “roping” (grading the right of way to reduce the number of required factory and field bends) and “bending” (using field bends to accommodate undulations discovered along the right of way during construction).

Roping analysis can be performed at the design stage to define bend radius limits for roped pipe (e.g. ~1500m minimum bend radius limit for NPS 20 pipeline).

The benefits of roping include defining during the pre-installation phase the optimal pipeline profile to reduce global buckling risk, minimizing trenching cost, and enhancing construction efficiency.

Inertial measuring unit (IMU) in-line inspection (ILI) tool runs are commonly used to as-built newly constructed buried midstream pipelines. Bending strain reports generated from the IMU data can be used to identify locations along the pipeline where cold/shop bends are present. Bending strain reports generated form the IMU data can also be used to identify location along the pipeline where sections have been “roped” to match the contours of the route curves (either horizontal or vertical).

IMU data can be used to generate an as-built pipeline profile in an FEA program. Using the IMU data to identify cold/shop bends versus roped sections allows one to identify areas where no elastic bending stress remains (i.e. cold/shop bends) versus areas where some nominal elastic bending stress exists (i.e. roped sections).

Question: When performing stress analysis on the as-built pipeline system, if roped sections can be identified, should the elastic bending stress be calculated and superimposed on top of operating stresses in the FEA analysis?

HDDs are a good example where pipe is roped to match the route curve (i.e. HDD profile). Since roped pipe remains in an elastic bending state, it is common for HDDs to account for the bending stress that permanently remains and the bending stress is added to the operating stresses (gravity, pressure, thermal). However, I do not typically find that people spend the time to review the IMU data to identify every mainline location where roped pipe is present and then include the elastic bending stress calculated at those locations in the stress analysis. Typically roped sections on mainline have extremely high bend radii (~500m to 1500m) but even with a high bend radius, the elastic bending stress can make a significant difference in the stress analysis results.

I find it interesting that Australian pipeline standard AS-NZ 2885.1:2018 directs users to omit roping stresses. The "note" is particularly interesting. I was recently qualifying a fusion bond epoxy external coating and when we removed longitudinal coupons from the coated line pipe, the coupons (1" wide x 24" long) sprang up like a canoe (evidence of residual stress remaining from manufacturing). The point being that there may be other residual stresses in addition to roping stresses that are not accounted for during stress analysis (e.g. pipe manufacture and girth welding).

The following excerpt taken from Australian / New Zealand pipeline standard AS-NZ 2885.1:2018:

5.7.5 Stresses due to construction

This Standard does not limit stresses prior to hydrostatic testing. Strains, deflections and displacements shall be controlled so that—

(a)strain does not exceed 0.5% except where strain is displacement controlled, (e.g. cold field bending within an approved procedure, forming of pipe ends for mechanical jointing, weld contraction etc.); and
(b)diametral deflection does not exceed 5% of diameter.

Residual stresses left in the pipe after construction (e.g. roped bends) do not need to be considered in the calculation of operating stresses, provided the pipe has good lateral restraint (e.g. laid in soils of normal strength). Where lateral restraint is weak or absent, consideration shall be given to preventing the possibility of uncontrolled strain due to the combination of residual stresses with either hydrostatic pressure test stresses or operating stresses.

NOTE: Pipe manufacture, girth welding and pipe-laying all result in residual stresses (potentially as high as yield stress), which conventionally are neglected in pipe stress analysis because they are not associated with any failure mode; however, it is conceivable that failure by deformation or buckling during hydrostatic testing may occur in a pipe containing high longitudinal residual stress but lacking lateral restraint (or during operation if lateral restraint is removed subsequent to a successful hydrostatic test).

 

[1503-44]

Where elastic bends are used and bend radius is shown on alignment sheets, I do include the bending stress in the combined stress calculation. It is especially important for offshore construction where bending machines are not used and considerable elastic stresses may be introduced during laying by barge. When it isn't practical during the design engineering phase to check that combined stresses remain below allowables considering all the possible elastic bending stresses that might be introduced during construction, I do not allow elastic bends to remain in the pipe at final in-the-ditch position via a clause in the construction specs., where all bends shall be made using a bending machine. Otherwise I establish a minimum bend radius, note it as such on alignment sheet notes, and allow for the corresponding bending stress in the combined stress calculation when establishing the minimum required wall thickness.

Even if a standard says you can do it, it may not be in your best interest to do it. Standards and specs should always be viewed as stating the minimum requirements. In the AS case, I would interpret that clause to mean, you are allowed to ignore roping stress, which I might do if I knew that it was not going to be significant enough to push my combined stress calculation over the allowable stress. If the line failed at a point where bending stress pushed you over limit stress, could you explain to your boss, your client, the judge, or yourself, why you decided to ignore bending stress when you knew it would theoretically cause the pipe to be over-stressed.

Frankly I don't see how an AS could be written in that fashion, when you know damn well that elastic bending alone can exceed yield stress. Sounds like they are depending on the soil to keep the pipe straight enough to avoid doing anything more than ovaling and locally buckling, but that = failure in my book

 

[auzie5]

Thanks for the reply.

What I have noticed is that most ILI vendors do not report bends with high bend radii or very small bend angles in a typical ILI pipe talley or bending strain report. It is not uncommon for ILI vendors to only report bending strains above 0.2%. "Roped" sections of pipelines where bends remain in the elastic region have bending strains less than 0.2% and therefore are not typically reported by ILI vendors. ILI vendors will report whatever the customer asks for. And therefore "roped" sections are identifiable if the customer asks for a non-standard reporting criteria. But in my experience, many stress engineers do not account for "roped" elastic bending stress during FEA stress analysis since they do not typically have the necessary IMU data to identify roped bend locations. This is of course for onshore applications.

For offshore applications, I suspect it is common to have pipe layed on the seafloor with elastic bends present. Is that correct?

 

[1503-44]

Yes. That's pretty much the only practical way to do general offshore installation. But in most cases the bends are very long radius anyway. In limited cases it is possible to do a direct vertical placement of a piece of line with a fixed bend angle, say near a platform, or subsea tie-in, but its complicated. But yes, offshore lay is most easily done by keeping things in the elastic bending range

 

[auzie5]

It is also worth noting that running an IMU tool as part of an in-line inspection (ILI) survey is a relatively modern practice for many pipeline operators. Although there has always been expectations for pipeline contractors to follow design limits for "roping", it has historically been difficult to confirm, with precision, if this expectation was satisfied. We can now run IMU tools through pipelines built in the 1950s and accurately describe the current, as-built piping configuration. By requesting specific (non-standard) reporting criteria from the ILI vendor, we can now identify locations along a pipeline that are likely to be in an elastic bending state. However, I rarely see this being requested of our ILI vendors (aside from HDD locations). Moreover, I hear from many ILI vendors that this is an extremely rare request (i.e. to report locations with strain levels <<0.2% and/or bend angles <<5 deg).

With more advancements in MFL technology, perhaps we will soon be able to identify other types of residual stress that commonly exists in most pipelines (e.g. from welding or pipe manufacture). I suppose then when armed with the data, one might also include those other residual loads when performing stress analysis. We will see.

 

[1503-44]

Those residual stresses have been accounted for in the design methods and the corresponding design factors and allowable stresses. That has worked rather well so far, so I hope I won't have to know anything more about them. I'll be happy if ILI finds corrosion, service related defects and induced cracking. Corrosion is public enemy #1. Residual stresses not so much.

 

[BJI]

The answer to your question already is already given in the extract from AS2885, it is not that residual stresses have been overlooked by analysts. These are displacement controlled stresses, therefore, don't influence the failure mode you are assessing under operating conditions. It is either load controlled (plastic collapse) or displacement controlled cyclic. The displacement controlled forming strains will not influence your displacement stress range since they are not cyclic and the mean stress effect is relatively negligible. However, just like in structural analysis, residual stresses still need to be considered for buckling analysis, and there are precautions noted for piping systems without adequate lateral restraint. The other consideration where residual stresses are relevant is fracture mechanics. Given your additional questions on pipe manufacture, I would recommend the papers by Ted Anderson, that discusses these types of considerations for fracture assessment of pipelines.

http://fracturemechanics.com/downloads/