Fracture control and pipeline assemblies

Good question recently from Josh Wickham of GPA.  He’s not the first to ask this and there have in fact been a number of official requests for interpretation from the Standards Australia committee ME-038 (responsible for AS 2885 ).  Josh’s question relates to design of scraper traps for which the client has already specified the materials, which happen to be linepipe for the barrels:

My query is what dictates the toughness requirement for these vessels, and does the fracture control plan (and hence the toughness requirements) then apply to this application?  AS 2885.1 is vague on this matter – clause 4.8.2 (with amendment 1) states “the fracture control plan shall only apply to the linepipe as defined in fig 4.1. It shall not apply to accessories.”  Accessories doesn’t appear to be defined – but figure 4.1 shows scraper facilities and main line valves as part of the ‘pipeline’ and therefore infers assemblies are part of the fracture control plan.

Quite right that there is some unfortunate wording in the standard here – it’s one of the things that needs attention in the next revision.

The short answer is that no, the fracture control plan is not meant to apply to pipeline assemblies.  Note that “pipeline assembly” is a defined term, although it is defined in Clause 5.9.1 rather than the Definitions section:

Pipeline assemblies are elements of a pipeline assembled from pipe complying with a nominated Standard and pressure-rated components complying with a nominated Standard or of an established design and used within the manufacturer’s pressure and temperature rating.

So that includes mainline valves and scraper station piping, and a variety of other things ranging from simple branch connections up to slug catchers (if you wish to design them to AS 2885 rather than a piping code).

Josh goes on to ask:

… I would have expected that given that the intent of fracture control design is to select sufficiently tough materials to arrest crack propagation, and the fact that the connection between the vessel and the pipeline is a flanged connection, then a crack initiating on the vessel side would be arrested by the flange? If this is the case – then should the fracture control plan really apply?

This is definitely on the right track.  Why do we need fracture control plans?  Mainly to control propagating fracture.  Clause 4.8.1 defines four pipeline failure modes:  leak (without rupture), rupture (full bore opening, but arrested), propagating brittle fracture, and propagating tearing or ductile fracture.  Just to be quite clear, propagating failures in the past have run for hundreds of metres, unzipping the pipe and completely opening up the trench over that distance.  Generally a very nasty form of failure.

Does that sound as if it’s relevant to a scraper trap?  Clearly not, nor to any other type of pipeline assembly (unless perhaps you have a really long slug catcher).

Josh continues with a third question:

If the fracture control plan does not apply, but the client would like to maintain the design temperature rating of the scraper vessel…could you then use the approach of the pressure piping standards – AS 4041 (section 2.11) or ASME B31.3 (section 323.2.2) which give guidance on impact testing requirements to meet a minimum design temperature (using graphs based on material wall thickness etc)?

At this point I’m going to stop expressing opinions because I’m not a materials engineer and will rapidly get out of my depth.  However the approach suggested by Josh sounds entirely reasonable.  Comments welcome from those with more expertise!

Advertisements
This entry was posted in Pipeline design, Standards. Bookmark the permalink.

4 Responses to Fracture control and pipeline assemblies

  1. Anonymous says:

    All good questions, and unfortunately questions that AS2885.1 leaves largely unanswered.

    Fracture control plans should state that they apply to linepipe and that other components will require fracture control by other assessment. AS2885.1 gives plenty of guidance, but requires considerably more materials knowledge than most pipeline designers possess, and I am concerned that the standard is insufficiently prescriptive to ensure that all pipeline designs are fit for service. The gurus will be on top of this, but pipelines are designed by a wide variety of people, some of whom have minimal knowledge of fracture control. The old chestnut that, “ASTM A106 Grade B is good to -29 Centigrade”, is still very much alive and well. For scraper traps, one can use a fracture initiation control approach as indicated by Equations 4.8.5(4) & (5). The Charpy test would then be done at the design minimum temperature. This is a completely different purpose than using Charpy tests for fracture propagation control, and the test temperature will likewise probably be lower. I am sceptical about adopting guidelines from other standards such as ASME B31.3 because one can always find a standard to suit, and do we really know how relevant it is unless it is being used for the total design? Stress is an important parameter in design, and the allowable stress under AS2885.1 and ASME B31.3 may well be different. Mixing and matching standards is asking for trouble. The Charpy requirement will be higher with higher strength materials, as illustrated in ASME B31.3 Table 323.3.5 and AS2885.1 Equations 4.8.5(4) & (5). However, I know from experience that the answer obtained will be different with each approach.

    I suspect the reason there haven’t been failures is that materials are remarkably robust and engineering codes have inbuilt conservatism. My opinion only, but I would like to see AS2885.1 cover the fracture control of pipeline assemblies and fittings much more explicitly to resolve such questions without relying on pipeline designers being materials engineers, which they are clearly not. I am alarmed by the poor understanding of this issue and the lack of prescripttion in the standard.

  2. Mark Coates says:

    Some additional points to note:

    Pipeline assemblies such as pig launchers and recievers are not always under pressure, and hence are not subjected to the same operational conditions as included in a fracture control plan. I would question including launchers and receivers into an operating pipeline fracture control plan.

    Vessels can be designed to standards such as AS1210 and ASME BPV Section VIII which have specific requirements for low temperature excursions – including (as stated above) the need to low temperature impact testing. Codes such as AS4041 and ASME B31.3 also have good sections on low temperaure designs. Note that these pressure piping codes cover a variety of materials, and that low temperature excursions also line up with the use of fine-grain carbon steels for low temperature service (A333). Pipelines are almost exclusively high strength steel (X70, etc) accompanied by Grade B steels for above-ground facilities. Apart from low temperature impact testing (and use of fine-grained steels), another option for low temperaure is to design at a lower design strength. This may result in larger, thicker pipe – good for fracture propogation; bad for economics (mroe expensive).

    On larger pipelines, the pipelines are separated from the above-ground assemblies by monolithic insulation joints (insulation flanges are also used, but tend the be for the smaller pipelines). These joints have enough “bulk” to act as fracture arrestors. Again, this would question including these assemblies into a pipeline fracture control plan.

    Pipeline assemblies may tend to be more shop fabricated, and hence have difference welding and inspection procedures applied (compared with those used on the ROW for pipeline construction). This has an impact on the fracture resistant properties of the assembly.

    Another option may be to have multiple fracture control plans for the different pipeline segments. This requires more work, but avoids the problems associated with a “one plan fits all” approach. This would be more beneficial to a pipeline which undergoes varying operational conditions (pressure, flowrate, gas quality) along its length.

    Thanks

    Mark C.

  3. Chris Hughes says:

    As stated in the original blog, fracture control is intended to prevent a defect from resulting in a non-stop running rupture: this is clearly demonstrated by the two-curve method where the criterion is to ensure that at any pressure the decompression velocity of the gas is greater than the crack propagation velocity, thus ensuring that the pipe hoop stress at the crack tip is always decreasing as the gas decompresses through the rupture.

    Fracture control, as defined in AS2885.1, is not meant to ensure that there are no defects in the steel which can initiate a rupture: that function is performed by the hydrotest which ensures that any defects latent in the steel will not grow to cause failure at the operating pressure. We cannot design to totally eliminate the possibility of defects being created during operation (although the SMS should minimise the likelihood). We can ensure that the failure of such a defect will be localised in effect by specifying the toughness of the material.

    Practicality of fracture control in the pipeline industry lags well behind the research. Back in the 1990’s I did a literature search on fracture control in the PRC proceeding publications, and it was well accepted that using Charpy values to determine fracture resistance was prone to large errors, particularly on high strength steels operating at high hoop stress. Crack tip opening angle (CTOA) is generally considered a much better predictor of fracture control properties, yet I have never seen this being specified nor have any of the pipe mills suggested its use. Probaly because it is a complex and expensive test compared to Charpy and DWTT.

    One factor which is easy to determine, which used to be included in every pipe specification but has disappeared from sight (I am trying to bring it back) is to perform DWTT at decreasing temperatures (well below minimum design) to determine the ductile/brittle transition temperature. If we KNOW this temperature from mill tests then all the kerfuffle about using low temperature steels for vent lines and concerns about low temperatures during blowdown and refilling can be forgotten.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s