Evaluating small buckles, ripples, wrinkles (B31.4 versus CSA Z662)

[auzie5]

CSA Z662-15 Clause 10.10.8.3 provides equations for evaluating small buckles, ripples, wrinkles. It requires that the user input the crest-to-trough height and the “maximum operating stress”.

Should the “maximum operating stress” used to evaluate a buckle be:

a) Hoop Stress
b) Longitudinal compressive stress
c) Combined stress (Hoop & Longitudinal)

I thought that "combined stress" was the obvious answer since it accounted for axial loading (which seems like an obvious thing to be concerned with when evaluating a buckle).

However when I checked against B31.4 Para 451.6.2.8, I noticed that B31.4 explicitly states that the “maximum operating hoops stress” be used.

The formulas are exactly the same in CSA Z662-15 versus B31.4 as follows:

CSA Z662-15 / B31.4:

For pipe with “maximum operating stress (S)” between 20 ksi and 30 ksi:

A feature (ripples, wrinkles, and buckles) are considered to be a defect when the crest-to-trough height results in a ratio to outside diameter greater than

h/d < 0.01 x [(30 – S)/10 + 1]

So why is longitudinal stress not considered in the equation? For a pipeline operating at:

• Operating Pressure = 9,930 kPa
• Operating Temperature = 85 deg C
• Installation Temperature = 15 deg C

The resulting stresses are:

• Hoop Stress = 212.4 MPa
• Longitudinal Compressive Stress = -110.16 MPa
• Combined Stress = 322.56 MPa

Can we really only use hoop stress for the formula noted above? How can we not take longitudinal compressive stress into consideration when evaluating a buckle feature (seems counterintuitive). Or maybe B31.4 never realized we would be designing lines to operate at 85 deg C where the lines are subjected to a significant compressive stress?

I trust I am missing something here since B31.4 is usually a reliable source to clarify confusion found in CSA Z662.

 

[BigInch]

After a buckle, it is quite possible that the displacement has allowed most all of the longitudinal forces to have relieved themselves entirely, hence only hoop stress is probably remaining anyway.

 

[auzie5]

Very much appreciate the reply BigInch.

I agree. A buckle that is discovered during operations will not likely see a significant increase in compressive axial stress. However, consider when minor wrinkles/buckles are found during installation of a new pipeline (ex: from bending during lowering, field bending, etc.). For example, a buckle with a crest-to-trough height of 3.5mm.

I'm not as familiar with B31.4 as I am with CSA Z662, but the installation requirements in CSA Z662 apply the same evaluation formula to accept features found during new construction versus operations. For hot pipelines, it is concerning that a feature discovered during installation would only use hoop stress for evaluation when it has yet to see the axial loading associated with the large temperature change from locking in the pipeline at ambient temperature (ex: 15 deg C) to operating the pipeline at 85 deg C.

I have been told that B31.4 also uses the same acceptance criteria for features found during construction versus operations (i.e. uses hoop stress only).

As always, I would appreciate hearing your thoughts.

Thanks in advance for anyone interested in commenting.

 

[BigInch]

The highest stress occurs just before buckling. Once buckled, any compressive load would quickly be dissipated by doing just a little more work on the already buckled area. Force to Displacement ratio (stress strain curve) after a buckle is relatively flat compared to before buckling.

 

[auzie5]

I had more time to investigate the problem. Here are some of my key takeaways (below).

•    The formula was established based on fatigue testing related to hoop stress variations which is why the evaluation formula that appears in B31.4 and CSA Z662-15 is based on hoop stress not combined stress. There is no evidence that using combined stress in the formula is an effective way to address large compressive stresses associated with large differential temperatures and/or high axial loading.

•    CSA Z662-15 and IPC2002-27124 both note that more restrictive limits than those calculated using the evaluation formula can apply to hot oil pipelines. No guidance is provided on how to establish more restrictive limits. Using combined stress in the formula may be one method to establish more restrictive limits but that is speculative.

•    In any case, hoop stress alone cannot be used in the evaluation formula as a “go / no-go” test for hot oil pipelines when evaluating ripples, wrinkles, buckles. Therefore consider cut-out features and replace with sound pipe or repair.

•    A composite repair sleeve (ex: Clockspring) will restrict the feature from growing outwards but in turn the feature may simply push inwards.

•    A recommendation for a repair (if one is required) would be a steel sleeve with an epoxy fill. It will not stop the pipe from buckling inwards; however, it will apply high hoop stiffness and it will help arrest any change in ovality (i.e. if you imagine the wrinkle starts to grow inwards, the sides of the pipe away from the wrinkle will need to push outwards). Provided the length of the sleeve is at least 3 x pipe diameter and surface preparation is good, the transfer of axial load from the pipe to the sleeve should be good.

•    A composite sleeve on the other hand has lower stiffness, and will allow a certain amount of deflection.


•    Evaluation formula that appears in both B31.4 and CSA Z662-15 is based on hoop stress not combined stress. This formula is a qualitative test based on operating experience and failures of ripples, wrinkles, and buckles. The formula is based on hoop stress since for typical pipelines with these features, the failure mechanism is fatigue related to hoop stress variations.

•    Hot oil pipelines are not typical cases. For these applications we have additional concerns with high axial compressive stresses or fatigue related to axial stress variation from thermal stressing. CSA Z662-15 notes that more restrictive limits than those calculated using the evaluation formula in B31.4 and CSA Z662-15 can apply to pipelines with large differential temperatures and/or high axial loadings.

 

[auzie5]

• Need to establish exactly where these features lie along the ROW. This will allow one to determine if the feature can be daylit for non-destructive inspection. Excavate as short a section as possible so as not to reduce the constraint on the pipeline at the feature location. This will also help one to evaluate the environment surrounding these feature in case there is a need to adjust the emergency response plan accordingly.

• If the feature cannot be removed, should investigate the lowest pressure to operate line for the first few months after construction so there is time to measure and record how the features have responded to the increase in temperature. If that is the approach, schedule a date to complete a second tool run to measure and record any change in feature dimensions in response to the large differential temperature from lock-in to operating conditions.

• The wrinkle will likely get worse. The question is whether or not the wrinkle will grow to the point where it will tighten and result in a weep from a crack forming in the surface (ID or OD) or if it will simply flex a little bit within acceptable limits.