Fracture Control

From AS2885 INFO
Jump to navigation Jump to search

Overview

For detailed explanation of fracture control, see the EPCRC Fracture Control Code of Practice, via a link to that report at the bottom of the page. This CoP is only available to members of the EPCRC/FFCRC.


See also APGA Webinar 3 - Fracture Control: (Feb. 3, 2021), presented by Nick Kastelein.


Threshold stress is a common term and is explained in the Code of Practice.



Questions from Webinar, 3 Feb 2021

Is the requirement for a fracture control plan mandatory for an existing pipeline operating under AS 2885.3?

AS 2885 in general is not retrospective - see Clause 1.6.1 of Part 0. Hence the fracture control requirements in AS 2885.1-2018 do not apply to pipelines built earlier than that. However the forthcoming revision of Part 3 will require a fracture resistance assessment. Among other things that includes an assessment of risks using the SMS process of Part 6, and further actions to manage risk may arise from that. (See also a similar question and response under the heading Questions From March 2019 Launch Seminar below.)


When the standard says <85 MPa stress does it mean hoop stress?

Yes. This is explicit in most references to the 85 MPa threshold stress (e.g. Fig 5.3.2) except in Appendix C, C3.3.2 where it is only implied.


Where is "Stable liquid" defined in the standard?

"Stable liquid" is not a defined term in the Standard (perhaps should be). However it is explained in the Note 1 to Cl 5.3.1: "Stable liquids have no significant vapour phase at atmospheric pressure, e.g. distillate or processed crude (not wellhead fluids)."


When should crack arrestors be considered?

It seems reasonable to expect that crack arrestors for a new pipeline will be necessary only as a last resort when the steel toughness required to provide adequate fracture control cannot be practically achieved by pipe manufacturers. It is also conceivable (hypothetical) that retrofitting crack arrestors to an existing pipeline may be worth considering if the fracture resistance assessment (anticipated requirement of the forthcoming revision of Part 3) indicates a high risk and other risk management options are limited.


What is the Code approach on using A106 material where low stress and not impact tested?

ASTM A106 can be used whenever the code doesn’t require impact testing, which is described by the exemptions in clause 5.3.4.2(b) and 5.3.4.3(a). This covers DN100 and smaller pipe at low stress, and any pipe carrying stable liquids above 0°C. When toughness testing is required, it can still be used if supplementary testing is obtained. Note, however, that it is not uncommon for ASTM A106 to have less than 27 J of toughness at 0°C and so it should not be assumed that it will pass.


What is meaning of achieving 80 % of CDL, how it is related to Min toughness 27 J?

Critical defect length increases with toughness, approaching a maximum achievable value called the ‘high toughness CDL’. It is recommended to have sufficient toughness to achieve more than 80% of the high-toughness CDL. This is a pragmatic recommendation – an increase in toughness provides diminishing returns above this point. However, it is only a recommendation, not mandated, because there can be a range of other considerations that are more critical and relate directly to risk. This requirement does not relate to the 27 J threshold; for some pipelines it will need less than 27J and for some it will require more.


For brittle fracture control, there used to be the allowance to establish FATT via Charpy at 30°C below the min design temp in AS 2885.1-2012. Why was this removed?

This was removed because it was logically flawed. The difference between the DWTT and Charpy fracture transition temperatures is generally a minimum of 30°C, but is often larger, even up to 100°C. Consequently, it is logical to state that the Charpy transition will be at least 30°C below the DWTT transition. It is NOT valid, however, to state that the DWTT transition will be no more than 30°C above the Charpy transition.


Does spiral seam weld support provide a means for fracture arrest?

No. The spiral weld is unlikely to arrest a propagating fracture, just as the girth welds are not relied on to arrest a propagating fracture. Unless experimental evidence shows otherwise, the spiral weld would not be expected to assist.



Minimum Toughness for Hot Tap Fittings

Reference Clause:  5.3.3 Minimum fracture toughness

A query was raised regarding the application of clause 5.3.3.2 - minimum toughness requirements for hot tap fittings. Does only the run cylinder of the hot tap fitting require the minimum toughness of Clause 5.3.3.1, since it is welded to the pipeline, or must the branch and flange materials also meet the toughness requirements?


Clause 5.3.3 (and subclauses) specifies a minimum level of toughness for steel materials. The minimum toughness is required for three categories:

  • Mainline pipe (clause 5.3.3.1) —mainline pipe is a defined term. The mainline pipe is required to have minimum toughness of 27 J, (or 40 J if measured in the transverse direction). There are several reasons for this requirement. Firstly, it will achieve a minimum steel quality. Secondly, this level of toughness is assumed for some other clauses of the standard, such as defect tolerance in AS 2885.2 and some of the exemptions in Clause 5.3.4.
  • Pipeline Assemblies that are designed under AS 2885.1 (clause 5.3.3.2) – pipeline assemblies can be designed in their entirety to AS 2885.1 OR they can be designed to ASME B31.3 or AS 4041. In the latter cases, the requirements for those alternate standards will determine the required toughness, but in the former case, the minimum toughness of 27 J must be achieved. Many common materials are not tested to 27 J, but are tested to a lower specified toughness at lower temperatures. A basis for converting toughness between different temperatures is provided so that such materials (e.g. ASTM A333 Gr 6) can be used.
  • Components integrally welded to the mainline pipe (clause 5.3.3.2) – even within pipeline assemblies that are designed to other standards (ASME B31.3 or AS 4041), if there is a component that is directly welded onto the mainline pipe, it is required to achieve the minimum toughness. An example would be a hot-tap branch fitting or a pipe pup on a weld-end valve. The rationale behind this requirement is to provide the minimum toughness where there will be welds that are qualified to AS 2885.2.



Minimum Toughness for Valve Bodies

Reference Clause:  5.3.3 Minimum fracture toughness

Please clarify requirements for Charpy testing of valve bodies in pipeline assemblies to AS 2885.1. At face value Clause 5.3.3.2 requires Charpy testing as per Clause 5.3.3.1 (27 J at 0 deg C). This requirement does not appear to differ between butt weld or flanged valves.

The intent is that any component that is welded to the line-pipe shall have the minimum toughness, in order to be suitable for a weld completed to AS 2885.2. If a component is not welded to the line-pipe then the user needs to consider what sort of component it is and what standard it is being designed to.

If the valve is flanged then its design is governed by the valve standard (e.g. API 6D). You should refer to the valve design code, if it has temperature rating guidance, or the valve manufacturer’s stated temperature rating. If that won’t work then nominate ASME B31.3 as the design code for the valve assembly (which AS 2885 allows you to do) and base the material selection and toughness requirements on that standard. Generally the outcome will be that toughness testing is not required unless the valve gets very cold (e.g. flow control valves for commissioning, venting or repressurising). Specifically, API 6D requires impact testing only for temperatures below -29ºC.

Here is a summary of where to find the component toughness requirement for different situations:

  • Welded fitting on the mainline - AS 2885.1
  • Welded fitting in an assembly designed to AS 2885 (e.g. reducer in a pig launcher) - AS 2885.1
  • Welded fitting in an assembly designed to ASME B31.3 - ASME B31.3
  • Welded valve at the boundary of an assembly designed to ASME B31.3, welded onto the mainline pipe - AS 2885.1 and ASME B31.3
  • Flanged valve designed to API 6D - API 6D if it specifies toughness, or manufacturer’s stated temperature rating, or designate as an ASME B31.3 assembly and follow that standard

(Peter Tuft with advice from Nick Kastelein)


Questions From March 2019 Launch Seminar

Should supply disruption be considered when developing a fracture control plan? E.g. time to repair long fracture in R1 may mean shorter arrest required?

Definitely. There is a subtle hint about this in Table 5.3.2 which specifies required fracture arrest lengths. For R1 the arrest length is "5 pipes unless otherwise justified in the SMS". It is conceivable that propagating fracture in a remote area may have negligible safety consequences but major supply consequences. The required arrest length should be considered through the SMS process by looking at both safety and supply impacts and the implications of time to repair.

This is also touched on in Part 3 (as at Feb 2020 in public comment draft) which requires fracture resistance assessment for existing pipelines that do not have a fracture control plan complying with AS 2885.1(2007) or later. A note mentions that in a remote area (no safety consequences) adequate mitigation for long-running fracture may be to ensure there is sufficient stock of spare pipe to allow rapid restoration of supply.


2012 version did not require minimum toughness below 85MPa. Is the addition of the 27J min below 85MPa for good practice or did new evidence prompt this change?

85 MPa was used in the 2012 revision as the threshold requirement for any fracture propagation control. This value was most likely used because it is about 40% of the SMYS of Grade A pipe – the lowest grade used under the standard. The justification of exemptions were revisited for the 2018 revision. The requirements were simplified, but they were also made more conservative. The basis for the new requirements is to simultaneously meet two conditions: (1) operating below 40% of SMYS, which protects from ductile fracture propagation, and (2) operating below the threshold stress for brittle fracture, which is calculated by comparing strain energy to the energy required to create new fracture surfaces.


Is performing DWTT on small diameter pipelines such as DN150 going to be an issue?

There is potential difficulty applying the standard in this range.

One option is to have an operating stress below 85 MPa (which is frequently achievable for such small diameters). An option for DN 150 pipe is also to conduct testing on the strip or plate from which the pipe is manufactured. Two other options that could meet the requirements of the standard, but are less likely to be practical, are to: (A) accept the risk of brittle fracture propagation by subjecting it to review in the SMS, which is acceptable for R1 locations only, or (B) meet the fracture length limits through the use of crack arrestors.

No other options are provided in the Standard itself. However, the EPCRC Code of Practice for Fracture Control of Australian Pipelines provides formulae for calculating the threshold stress for brittle fracture control. The calculation should use the lower-shelf Charpy Energy (conservatively assume this is less than 5% of the upper shelf energy, if lower-shelf data is not available). If the pipeline operates below the threshold stress, then propagating brittle fracture may be considered to be controlled.


Why contrary to CL 3.3.2, AS 2885.1 is less conservative than API 5L regarding CVN on the weld seam? API 5L requires 27J on the weld seam for pipes less than 56” diameter.

This reference is incorrect, so it is unclear where the standard expresses this.

The standard will require seam weld toughness only for the purposes of fracture initiation control, not propagation control. This is because the construction method already requires that the seam welds of adjacent pipes should not be aligned, and hence a propagating fracture that initiates in a weld seam cannot propagate more than one pipe length, regardless of the seam toughness.

Propagation control is based on the needs of the safety management study (SMS), and the “no rupture” requirement of Clause 4.9.2, which must be high enough to provide a safety margin of 1.5 between the Critical Defect Length and the largest defect that can be created by the largest credible external interference threat.


In relation to a comment relating to weld-seams having lesser toughness, would seamless pipe provide better arrest of propagation?

There’s no particular reason why seamless pipe would be superior for fracture arrest. Generally, other factors will be more significant for selecting seamless pipe over welded (for instance, the manufacturing tolerance).


Is a retrospective fracture control plan required for all pipelines designed before 2007 in the new revision of AS2885?

This will be addressed in Part 3 of the standard, currently under revision (as at April 2021). It is expected that pipelines that were constructed before the fracture requirements were introduced in 2007 will require an assessment of fracture resistance. However, they are not required to meet the current requirements of AS 2885.1 which is a design standard intended for new pipelines. Rather the compliance gaps are expected to be subject to Safety Management Study review using the methods of AS 2885.6. Guidance for retrospective reviews is also provided in the EPCRC report Code of Practice for Fracture Control of Australian Pipelines.


Retrospective fracture control: Is PRCI L51691 an acceptable tool for brittle & ductile fracture for lines that historically didn’t need fracture plans (liquids)?

PRCI L51691 is a very good reference for fracture control methods and formulas. However, this document is intended for gas pipelines (per the title) and so liquid lines are one application where it will be excessive. The current requirements of AS 2885.1 are minimal for liquid pipelines and generally for these lines controlling fracture is achieved easily.


Please explain why a DN600 pipe operating at a hoop stress of less than 85MPa will require brittle fracture to be controlled?

Calculation of the threshold stress for brittle fracture indicates that the threshold stress for brittle fracture may be higher than 85 MPa, for a pipe that is DN600 and has the minimum upper-shelf toughness of 27 J. Consequently, DN600 was set as the cut-off for exemption from brittle fracture control. The limit is not onerous, as pipe larger than DN600 will typically have higher stress and be provided with Drop-Weight Tear Testing anyway.


What prompted the change from 100J to 150J as the cut-off for the requirement on expert validation of the fracture control plans?

Sufficient evidence was available to the Standard committee that the codified requirements would provide safe solutions up to 150 J.