High pressure natural gas transmission pipelines are a vital part of Australia’s infrastructure, however, it appears that some of the engineering lessons of the past are not being followed today.
They deliver over 20 per cent of our energy requirements, and in most cases gas is the most economical form of energy for both domestic and commercial users.
We have vast reserves of gas, which is ideally suited as a transition fuel between coal and renewables and delivers a reduction of more than 50 per cent in greenhouse gas emissions.
Without these pipelines the gas cannot be transported to users, and compared to electricity transmission lines, pipelines are out of sight buried safely underground.
Although there have been several ruptures, there has never been a fatality due to an operating failure of an Australian gas transmission pipeline, unlike notable examples overseas in San Bruno, California, United States of America and Ghislenghien, Belgium.
Therefore, on record, Australian energy pipelines have been very safe, and we all hope that this will continue.
The Australian Standard for natural gas pipelines
The Australian Standard (AS) 2885 (the Standard) has been adopted by the State and Territory governments as the single and sufficient set of requirements for petroleum pipelines, and is arguably better in many key respects than equivalent Standards elsewhere.
An important feature of the Standard is that it was never the intent of the responsible committee to allow “˜grandfathering’ of safety requirements, which is to say that when standards are updated, the new requirements do not apply to existing pipelines, only the new ones.
Pipelines are required to be licensed and the licensee is accountable for the safety and integrity of the pipeline, and cannot delegate this responsibility.
The pipeline is subject to metre by metre risk assessment over the entire length, and all threats are “˜designed out’ such that risk is reduced to low or negligible, or at least to As Low as Reasonably Practicable (ALARP) providing the assessed risk is not higher than intermediate.
An important and special feature of the Standard is a requirement that in High Consequence Areas (HCA) where population density or other considerations dictate the need for increased levels of protection for the public, the pipeline must be designed such that it will not rupture i.e. that it will leak rather than break if impacted, for example, by an excavator, or if some material failure occurs.
The main factor influencing the susceptibility to rupture is the stress level, and unfortunately, the effect of mandating no-rupture retrospectively on an existing pipeline will be likely to reduce the allowed operating pressure, and that will have serious implications for the economics of the pipeline.
A useful analogy here is to think of an inflated toy balloon which will fail catastrophically if pricked by a pin, whereas a bicycle inner tube will simply leak and deflate slowly.
Pipelines in HCAs should leak not break, and thus should not allow a full-bore rupture to occur with the resulting very high radiation intensity that will occur perhaps hundreds of metres or more from a ruptured pipeline if ignition occurs.
AS 2885 clause 4.7.4 Change of Location Class requires that: “where there are changes in land use planning (or land use) along the route of existing pipelines… a safety assessment shall be undertaken and additional control measures implemented until it is demonstrated that the risk from a loss of containment involving rupture is ALARP.”
Unfortunately, these requirements around HCAs that have arisen from changed land use are not always being achieved nowadays.
Despite the intentions of the Standard and those that nurture it, the encroachment of suburban growth upon pipelines designed for rural areas has resulted in hundreds of kilometres of pipelines spread over most of our capital cities that don’t meet the no rupture and limited discharge rate requirements.
As such, these pipelines expose large numbers of the community to risks that they are blithely ignorant of, and have no say or choice over.
Thus, while the Standard requires old pipelines in HCAs to be upgraded to the same safety level as would apply to a new pipeline, in practice we now have different requirements for new pipelines from that applying to old, now encroached, pipelines.
Another matter that I find disturbing is that, contrary to clearly stated principles of AS2885, fracture control, including prevention of propagating fracture, is not a mandatory requirement for all pipelines – including pipelines constructed before current fracture control requirements were implemented.
We not only have pipelines in populated areas that can suffer full bore rupture, but it may also be possible that a fracture initiated as a rupture could run for hundreds of metres or more because of the lack of sufficient toughness to cause arrest.
It is understood that this exception to the principles will be corrected in the next revision of the Standard.
Risk Assessment practice
This is a very difficult and somewhat fraught situation.
The middle management and independent consultants that routinely get involved in risk assessments become accustomed to accepting risk rankings for old pipelines that they wouldn’t accept in new pipelines.
They find themselves constrained to produce outcomes that won’t cause the need for very expensive measures, like many kilometres of slabbing or even total replacement of the pipeline.
They are forced, possibly against their professional engineering judgement, to rely on procedural measures like patrols and dial-before-you-dig, even though those measures are fallible like all human endeavours, instead of strong physical measures like no rupture and limited release rate.
Therefore, we routinely see outcomes of risk assessments that are not always as transparently rigorous and as effective in ensuring public safety as they should be.
A recent issue paper written by Peter Tuft, Chairman of the Standards committee responsible for AS 2885.1 Design and Construction that is currently under revision makes the point well.
The issue paper notes: “The current SMS [safety management study] process appears to have been useful but has never been under the spotlight expected after a failure that has disastrous effects for the public. It has also never, as far as we know, reached a conclusion that an existing pipeline presents unacceptable risk even though there are a small number of pipelines whose failure would have very serious consequences (multiple fatalities) and attract intensely critical attention.”
It then goes on to say:
- SMS is completed by middle-level technical people in accordance with AS 2885
- High-consequence risks are inevitably shown to be at worst Intermediate and ALARP, and hence tolerable
- The person authorised to issue approvals on behalf of the licensee will see a report that shows risk to be tolerable and hence accepts it without deep consideration
- The CEO and/or Board may or may not be informed of the consequences of failure.
It is an anomaly that the Standard holds the licensee accountable for safety and integrity without the ability to delegate, and yet the system allows the Directors, who hold ultimate responsibility to be uninformed of the risks they bear.
Along with this, it is not always the case that the Board includes people that are expert in pipeline technology.
An example of this is cited by Jan Hayes and Andrew Hopkins in Nightmare Pipeline Failures (CCH Australia Ltd 2014), where they say that the Board of the Pacific Gas and Electric Company (PG&E), the operator of the pipeline that failed at San Bruno, did not include anybody with experience in “gas engineering or high pressure gas transmission”.
These matters are of course issues of Corporate Governance over which the Standard has no control, but they are issues for the pipeline industry, and for the community.
A current example
The recent rupture of the Moomba-Adelaide pipeline (MAP) Port Pirie lateral as a result of stress corrosion cracking (SCC) is a good example of the potential consequences of the currently fraught risk assessment practices.
The pipeline system in question is typically 40-50 years old, and it is out of its original design life.
It consists of hundreds of kilometres of pipeline, made up of different diameters, age, pipe manufacture, and coatings.
It is also known to have a history of SCC, which it is particularly susceptible to given that much of it was coated over-the-ditch with a tape coating.
The coating is acknowledged by the owner to be in a deteriorated condition.
Despite the diversity of the pipeline, widely differing risk, and known issues with SCC due to over-the-ditch coating, the risk of SCC was assessed as Intermediate and subsequently deemed ALARP.
The 2013 Annual Report on the Technical Regulator’s website, upon which the pipeline licence is presumed to depend amongst other things, asserts compliance with the AS 2885 risk criteria, and earlier reports say that the risk of SCC was assessed as Intermediate without a case being made.
Similarly, the risk of SCC failure is deemed ALARP without explicit reasoning.
The term “˜ALARP’ comes out of safety legislation where the onus is on the operator to present a case to the regulator that the pipeline is safe.
No such case was found in the public domain.
The fitness for purpose reports on the website do not review the design parameters and pipe material properties against current requirements.
For example, no fracture control plan information was found, and this would be a very important consideration in the event of a rupture on the main line.
The mitigation cited to support the ALARP outcome was Direct Assessment, a method of inspection involving dig-ups and inspections at sites selected on the basis of maximum expected susceptibility.
The methodology is described in a NACE International Recommended Practice RP0204, which has recently been called into question by the National Transportation Safety Board (NTSB) in North America, and which is self-described as: “complementary with other inspection methods such as in-line inspection (ILI) or hydrostatic testing and is not necessarily an alternative or replacement for these methods in all instances.”
Reference was made in the reports to ILI with Magnetic flux leakage (MFL) tools, which it is known cannot find SCC.
Other tools, for example those employing ultrasonics or the very recently proven electromagnetic acoustic transducer (EMAT) technology tools, are needed to discover SCC.
Since the case for ALARP was not made in publicly available documents, the argument that the cost of further mitigation was grossly disproportionate to the reduction of risk achieved cannot be examined.
Perhaps it was considered that the cost of ultrasonic ILI at some tens of millions including supply interruptions, say around 5-10 per cent of the asset value, does approach a disproportionate level of expenditure, but the new EMAT technology available is much cheaper and would not be a disproportionate level of expenditure.
The type of SCC that caused rupture in the Port Pirie lateral, whether high pH or near neutral (NNSCC), has not been made public.
The former is historically more likely in regions of elevated temperatures downstream of compressor stations, but NNSCC is not regarded as particularly temperature dependent.
Both forms have been found in pipelines with deteriorated over the ditch coatings.
The NACE Direct Assessment standard refers to the risk of NNSCC near the weld seam in Electric Resistance Welded pipe as used on the lateral that failed.
In any case, temperature is not the only parameter affecting high pH SCC and selecting direct assessment sites immediately downstream of compressor stations, especially in a very old pipeline with serious coating degradation, is not conservative in seeking to
On this basis, Direct Assessment cannot be relied upon alone to demonstrate ALARP, and this is graphically demonstrated, albeit with hindsight, by the rupture event.
Of course, any commission of enquiry that is set up to investigate a catastrophe will also have the same benefit of being able to apply hindsight.
It seems that the risk of rupture was not Low or Negligible, and not demonstrably and transparently ALARP.
The saving grace was that, very fortunately, the rupture occurred in a small diameter lateral instead of the main line, that no one was around at the time, and that ignition didn’t happen.
If, just as likely, the rupture had occurred on the much larger diameter main line in the region of encroachment in the populated suburbs of Adelaide, the consequences could have been catastrophic with multiple fatalities and extended loss of supply.
Change of ownership and diligence
In the June 2015 edition of Pipelines International there is a timely article by W.K. Muhlbauer entitled Due Diligence – risk costs of asset ownership, which says that “Pipeline systems are assets. Their value stems from their ability to generate revenue for their owners” and that “too often insufficient attempts are made to fully quantify the implications of acquiring an asset.”
In the July 2015 edition of The Australian Pipeliner, Max Kimber commented that we seem to forget the engineering lessons of the past, and how we still suffer welding problems on new pipelines.
The Australian Pipelines and Gas Association (APGA) has commendably introduced accredited competency standards for pipeline personnel, and maybe one day there will be sufficient competent welding personnel to fill the gap that exists in that area.
However, until this happens we will go on having problems in the quality standards of welding in new pipelines and in the due diligence processes when those pipelines change ownership.
The outcome is that there may be systemic quality shortcomings so that the pipeline may not meet the AS 2885 requirement of low or negligible risk in the event of the kind of washaways that have been seen in some of our pipelines in recent years during flood events.
The other example is the Moomba-Adelaide pipeline which changed hands only a couple of years ago.
If, as it seems, the risk of failure in that pipeline was not really ALARP, should that situation have been discovered by the purchaser’s due diligence process?
It is understood that the cost of the acquisition was $400 million.
If the cost of EMAT ILI was around $5 million, that would have been a relatively small expense and would have been something that could have been undertaken as part of due diligence.
The price of safety is eternal vigilance.
I believe that the industry practice and the technical regulation system needs to be reviewed and improved, especially in relation to the safety of the public in high consequence areas that have come about through land use changes.
If there was to be a catastrophic failure with fatalities and major economic loss there is no doubt that a judge would find serious flaws in our systems and practices.
It is very important that we make the necessary improvements without the need for a catastrophe as a catalyst.
The Australian Standard already has quite stringent requirements for change of location class in land use change.
It also has a very clear set of principles set out in AS 2885 Part 0 that require the integrity of a pipeline to be maintained throughout its life, including where changes occur that affect the integrity, and at the end of the design life, the pipeline is to be abandoned unless an approved engineering investigation determines that its continued operation is safe.
The first and most important thing that must be done is that these requirements must be thoroughly complied with.
This is not the case at the present time.
In summary, the issues requiring attention by the industry are:
- Risk assessment processes, especially in relation to the ALARP principle;
- Land use change due to residential and other development around pipelines that were built to a rural specification;
- The application of fracture control to existing pipelines, as well as new pipelines; and,
- The use of competent people in girth welding.
Finally, for want of a persuasive example, this article has focussed more than I would like upon the Moomba-Adelaide pipeline.
However, it is just one example.
There are probably other pipelines for which the details might be different but the failure consequences similar.
I would like to acknowledge the assistance and support of the following people who have helped in the preparation of this article:
Andrew Hopkins, Susan Jaques, Max Kimber, Jim McDonald, Sarah Maslen, and Peter Tuft.
The opinions expressed in this article are my own, and the acknowledgement of the people listed above does not indicate that they agree with all of the views expressed.