I’m interested in feedback from those of you involved in pipeline maintenance or corrosion management.
My experience of Australian safety management studies is that when the threat of external corrosion comes up we say something like “It’s got factory applied linepipe coating, high quality joint coating, cathodic protection, regular cathodic protection monitoring, periodic coating defect surveys and periodic in-line inspection, therefore corrosion failure has been designed out and does not need any further consideration.” Nobody ever argues much against that position.
My question for you: Is this a bit glib?
Australian pipelines have a negligible rate of corrosion failures. Apart from about 15 leaks on one pipeline that was very poorly built and very poorly managed in the 1960s, the Australian pipeline incident database contains a mere two loss-of-containment incidents caused by corrosion, and both of those were stress corrosion cracking (including the infamous 1982 rupture of the Moomba – Sydney pipeline). So that makes the “designed out” conclusion look pretty reasonable.
In contrast, pipelines in North America seem to suffer corrosion failures at a great rate. A plausible (but unverified) explanation for the difference is that Australian pipelines are younger. Apart from the fact that they have had less time to corrode, they are also less likely to suffer serious coating failure because almost all Australian pipelines have been built since the development of modern factory-applied coatings. Many North American pipelines, on the other hand, are several decades old and were coated over the ditch with limited surface preparation.
For this reason pipeline engineers in North America pay a great deal of attention to evaluating the likelihood of corrosion failure. There is a variety of techniques available including a sophisticated reliability-based analysis (RBA). RBA has been done for a few Australian pipeline segments (as part of an APIA Research and Standards Committee project, which will be reported in due course). Interestingly, that analysis showed a small but non-zero probability of corrosion leak, notwithstanding all the design and operational features that we assume eliminates the risk (see footnote).
The main reason for the non-zero leak probability is the imperfect accuracy of monitoring, mainly in-line inspection tools.
With all that as background,
- Is it reasonable to say that we’ve “designed out” the risk of corrosion failures in the more modern Australian pipelines?
- Have there been corrosion failures (leaks) that were not reported to the incident database?
- Have you had any scary experiences of finding (by chance) unexpectedly severe corrosion despite a clean bill of health from in-line inspection?
- How do you think the AS 2885 safety management process could better address the corrosion threat?
- Is the AS 2885 process a suitable approach for assessing corrosion risk?
In asking these questions I’m not trying to cast doubt on the AS 2885 process as a whole, but rather probing and testing it for any possible weaknesses that we can correct if necessary.
Let me have your thoughts.
(Footnote: The RBA calculated probability of corrosion leak might be non-zero, but given that the consequences of a small leak are minor the risk is well within the tolerable range. There is nothing here to suggest that corrosion failure of Australian pipelines presents a serious risk to the public.)
Pipelines OZ 
Hi Peter
As an ex corrosion engineer and someone who monitors US failures I think the differnce that exists comes from two areas.
First a lot of the pipelines, such as the recent San Bruno event, are (in Australian terms) very old pipelines with I suspect deteriorated coatings. I’m not sure about the level of coating or metal loss pigging inspections that occur but this may contribute to the failure rate. While our environment where most pipelines are is dryer than the US we may need to be aware of aging coatings.
I have been concerned that on pipelienmoperatro told me recently that he doesnt do coating ispections just pig runs. If they show a problem then he repairs it. It would be interesting to know what the pipeline safety managemnt study said.
The second and I think more imporatnt factor is “wet” gas. My understanding is the US has a lot of gas pipeline with higher levels of water in the gas. As a consequenmce internal corrosion failures are higher than here. I think this less imporatnt for Australia and may be the main contributor to the differnces in corrosion failures.
Regards
Adrian Amey
Adrian,
Thanks for the observations. I’ve too have heard the argument for just relying on pig runs and not doing coating defect surveys, and as it was presented to me it made sense, but then I’m not a corrosion expert. I think the reasoning was that just because there is a coating defect there isn’t necessarily significant corrosion there – in fact if the CP is doing its job there shouldn’t be. And you can get significant corrosion at shielded defects that are less likely to be picked up by the coating defect survey. What really counts, because it has potential to result in failure, is metal loss. And in-line inspection directly detects metal loss.
Peter,
“Have there been corrosion failures (leaks) that were not reported to the incident database?”
I take this to be a rhetorical question, particularly if a pipeline is defined as anything designed to AS 2885.
On a more general note, I don’t believe it is glib to say that corrosion is designed out if there is factory applied linepipe coating, high quality joint coating, cathodic protection, regular cathodic protection monitoring, periodic coating defect surveys and periodic in-line inspection. If all these are in place, the chances of a corrosion leak must be just about zero. Society should be more concerned about those existing pipelines for which these elements are not in place. This comment is probably more applicable to in-field gathering lines and minor product pipelines than the major transmission pipelines. Even the proverbial drover’s dog will be aware that corrosion leaks on older pipelines are occurring out there. I definitely wouldn’t get too complacent about us here in Australia being superior to the Americans. Their economy is much larger than ours and they are going to have more frequent incidents as a result.
Mike,
The question about unreported incidents wasn’t really rhetorical. I’m in fact not aware of corrosion leaks on older pipelines (other than production flowlines which are in a separate category).
If there really are corrosion leaks occurring but not being reported to the incident database then that’s very disappointing, given the effort that POG and APIA have put into encouraging incident reporting, including annual declarations by every POG member that they have reported all incidents (or had none). Maybe you could let me know (privately if necessary) of any corrosion leaks that you know of.
Having said all that, there was a period of several years around 2000 when incident reporting was very poor, so perhaps the incident database missed a few then.
Hi Peter,
Can you please explain what you mean when you say that production flowlines are in separate category when both transmission pipelines and production flowlines are covered by AS2885.
Hmmm, I guess I tend to focus pretty strongly on transmission pipelines, and the pipeline incident database in practice only captures incidents on such pipelines, so my observations on corrosion failures apply only to transmission lines. I should have made that clear. However I know PIRSA regulate a huge network of flowlines so understand where you’re coming from.
Peter and others,
I don’t have any real comment on what has been written (I am not in an operations or inspection role), but I feel an interesting observation is the way our coatings have progressed over the last 10 to 15 years, and how this affects the long-term integrity of pipelines – at least from an external corrosion protection perspective.
Previously we were installing pipelines with yellow jacket, along with FBE for the higher temperature services. Over the years. FBE has become more common, along with the continual development of tri-laminates, and combined coating systems. One of the reasonings for these (I have beentold) is that these more advanced systems perform better in wetr conditions when coupled to the CP system. Another reason may be the durability and longevity of the multi-layer systems.
Anyway, it does show the level of importance given to corrosion protection on our pipelines. This work is also supported by the EPCRC with its coatings and corrosion protection research initiatives.
For internal corrosion protection, we have the option of internal linings (which also lowers pressure drop). This doesn’t seem to be as popular – obviously it has a capital cost and schedule implication. It may, however, help in services where higher levels of moisture are likely.
Thanks
Mark C.
Like most others I would normally state that high quality factory coating, high quality field joint coating compatible with the factory coating, CP, regular maintenance, and periodic MFL pigging and DCVG surveys would be full mitigation from external corrosion: internal corrosion would depend on the product being transported but for sales quality gas would also be assessed as zero.
However, if this got as far as a risk assessment it would be Rare, or even Hypothetical, with a consequence rating of Minor or Trivial: in any case the risk is Negligible.
SCC is a different beast and needs to be considered separately, since SCC can exist undetected until enough little cracks join together to form a Critical Defect and the pipe ruptures. If conditions suitable for SCC exist then they have to be treated and mitigated.
Hi Peter,
PIRSA believes the AS2885 safety management study process adequately addresses the threat of corrosion, if undertaken in accordance with the code.
If all of the corrosion mitigation methods as referred to above, such as intelligent pigging, factory applied coating, etc, are applied then the likelihood of a corrosion related failure is reduced enough for the risks to fall within the tolerable range. Where we should start worrying is when not all of these are implemented (e.g. lack of adequate coating) in which case that likelihood is elevated.
In relation to your footnote, we have to be careful about the message we are sending out when we state that consequences of pinhole leaks are minor. Firstly, the severity of the consequence may not necessarily be minor as these will vary depending on the product and the surrounding land use. For example; the environmental consequence from a prolonged liquids leak (even from small hole) into a sensitive environment (e.g. water body) could be classified as severe or major, which could put it out of the tolerable risk range. It is important for operators to adequately assess failure consequences and event likelihoods to demonstrate that the risks are being adequately assessed and accepted (if tolerable), as per the AS 2885 process. Broad statements that consequences are minor and risks tolerable are misleading and inconsistent with the intent of a risk based standard such as AS2885.
Furthermore, according to AS2885, structural integrity is achieved when the pipeline is leak tight. Therefore the occurrence of pinhole leaks on the pipeline is at least by definition, an integrity failure which would ask questions of the operator’s integrity management systems. If pinhole leaks are occurring (even with minor consequences), there is nothing to suggest that integrity failures yielding more severe consequences will not occur, making the risks associated with ongoing operation intolerable. It is because of this that PIRSA classifies pinhole leaks as serious incidents regardless of their consequence. PIRSA’s response to such an incident involves probing the operator’s integrity management system to gain confidence that pipeline operation is being undertaken in accordance with the operator’s own systems and AS2885.
Michael Malavazos, Belinda Hayter, Michael Jarosz
Glad you feel that the AS 2885 approach isn’t underestimating corrosion risk.
When I said “minor” I was thinking (loosely) in terms of the AS 2885 risk matrix severity ratings. So it contrasts with the Major rating (fatalities, major offsite impact, long term effects) and Catastrophic (multiple fatalities, permanent major changes, ecosystem viability affected). “Minor” wasn’t meant to mean insignificant. However if we are using the matrix I could agree that a pinhole leak in a liquids line might be Severe rather than Minor. (Showing my gas transmission bias again, which I really need to overcome.)
I am presently employed as design engineer in Nabucco Natural Gas Pipeline Project. Many railways and highways will be perpendicularly crossed over along pipeline route. As per the regulations of the authorities responsible for railways and highways in my country, the carrying pipe is to be through a caisson. However, caisson pipes are no more in practice recently due to difficulties in cathodic protection. Pipeline comparies recommend crossings without caisson. We, too, would like to adopt crossing without using caisson pipes. However, I am having difficulty to give actual examples of worst cases of using caissons, in my efforts to persuade the relevant authorities. Which practices are adopted by the authorities responsible for highways and railways in your country? I would appreciate very much if you could provide me with any example of worst case of using caissons, locations, reasons and results thereof, and photographs if any.
Respectfully Yours,
Nil AÇIKGÖZ
Pipeline Desing Engineer
The situation in Australia is much the same as you describe – pipeline companies want to avoid cased crossings, road and railway authorities tend to think they are necessary (because they don’t understand high pressure transmission pipelines at all). There is an Australian Standard for Installation of underground utility services and pipelines within railway boundaries (AS 4799-2000). For pipelines carrying flammable and combustible fluids it says “Where a carrier pipe is a steel pipe it may be designed and installed without an encasing pipe, providing sufficient mechanical and cathodic protection are provided as approved by the [railway authority]”. That standard is fairly modern (2000). There are many older crossings that do have encasing pipes.
I can’t give any examples of casing-related failures. As noted in the main blog post, there are only two recorded corrosion failures in Australian onshore transmission pipelines and both of those were stress corrosion cracking. While I’m aware that some pipeline operators are quite concerned about possible corrosion on some of their old cased crossings, that information has to remain confidential. And in any case, it’s only a concern, not a serious problem at this stage and certainly not a failure.