Richard Kuprewicz's commentary on the Corrib report

Observations by Dr Richard Kuprewicz on the Advantica review into the proposed Corrib gas pipeline through Rossport, which was released by the Department of Communications, the Marine and Natural Resources on 3 May, 2006. Dr Kuprewicz was the author of the Accufacts report, published by the Centre for Public Inquiry in November 2005. (This article has not been published in the print edition of Village, but the current issue includes an overview by Harry Browne of the issues raised by Dr Kuprewicz, as well as a timeline of the Corrib gas controversy and a story by Frank Connolly into Tony O'Reilly's oil and gas millions)

This summary briefly presents several key observations comparing the recently released Advantica review and Accufacts Inc.'s ("Accufacts") earlier independent report on the proposed Corrib Pipeline. We invite the reader to evaluate these two reports, paying particular attention to those issues in which Accufacts and Advantica are in general agreement, those areas in which we are not in agreement, and those critical areas that Advantica identified were not in their work scope, but that Accufacts has addressed as system critical issues concerning the onshore pipeline. It is hoped that this information provides assistance to all parties to permit additional informed discussion and constructive dialogue concerning this project.

Appendix G of Advantica's review for the government provides a commentary on Accufacts' Corrib report. In reading this appendix a reader might wrongly conclude that Accufacts issued its independent report after Advantica's draft report was issued, and that our analysis may somehow be incomplete because we have restricted our investigation to information in the public domain. The Advantica draft report was not issued to the public, nor made available to Accufacts, until December 2005, well after Accufacts had made public its Corrib report. Accufacts is often called upon to perform critical energy system analyses, especially on sensitive pipeline systems. Our process approaches are well-known, based on tested sound engineering science and experienced management practices, and for the most part a matter of public record. Our public disclaimer is intended to serve notice that we are obligated to protect those clients where information may be of a highly sensitive nature, such as trade secrets.

Our very detailed system analyses seldom require volumes of additional information to address very specific key "system critical" questions. It is our practice that, should certain information be unavailable to complete a thorough analysis, we indicate those areas of information deficiency in our report. Clear examples illustrating this point are the identification in our Corrib report of further load analysis needed for the pipeline in peat environments, as well as the cycle frequency potentials that can be associated with third-party damage on thick-walled pipe. Accufacts does not raise these clarification issues to be argumentative, but to stress the importance of the concepts that both Accufacts and Advantica raise as independent neutral observers regarding important safety concerns on this project.

After completing a detailed review and analysis of the Advantica report, made public on 3 May, 2006, Accufacts believes three areas of concern are critical to the current onshore pipeline route proposal:

1. Pressure Control Approaches to Limit Pressure are in All Probability Not Failsafe and Need Further Definition/Analysis

Advantica's observations concerning the possible inadequacies of the original pressure limitation designs are not a surprise to Accufacts. Confusion as to a pipeline system's pressure capabilities are all too often observed by Accufacts as a serious indication of project confusion and management process breakdown on a pipeline. All pipeline operators, regardless of the pipeline system, should be able to clearly indicate and communicate the pressure limitations and the various safety system approaches utilized to ensure pipeline pressures stay within acceptable ranges to prevent failure.

Advantica has suggested that "a full and technically thorough reliability analysis be carried out for the subsea pressure control and isolation systems specifically in the field design," that pressure be limited on the onshore segment to a 0.3 design factor, and that "some form of isolation over and above that currently planned (a locally operated isolation valve) appears to be appropriate and practical." We concur with the reliability analysis, as this is one mechanism to reduce high uncertainty in system design, especially for those systems prone to operator error or remote communication breakdowns that can cause excess pressure.

We must caution, however, that Accufacts, in developing our original report, considered and rejected the potential to reduce or maintain pressures on the onshore pipeline via a pressure letdown/isolation valve at the beach/shoreline. This design scheme is intended to prevent the onshore pipe from reaching pressures in the corrosion-induced rupture pressure zone. Accufacts, as an experienced certified process hazard and operability analysis (HAZOP) team leader, performs many system reviews and analyses. We have observed that it is all to common for engineers, especially instrument engineers, to mislead themselves by citing reliability analysis for their design approaches while missing the interactions that can result and actually drive pipeline systems to pressure failure should the instruments fail. One of the purposes of a HAZOP is to screen these poor design approaches out of a project, hopefully in the early stages of design.

In areas where safety over pressure protection equipment failures can result in high numbers of casualties, ultimate overpressure design philosophy must rely on the concept that the engineered safety approach cannot fail its design intent, a scheme we call failsafe. Failsafe is a term much misunderstood and misused in the industry, and can be the subject of substantial debate. Most gas transmission pipelines incorporate failsafe overpressure protection in their designs, a concept that is much more difficult to include in this production pipeline. Even though Accufacts rejected the pressure letdown/isolation scheme in our earlier analysis, as we could not conceive a failsafe scheme, we remain open to design proposals that would ensure onshore pressures never exceed the 0.3 design factor, or 144 bar. Given the societal risks associated with the current proposed route, the burden of proof for such an over pressure protection design falls on the pipeline operator and must go beyond reliability analysis. Without such a proper demonstrated failsafe design approach, Accufacts must continue to advise relocation of the onshore pipeline away from people, especially in the area of Sruwaddacon Bay, where terrain makes escape routes for the clustered population essentially impossible in the event of a rupture. The ideal ultimate failsafe design scheme is distance (i.e., remote and/or proximity siting), which cannot always be provided. Given the wide open areas surrounding the gas processing plant, a rational remote pipeline reroute should be able to be determined that would allow the pipeline operator to take full design pressure advantage of the Corrib field and any subsequent fields developed, without incurring pipeline safety risks to local civilians. It is Accufacts' opinion that the pipeline reroute option difficulties have been substantially overstated.

2. Serious Concerns/Questions Still Remain Regarding Internal Corrosion Risks

Corrosion rates can be general or site-specific, usually called selective corrosion. Selective corrosion rates can be much higher than that for the system general corrosion value. As a result, predicted corrosion rates can be substantially different than the actual corrosion experienced on a pipeline. Accufacts is not able to reach any specific conclusion in the Advantica report as to the predicted general corrosion rate, except that these specific studies appear to conclude that inhibitor treatment should prevent excessive internal corrosion. The calculated growth corrosion rate with no corrosion inhibitor mentioned in Appendix D.2.3 of 6.6 mm over 30 years (0.22 mm/yr) appears to be very low for a gas production pipeline. Specific reasons for such a low non-inhibitor value need further clarification (e.g. very dry gas, low temperatures, low CO2 concentration, etc). Accufacts remains very concerned that internal corrosion may be the predominant risk of concern for this pipeline. We agree with Advantica's recommendation to re-evaluate the internal corrosion rate prediction, and advise that such prediction methods consider that the onshore production pipeline: 1) has been exposed to the elements for quite some time, 2) may sit in inhibited water for an extended period of time, and 3) will not be periodically pigged with a cleaning pig. All of these factors can significantly increase the likelihood of accelerated selective internal corrosion sites, especially on a gas production pipeline.

As recently demonstrated by the British Petroleum release on the North Slope of Alaska, internal corrosion rates for pipelines associated with production facilities can still get away from the operator if certain internal corrosion protection processes are not properly in play. While the Alaska release was on an oil pipeline, it helps to illustrate that even for a pipeline that had a corrosion inhibitor program in place, corrosion rates can suddenly change for many reasons, sometimes resulting in a rapid acceleration of internal corrosion and pipeline failure. Press releases suggest internal corrosion on the Alaska pipeline increased rapidly since last October at this site. The Pipeline and Hazardous Material Safety Administration (the realigned Office of Pipeline Safety) has also indicated that other serious metal loss from internal corrosion is evident on this pipeline in many places despite previous smart pigging on this line and a corrosion inhibitor program (there was no cleaning pig program).

3. Integrity Management Program is Not Well Defined

Both Advantica and Accufacts are in very strong agreement that the critically important integrity management ("IM") program lacks sufficient detail to insure confidence that the program will be effective. Given the higher risks associated with internal corrosion as discussed earlier, lack of an effective IM program should be of very grave concern. While many international pipeline standards have areas in their regulations that are superior to U.S. pipeline regulations, Accufacts believes that U.S. pipeline regulations addressing IM lead the world in implementation and practice. Tragically, the catalyst for this recent U.S. advancement in pipeline safety was two major pipeline failures with multiple loss of life (Bellingham in 1999 and Carlsbad in 2000).

In the U.S. the first round of IM inspections required by new pipeline safety regulation are now complete for both oil and gas transmission pipelines and this experience should serve as a database on the inspection types and technologies that best serve to identify certain risks on pipelines. The U.S. has at least an order of magnitude more pipeline miles actually inspected under integrity management compared to many other countries. Accufacts must caution that in the U.S., pipelines continue to experience pipeline failures from corrosion despite the running of smart pigs (the Alaska North Slope pipeline mentioned above is just one example). While smart pig tools can prove highly advantageous, their effectiveness can be limited for a variety of reasons, one being changing corrosion rates as discussed above. In addition, thick walled pipe, such as that used for the Corrib pipeline, presents technical challenges that also limit the smart pig technology that can be applied to this system. It is critically important that any smart pig tools be substantiated by field confirmation as defined in API 1163, and the capabilities of smart pigs to determine both external and internal corrosion not be overstated. An integrity management program utilizing smart pig inspection for the Corrib onshore system would want to deal with risks associated with 1) external and internal corrosion, 2) pipeline movement, and 3) third party damage.

* Please credit Dr Richard Kuprewicz and Village magazine if using this commentary elsewhere. If reproducing for commercial use, please contact Village for permission