Pipeline report fails real-world test
Opinion: Risk assessment report on the likelihood of future marine spills full of numerous errors that threatens credibility
By Keith Michel and Audun Brandsaeter, Special to The Vancouver Sun May 14, 2013
As professionals in the field of risk assessment, we’re faced daily with the task of separating legitimate risk from its misinterpretation.
A recent paper claiming Enbridge Northern Gateway’s risk assessment underestimated the likelihood of future marine spills is an example of a report that makes serious errors in its assumptions and as a result, significantly inflates the risks involved in the future operation of the project.
Lead author and Simon Fraser University professor Tom Gunton, along with co-author and PhD student Sean Broadbent, have failed to recognize the number and volume of tanker spills are declining markedly worldwide — even as the total number of tankers plying the world’s waters increases. All data seem to point to the fact that this trend will continue. (SEE FIG. 1 and FIG. 2)
This downward trend in marine spills is due to a number of positive developments in marine safety, including the phased-in acceptance over the past two decades of double-hulled tankers as the absolute industry standard in liquid fuel transportation, as well as a raft of new marine transportation regulations that have come into existence over the past 20 years, and a wholesale shift in the safety culture of the industry.
Yet authors Gunton and Broadbent, in their criticism of findings we presented at the Joint Review Panel examining the pipeline project, fail to recognize the significant downward trend in oil spills from tankers, neglect the positive impact of double-hulled tankers by applying spill statistics dominated by single-hulled tanker accidents, and incorrectly apply combined in-port and at-sea spill statistics when assessing spills during transit from Kitimat to the Pacific. The report authors further criticize Northern Gateway for not using the U.S. Oil Spill Risk Analysis (OSRA) model drafted in 1975 and originally designed to assess the risks of offshore drilling.
But Northern Gateway chose not to use this model to assess the risk of its project for a simple reason: it does not properly assess the localized conditions specific to B.C. For example, the OSRA model fails to account for important regional factors such as the size of ships and local B.C. environmental variables.
By contrast, based on expert judgment, Northern Gateway’s assessment of oil spill risk specifically accounts for these regional factors. Furthermore, the chosen methodology enables assessment of risk and risk reducing measures for each segment of the route, something that would not be possible using the OSRA model.
Even more curious is Gunton’s and Broadbent’s focus on only a small subset of historical failure incident data. Their failure frequency analysis is based on data from only two pipelines. Spill incidents on these pipelines are reflective neither of the industry experience nor of the new technology proposed for Northern Gateway.
But perhaps what’s most telling in the Gunton-Broadbent paper is the fact that its claims fail the real-world test. Based on Gunton’s and Broadbent’s estimates, we would expect 21 to 77 large tanker spills every year around the world. Instead, there are now on average fewer than two large spills per year worldwide. In 2012, there were none.
Had the authors made their paper available to the Joint Review Process for scrutiny — as in fact Northern Gateway did with its reports — then these flaws might have been more openly reviewed and discussed. The public would no doubt have gained from the interaction.
Employing flawed methodology, the authors fail to account for a range of mitigating factors that will make Northern Gateway truly world class — including a commitment to double hulls both for the crude being transported as well as for the fuel driving the vessel, two-tug escorts for laden tankers, enhanced navigational technology, better marine transportation procedures, improved ballast coatings, reduced speed requirements and tough environmental limits.
At the end of the day, readers should know our own peer reviewed analysis finds no reason to think the probability of marine spills here is any greater than the probability anywhere else in the world, where improvements continue and spill incidents are in a steady decline even as tanker traffic increases worldwide.
Keith Michel and Audun Brandsaeter are experts in maritime and oil and gas risk assessment. They appeared before the Joint Review Panel on behalf of Northern Gateway Pipeline project to present their detailed risk assessment findings.