Enbridge underestimates oil spill risk: report
Consultants for Enbridge Northern Gateway recently criticized a report we prepared identifying deficiencies in their oil spill risk assessment. They made a number of assertions about our report that are incorrect.
The purpose of our report was to use international best practice standards to evaluate Enbridge’s risk assessment. Our evaluation identified 28 major deficiencies in Enbridge’s analysis.
Enbridge itself confirmed the existence of many of these deficiencies during cross examination at the Northern Gateway hearings. For example, Keith Michel, one of Enbridge consultants and co-author of the critique published May 14 in The Vancouver Sun, stated under oath that the data Enbridge used to forecast tanker spills under-report incidents by 50 per cent and spills by 25 per cent. Other studies suggest even higher under-reporting.
Despite agreeing that the statistics they use significantly under-report spills, Enbridge made no effort to correct for this under-reporting and consequently relied on data that understate the risk involved. It is also disconcerting that Enbridge failed to disclose this limitation in the data until they were challenged during cross examination.
Failure to clearly state and address deficiencies like under-reporting is contrary to sound risk assessment and is one of the many deficiencies we identify that result in an underestimate of risk for the Northern Gateway.
To assess the impact of these deficiencies in Enbridge’s risk assessment we did additional sensitivity analysis using Enbridge’s risk model and used other well-accepted models to compare the results to Enbridge’s results.
One alternative model we applied is the U.S. Oil Spill Risk model, which was developed by U.S. government scientists to assess spill risk. This model has been subject to rigorous testing and refinement in peer reviewed publications and legal proceedings over several decades. The model was just updated in June 2012 and incorporates recent mitigation measures and the resulting declines in tanker spill rates.
The U.S. model shows that even with these declines in tanker spill rates, there is still over a 95 per cent chance of an oil spill greater than 1,000 barrels and over a 65 per cent chance of a spill over 10,000 barrels during the operating life of the Northern Gateway. Recent research also suggests that these spill statistics may underestimate future risk because they do not account for the forecast increased risk resulting from the aging tanker fleet over the next several decades.
For pipeline spills we use data from a large number of pipelines and sources, including information supplied to us by Enbridge based on the operation of its own pipeline system, which Enbridge maintains is the “world’s longest, most sophisticated crude oil transportation system” and “is recognized as an industry leader in pipeline safety and integrity.” Enbridge’s experience therefore provides a relevant real world test of industry leader pipeline safety.
Enbridge’s spill record from 2002-2010 prorated for the size of the Northern Gateway is equivalent to over 15 pipeline spills per year, 31 times greater than Enbridge’s forecast of one spill every two years. While Enbridge maintains that new technology will reduce future spills, the accompanying chart, which shows that Enbridge’s pipeline spill rate is actually increasing despite new technology, raises cautions about claims of dramatic reductions.
Another concern confirmed in the hearings is that Enbridge will not accept full liability for oil spill damages. Consequently, the costs of larger oil spills could exceed available compensation and taxpayers could end up bearing some proportion of the costs of an Enbridge spill.
Many of these findings in our report including the spill risk estimates based on the U.S. model and identification of deficiencies in the Enbridge risk analysis were submitted as evidence to the JRP review process and our current report was subject to review by international experts prior to publication. It is interesting to note that Enbridge raised no concerns over our spill evidence during cross examination.
We acknowledge that all risk assessment methodologies have strengths and weaknesses as we document in our report and forecasting risk is challenging. However, all methodologies we review show a high probability of spills. Even Enbridge’s method, which underestimates risk, forecasts a 71 per cent chance of a tanker, port or major pipeline rupture over the operating life of the Northern Gateway, which increases to over 99 per cent if non-rupture pipeline spills are included.
In sum, the risk assessment provided by Enbridge contains serious deficiencies that need to be addressed to provide decision makers with an accurate assessment of the risk of the Northern Gateway.
What is needed to resolve these deficiencies is a more collaborative approach to risk analysis. By having experts work together instead of against each other in an adversarial process, we can provide decision makers with the best science available to make the best decisions possible.
Dr. Thomas Gunton is Director of the Resource and Environmental Planning Program at SFU. Sean Broadbent is an environmental consultant and PhD candidate at SFU.