Canadian Think Tank Study Reveals Relatively Good Safety Record for Oil and Gas Pipelines, Rail Transport, and Tankers

I just came across a news item on a recent study by Kenneth P. Green and Taylor Jackson at the Fraser Institute in Canada on the safety record for intermodal oil and gas transportation, including pipeline, rail, and ocean tanker. Based on the generally positive study results and in light of the contentiousness surrounding the Trans Mountain and Keystone XL pipelines and several news items about recent oil train accidents, I had initially assumed that the Fraser Institute might have a conservative slant, but on further research, learned that it is Canada’s leading independent think tank.

Source: Fraser Institute

The study results also point to the need to improve the safety mechanisms - including both leak detection systems and secondary containment structures - at midstream storage and distribution terminals. Significantly, while the report should help quell some of the irrational hysteria surrounding new pipeline projects in Canada and the US alike, it does not address the associated environmental damage caused by even the relatively small accident-related spills identified. It also does not address the fact while that the justification for much of the resistance to expanding North America’s oil and gas pipeline infrastructure is based on safety; a good part of that resistance simply relates to the NIMBY (not in my backyard) phenomenon to which we are all subject to a certain degree.

Here’s an abstract from the report’s executive overview:

Pipelines suffer few occurrences (accidents and incidents) given the amount of oil and gas that is shipped through them. Overall, between 2004 and 2015, pipelines experienced approximately 0.05 occurrences per million barrels of oil equivalent (Mboe) transported.

When petroleum and natural gas goods are evaluated separately, we find that the transportation of oil results in fewer occurrences than the transport of natural gas. Indeed, transporting petroleum products by pipelines resulted in approximately 0.04 occurrences per Mboe compared to 0.07 for natural gas products. This means that the rate of occurrences for transporting natural gas products was 1.67 times greater than the rate of occurrences for petroleum products.

The focus on the occurrence rate only tells part of the story for pipeline safety. In addition to having low occurrence rates, almost 70 percent of pipeline occurrences result in spills of less than 1 cubic metre (17 percent result in no spill).

Only 17 percent of pipeline occurrences take place in the actual line pipe, meaning that the vast majority of spills occur in facilities that often have secondary containment mechanisms and procedures.

The results were similar for rail, where the transportation of oil was found to result in fewer accidents per Mboe transported than natural gas. Also similar to the data on pipelines, most rail accidents occurred in facilities rather than in transit.

While both pipeline and rail transportation of oil and gas are quite safe, when comparing the two modes of transportation, pipelines continue to result in fewer accidents and fewer releases of product, when taking into consideration the amount of product moved.

Specifically, based on petroleum product transport data from 2004 to 2015, pipelines were 2.5 times less likely than rail to result in a release of product when transporting a million barrels of oil.

This study also evaluated marine tanker safety in light of the additional oil tankers that will result from the expansion of the Trans Mountain pipeline.

Since the mid-1990s there has not been a single major spill from oil tankers or other vessels in Canadian waters. One recent study conducted by the [Canadian] federal government on marine oil spill preparedness estimated that a major spill of over 10,000 tonnes was exceedingly rare and likely to only occur once every 242 years. Likewise, a spill of 100 to 1,000 tonnes is expected to occur once every 69.2 years.

Marine safety has also improved dramatically since the 1970s. For example, when comparing the number of spills in the 1970s to the 2010s (up to 2016) using international data, the number of spills between 7 and 700 tonnes has decreased from 543 to 35 and in this same period the number of large spills (>700 tonnes) has declined from 245 to 12. The amount of oil spilled has also dropped dramatically, falling from three million tonnes in the 1970s to only 39,000 tonnes in the 2010s.