Tanker study shows huge gaps in shipping and hazard data, documents show

The study of tanker shipping and tanker spills by Genivar for Transport Canada has revealed huge gaps in how the world monitors tanker traffic.

Genivar report
Cover of Genivar tanker report (Transport Canada)

Genivar says

Accident data was acquired from three main sources: the CCG Marine Pollution Incident Reporting System (MPIRS); the Lloyd’s casualty database; and spill incident records maintained by the International Tanker Owners Pollution Federation (ITOPF).

MPIRS lists all marine pollution incidents occurring in Canadian waters (CCG, 2013), with information on the region within Canada in which the incident occurred, type of material spilled, accident cause, and estimated pollution volume with multiple entries for a given incident showing updates of incident status and pollution amounts if applicable. The primary use of MPIRS in this study was for spill incidents in the smaller size categories… for which worldwide data was suspected to be unreliable due to under-reporting. MPIRS appeared to be a comprehensive listing of incidents that occurred in Canadian waters, and a summary of polluting incidents

It goes on to note that some key data has not been updated since the 1990s, largely prior to the introduction of double hulled tankers.

As noted, oil spill accidents were compiled on a worldwide basis.

In order to estimate the frequency for Canada, an exposure variable was required.

A series of studies by the U.S. Minerals Management Service (MMS, now known as
the Bureau of Ocean Energy Management, Regulation and Enforcement) investigated the occurrence rates of tanker accidents against various spill exposure variables and found that the simplest and most reliable indicator was volume of oil transported. Simply put, it was determined that spill rates could be expressed, for a range of spill size categories, as an average number of spills per billion barrels transported.

The MMS studies were updated periodically until the 1990s but have not been revisited since, but they did show a steady decrease in the likelihood of casualties and resulting spill volumes, due to a number of factors including tanker design, increasing governance and overall scrutiny of the marine transportation industry. The phased-in implementation of double-hull tankers may have also had a beneficial effect on spill rates in more recent years, particularly in the category of very large or catastrophic events… In any case, it is important in interpreting accident data to reflect current trends and implemented mitigation measures. The focus was on cargo volumes and accident rates over the past decade.

It goes on to say the volumes of crude carried is also under-reported to Lloyds.

In the case of crude oil and refined products carried as cargo, the exposure variable was simply the volume of each respective category carried on an annual basis for the period of interest. Information from the Lloyds APEX database was used for this purpose; it reports volumes of crude and refined products shipped worldwide, with a breakdown by year, country of origin, and country of destination. Compared with similar data from Canadian sources, the APEX data appeared to significantly under-report the carriage of refined products. As a result, the accident rates estimated and used in this study are likely somewhat conservative, that is, they overstate the likely frequency of refined products carried as cargo. For all calculations involving the potential spillage of refined products as cargo in Canadian waters, and for the apportioning of spill frequency among the various sectors and sub-sectors of Canada, Transport Canada commodity traffic data was used

Again about Lloyds data, until 2010, it was limited in its monitoring of the BC Coast.

In analyzing the Canadian movement data supplied by Lloyds, a major shortcoming was found in the data in that movements recorded prior to 2010 did not include broad classes of vessels such as ferries, passenger vessels, and pilot boats. Given that these vessels comprise a significant proportion of traffic movement in many sectors, only data covering the final two years of the record, 2010 and 2011, were used in the analysis.

The Lloyds data was also limited when it came to oil spills:

One limitation of the MPIRS data was that it did not classify spills as to whether they were from “cargo” as opposed to “fuel”, which would have been helpful in this study as these spill types were analyzed separately. As a result, for spills of refined products, which could have hypothetically been either cargo or fuel, assumptions were made based on the type of vessel involved, the type and severity of the incident, and other notes within MPIRS.

A database was acquired from Lloyds that detailed all marine casualties over the
past ten years regardless of whether the incident involved pollution…
This database was used to provide a breakdown of incidents by cause, and as an
initial listing of those incidents that did result in pollution. The Lloyds data was of
mixed quality when it came to the reporting of polluting incidents, with numerous
records only partially filled out, ambiguities in the reporting of spill volume, and
inconsistencies in the classification of the spilled material. A significant effort was
made to provide consistency and accuracy in the information, including cross-
referencing with other data sources.


So the Genivar report exposes a significant gap in the available data on oil spills.

It is certainly true that the number of major tanker accidents and spills have decreased since the Exxon Valdez disaster, a point frequently made by Enbridge at meetings in northwestern BC.

The expert panel report which said that Canada faces the risk of a major tanker disaster of 10,000 tonnes or more once every 242 years.

The Vancouver Sun quoted Transport Canada spokeswoman Jillian Glover on that risk of a spill on the Pacific Coast as saying. “This value must be understood in relative terms, such that the risk is considered high compared to the rest of the country only…Canada enjoys a very low risk of a major oil spill, evidenced by the lack of Canadian historical spills in the larger categories. Additionally, this risk assessment is before any mitigation measures have been applied, so that is a theoretical number before additional prevention initiatives are taken.”

Note that the government always talks about a “major oil spill,” but it appears from the gaps in the data that predicting the possibility and consequences of a medium sized or smaller oil spill is now not that reliable, even though such a spill could have disastrous effects on a local area. According to a map in both reports, the entire BC coast is at risk for a “low to medium” spill. This echoes the problems with the Enbridge Northern Gateway pipeline, where Enbridge based most of its projections on a “full bore breach” or major pipeline break and did little about a medium sized or smaller leak. Data analysis by Kelly Marsh of Douglas Channel Watch on the possibility of the cumulative effects of a medium sized and possibly undetected pipeline breach could have just as disastrous consequences for the Kitimat valley as a major pipeline break. The same is likely true at sea.