tankers increasing from 7,112 to 7,270, and the deadweight of the tanker fleet decreasing from 340 million tonnes to 299 million tonnes. Due to the phase-out schedule of OPA 90 and MARPOL Regulation 13G, the fleet is becoming younger, such that in 1999 more than 50 percent of the tanker fleet was less than 15 years of age. In recognition of the fundamental changes that took place after the Exxon Valdez accident, spill data from 1990 onward were used as the basis for estimating the amount of oil spilled from tankers (tank ships and tank barges). The U.S. Coast Guard consistently collected data on oil spills during the 1990s. These data are considered reliable, particularly for the larger spills, and they serve as the basis for the estimated spills in U.S. waters. The data were sorted geographically to remove inland spills. During the period from 1990 to 1999, there were 513 spills from tankers and tank barges in U.S. coastal waters of at least 100 gallons in size, and the average annual spillage was 4,000 tonnes. Of the recorded spills, 64 were greater than 10,000 gallons (34 tonnes) in size, and these accounted for over 95 percent of the total spillage. One single spill of 13,000 tonnes (3.9 million gallons) from the Mega Borg in 1990 accounted for 36 percent of the total spillage during this ten-year period.
Because comprehensive spill databases for Canadian and Mexican waters are not available, spill volumes for these waters were estimated by adjusting the U.S. figures by the relative movements of petroleum. On this basis, the calculated oil spillage to North American waters, from tankers is 4,000 tonnes per year (see Table 3-2). This calculated value was taken as the minimum estimate. Recognizing the completeness of the U.S. data set, the spill quantities were increased by 5 percent to 5,300 tonnes to obtain the best estimate, and by 20 percent to 6,400 tonnes to obtain the maximum estimate.6
International spill data were obtained from the Environmental Research Consulting database and includes information gleaned from the International Maritime Organization, ITOPF, and other national and regional agencies. These international data are not consistently collected and do not include spills under 10,000 gallons (34 tonnes) in size, and are therefore regarded as underestimates. The international spill quantities were increased by 25 percent to obtain the minimum estimate, by an additional 10 percent to obtain the best estimate, and further increased by 25 percent to obtain the maximum estimate. The international and North American figures were then combined to produce the worldwide estimates of spillage from tankers. The best estimate is 100,000 tonnes; the minimum and maximum estimates are 93,000 tonnes and 130,000 tonnes, respectively.
The NRC 1985 report and the 1990 study (IMO, 1990) both used data from the International Tanker Owners Pollution Federation Ltd. (ITOPF) to estimate the quantity of oil entering the marine environment from tanker accidents. In the NRC 1985 report, the spillage from tankers was averaged over the 7-year period from 1974 to 1980, establishing a best estimate of 390,000 tonnes per year. In the 1990 study, the spillage from tankers was averaged for the 10-year period from 1981 to 1989, establishing a best estimate of 114,000 tonnes per year. It is believed that the collection of international data prior to 1990 was even less consistent than it is today. Also, in the 1985 and 1990 reports no adjustments were made for the deficiencies in the database, and therefore care should be taken when comparing these estimates. In U.S. waters where the spill data have been more consistently recorded since 1980, spillage from tankers during the period 1990-1999 was less than one-third of the spillage recorded during the period 1980-1989.
During normal operations, certain tankers may discharge into the sea an amount of oil contained in the ballast and tank washings. Under regulation 13 of MARPOL 73/78, tankers of 20,000 tonnes deadweight and above are required to have segregated ballast tanks (SBT), dedicated clean ballast tanks (CBT), and/or crude oil washing systems (COW), depending on the vessel type, when they were built, and their size. Regulation 13F adopted in 1992 requires all new tankers to have double-hull or equivalent protection of all oil tanks within the cargo block. MARPOL Regulation 13G requires mandatory retirement for single hull tankers at 30 years of age. A revision to regulation 13G requires phase-out of all single hull tankers above 20,000 tonnes deadweight by 1 January 2007.
Generally, crude oil carriers of 20,000 deadweight tonnes (DWT) and above and product tankers of 30,000 tonnes deadweight and above delivered since 1983 must have SBT. Segregated ballast tanks are ballast tanks that are completely separated from the cargo oil and fuel oil systems, and are permanently allocated to the carriage of water ballast. SBT greatly reduces the likelihood of oily ballast discharge, because there are sufficient segregated ballast tanks for normal operation in ballast. For these vessels, ballast may be allocated to cargo tanks only when needed to ensure the safety of the vessel in particularly severe weather. Unlike SBT, the piping systems for CBT may be common or connected with the cargo oil pump and piping systems. There are few CBT tankers operating today.
Regulation 9 of MARPOL limits the amount of oil that may be discharged into the sea to 1/15,000 of the total cargo oil volume for tankers built prior to the implementation of
Roughly 5 percent of the spills reported in the available databases did not have adequate geographic information to place them in any region with confidence. The 5 percent factor applied to develop the best estimate reflects the committee’s confidence in the reporting of spills, the completeness of available databases, and a recognition that 97 percent of the total spill volume captured by these databases comes from spills that exceed 100 gallons. The likelihood that a spill much larger than that will go unobserved is, in the committee’s opinion, rather small.