Minerals Management Service indicates that there are 21,000 miles of pipelines in U.S. Outer Continental Shelf area (www.mms.gov/stats/), and DeLuca and LeBlanc (1997) estimate that there are 23,236 miles of offshore pipelines in North America. This represents a significant increase in petroleum hydrocarbons being transported during the past decade. In OCS waters, some 75 accidental spills occurred from 1990 to 1999, discharging a total 5,674 tonnes of petroleum hydrocarbons in the marine environment (Box 3-2). Based on the databases from Minerals Management Service on OCS spills and databases from coastal waters, the average annual discharge is calculated as 1,700 tonnes per year. Even though the database is fairly complete, there is always the uncertainty of estimating the volume of the discharges. The committee’s best estimate is 1,900 tonnes per year (minimum + 10 percent) with a maximum value of 2,100 tonnes per year (minimum +20 percent).5 An estimated 99 percent of the pipeline spills occurred in the central and western Gulf of Mexico, 66 percent (1,200 tonnes per year) of the spills occurred in state waters, while 34 percent (650 tonnes per year) occurred in OCS waters.

Estimating the volumes of discharge from pipelines worldwide is more difficult and uncertain. The number of miles of offshore pipelines in the world’s oceans (including North America OCS waters) has been estimated to be 82,748 miles (DeLuca and LeBlanc, 1997). This estimate is prob

BOX 3-2 Lake Barre Pipeline Break, Louisiana

In May 1997 an underwater pipeline rupture released 940 tonnes (280,000 gallons) of south Louisiana crude oil into Lake Barre, Louisiana. Lake Barre is part of a shallow coastal estuarine system separated from the Gulf of Mexico by barrier islands. Rates of subsidence in this area are among the highest in Louisiana, which among other factors results in extensive erosion of the marsh and creation of a very complex shoreline with isolated patches of sparse salt marsh and large expanses of open water. There had been such rapid change that aerial photographs taken in 1995 (two years before) were useless as base maps, and new aerial photographs had to be obtained. Water depth is about 2 m, and the tidal range is less than 1 m, and water-level changes are mostly wind-driven.

South Louisiana crude is a medium crude oil (API gravity of 31), and approximately 36 percent of the oil was estimated to have evaporated or dispersed within 48 hours. Steady onshore winds during the spill raised water levels in the marsh, and oil slicks and sheens eventually spread over 1,750 hectares (4,327 acres) of wetlands (Lorentz et al., 2001). With the high water levels and the sparse vegetation in the marsh, pockets of heavy oil trapped along the outer marsh fringe generated silver to rainbow sheens that dispersed north (downwind) through the marsh for over 6.5 km. Because of the high water levels, there was very little sediment contamination, and only 0.1 hectares (0.3 acres) of salt marsh died back. The oil left a black band a few tens of centimeters high on the vegetation on about 65 hectares (162 acres) of marsh. The widely spread sheens affected 1,685 hectares (4,165 acres) of wetlands over a two week period; two months later there was no evidence of oil or oil effects in these lightly oiled areas. Response efforts focused on recovery of floating black oil pooled in the edges of the marsh. Of greatest concern was that cleanup should not cause more damage to the vegetation, thus a rule was instituted, and strictly enforced, that responders were to remain in boats while working in the marsh.

Water samples collected five days after the release contained 0.6-2.2 ppb total PAH, dominated by naphthalenes. Limited sediment sampling indicated no large-scale contamination of bottom sediments. Residual petroleum hydrocarbons in sediment samples collected one year later were characterized as significantly weathered, with degradation of nearly all n-alkanes.

The only reports of dead fish or shellfish were three dead juvenile blue crabs (although adult crabs and fish were alive) in a total of seven crab pots inspected, some dead brown shrimp in a survey trawl, and a few dead forage fish in the marsh. An oil fates and effects model was used to estimate that 7,500 kg of fish and invertebrates were lost (Kern et al., 2001).

Estimating impacts to birds was difficult because the summer residents are mostly rails, small shorebirds, and similar birds that hide in marsh grass. Only two birds were found oiled and dead; 58 live oiled birds were observed but not captured. The bird impacts would have been significantly different if the spill had occurred in winter when very large numbers of wintering waterfowl are present.

Key factors that limited the biological impacts of this spill were: a crude oil that was highly degradable; rapid weathering of the oil because it spread over a large surface area; calm sea conditions so there was little mixing of oil into the water column or contamination of bottom sediments; time of year when relatively few birds were present; and little intertidal sediments so the oil did not strand onto the marsh substrate.


The 10 percent factor used to develop a best estimate and the 20 percent factor applied to develop an maximum estimate is somewhat subjective and 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.

The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement