Assessment of the risk of transporting propane by truck and train

Abstract

The risk of shipping propane is discussed and the risk assessment methodology is summarized. The risk assessment model has been constructed as a series of separate analysis steps to allow the risk to be readily reevaluated as additional data becomes available or as postulated system characteristics change. The transportation system and accident environment, the responses of the shipping system to forces in transportation accidents, and release sequences are evaluated to determine both the likelihood and possible consequences of a release. Supportive data and analyses are given in the appendices. The risk assessment results are related to the year 1985 to allow a comparison with other reports in this series. Based on the information presented, accidents involving tank truck shipments of propane will be expected to occur at a rate of 320 every year; accidents involving bobtails would be expected at a rate of 250 every year. Train accidents involving propane shipments would be expected to occur at a rate of about 60 every year. A release of any amount of material from propane trucks, under both normal transportation and transport accident conditions, is to be expected at a rate of about 110 per year. Releases from propane rail tank cars would occur about 40 times a year. However, only those releases that occur during a transportation accident or involve a major tank defect will include sufficient propane to present the potential for danger to the public. These significant releases can be expected at the lower rate of about fourteen events per year for truck transport and about one event every two years for rail tank car transport. The estimated number of public fatalities resulting from these significant releases in 1985 is fifteen. About eleven fatalities per year result from tank truck operation, and approximately half a death per year stems from the movement of propane in rail tank cars.

Cite this work

No public copy is stored for this record, so only the metadata is cited.

Stored only when both papers are in the corpus — 12 citations exist upstream.