ABSTRACT

Synopsis

The regulations governing the transport of nuclear materials are based on the IAEA Regulations for the Safe Transport of Radioactive Material. A major feature of these regulations is that the principal protection against the hazards of the radioactive material should be provided by the packaging. To ensure this the IAEA have specified a series of tests for the certification of packages. The most severe tests are designed to demonstrate the ability of a package to withstand, essentially intact, severe accident conditions. Recent assessments suggest that the IAEA regulations give protection in a very high percentage (over 99% for example) of expected road accidents. However, during air transport, package impact speeds could be an order of magnitude greater than during other transport modes. Hence, current IAEA accident condition tests may not provide such a high level of confidence in the package surviving an air crash. It has therefore been suggested that additional tests should be specified for packages intended to transport significant quantities of activity by air.

This paper reports a piece of work undertaken recently by SRD to provide material for discussions on the transport of radioactive materials by air and in particular on the formulation of fireball and sequential impact and fire tests. It comprises a search of aircraft accident reports for data relating to the occurrence, temperature and duration of fireballs, and the duration and temperature of fires following a severe impact. A qualitative analysis of the data obtained is also presented.

The study concentrated on a small sample of aircraft accidents during the years 1975 and 1985 inclusive, that

110fulfilled the following criteria:

Aircraft accident reports for the incident were readily available.

The aircraft involved had a maximum take-off weight of about 20 tonnes or greater.

At least one fatality among passengers or crew and/or destruction of the aircraft had occurred.

Only crashes fulfilling these latter criteria were believed to have the potential to damage a robust package significantly.

No attempt was made to quantify the impact parameters such as impact forces, fire temperature and fire intensity. Instead, the incidents were classified qualitatively into broad categories by examining the type and extent of damage done to the aircraft and persons involved in the crash.

Out of a total of 31 accidents, 12 involving a fireball, flash-fire or explosion were identified. Ten accidents were identified in which the crash environment was considered to be of such severity that the possibility of a package being breached must be considered. Six of these involved high impact forces, four involved large ripping or tearing forces in conjunction with a considerable fire. One extremely severe and prolonged fire (36 hours) was reported; the impact forces associated with it were negligible. The majority of the accidents occurred on, or close to, an airfield; however, fire fighting forces did not always find extinguishing fires straightforward.

From this limited study the following tentative conclusions were drawn.

A proposed fireball test is not supported; in general a fireball is followed by a ground fire with potential to cause much greater damage.

Very high impact forces are not reported in association with fires of greater than 30 minutes' duration. However, two incidents were identified in which considerable ripping or tearing forces were experienced followed by long fires (9 hours and 49 minutes). This type of scenario deserves more detailed consideration.

The possibility that an undamaged package may be subjected to a very prolonged engulfing fire does exist; however, the probability of this occurring will be very small.