The TNT and ANFO equivalences of the Beirut explosion

Abstract

On 4 August 2020, an estimated 2.75 kt of ammonium nitrate exploded in the Port of Beirut, Lebanon. Rigby et al. (Shock Waves 30:671–675, 2020) used observations from numerous video recordings of the explosion to estimate the time-of-arrival of the primary shock at 38 positions, the distances of which from the centre of the explosion were determined from Google maps. These data were analysed to make a preliminary estimate of the TNT equivalence of the explosion, with a best estimate of 0.5 kt and an upper limit of 1.12 kt. Rigby et al. have made their data available for other users, and the present paper describes how these data have been further analysed to obtain the TNT and ANFO equivalences as functions of radial distances from the centre of the explosion. The TNT equivalence varies from 0.15 to 0.7 kt, with a median value of about 0.5 kt, in agreement with the Rigby et al. result. The ANFO equivalence varies from 0.2 to 1.0 kt with a median value of about 0.75 kt. Some additional thoughts about the Beirut explosion are appended.

Appendix

The following are some thoughts by the author, not directly related to the analyses in the above paper, based on his own experiences with the study of ammonium nitrate as an explosive, and his measurement and analysis of several large scale air burst and surface burst ANFO explosions.

Ammonium nitrate, a widely used and easily obtainable fertilizer, normally comes in the form of a prill of small pellets. In this form, it is easy to transport and use, but, unless it is strongly confined in a borehole or metal container, it is very difficult to detonate and is therefore considered safe for public use. Even in large quantities with a booster charge, the prill is dispersed and does not detonate. It can be transformed into an explosive material by thoroughly mixing it with diesel oil in a ratio of about 5% by mass (ANFO).

Ammonium nitrate is hydroscopic and easily absorbs moisture from the atmosphere. After prolonged exposure to a moist atmosphere, the prill sets into a rock hard mass. In this form, it is detonable, and it has been suggested that it may spontaneously detonate when small changes in the ambient conditions initiate a rapid exothermic change in the crystal structure. It seems probable that the large amount of ammonium nitrate stored in Beirut for 6 years, only a few metres from the Mediterranean, would have absorbed a considerable amount of moisture and turned into a solid rocky mass. We know that the relative humidity was high at the site of the explosion because of the condensation cloud produced in the negative phase of the blast wave, as seen in the videos. This may have been one of the reasons why the stored fertilizer was not moved, because jackhammers would have been needed to break the mass into transportable pieces, and such a procedure is not recommended for use on a high explosive material. If in a solid form, it may have been possible to use hoses to slowly dissolve the mass because ammonium nitrate is highly soluble in water. The resulting run-off could have been collected and provided to local farmers as a rich liquid fertilizer. Heating the solution would not be recommended as it produces nitrous oxide, laughing gas.

The analysis of accidental explosions is usually not an accurate procedure because the physical properties of the blast wave can only be estimated from the damage produced. It was unusual therefore in the case of the Beirut explosion to have relatively accurate measurements of the shock trajectory obtained from a number of video recordings. The reason the recordings were being made at the time of the explosion was because the event was preceded, and caused by, a fire in a fireworks storage facility. However, this also meant that there were a large number of people watching the pyrotechnics through their windows, and this resulted in a large number of casualties from the shattered window glass. Many of the newsreels immediately after the blast showed people with lacerations to all parts of their bodies, undoubtedly caused by flying glass.

Even in the case of a nuclear explosion, it is estimated that a major proportion of casualties would be caused by flying glass. This is because there is a wide annulus around the centre of an explosion in which the only significant damage is to window glass. The area of this annulus is similar to that of the central circle of major damage and injury from all causes.

When a window is shattered by the loading of a blast wave, it typically breaks into elongated triangular shards, each with a molecular sharp point. The shards hinge inwards, breaking away from the frame to form a shower of high speed and very dangerous missiles. In the event of a large explosion, the initial flash triggers observers to look out of a window, which is the most dangerous thing they can do. There may be a few seconds before the blast reaches the window, enough time for someone to turn away and move out of the direct line along which the high-speed shards will travel.

At the extreme limit of window damage from a blast, it has been observed that for a significant proportion of the damaged windows the glass falls outwards instead of inwards. The suggested reason for this is that the glass does not fail under the initial pressure, but becomes detached from the frame on the rebound, and this effect may be enhanced by the negative overpressure phase of the blast wave, which is approximately twice as long as the positive phase. Because of this effect, in the event of a large explosion, people should not stand close to a tall building as there may be a lethal shower of broken glass falling from the extruded windows.