Pollution
Humain
Environnement
Economique

Two explosions occurred at 4 p.m. in a pilot plant of a chemical company, during the distillation of a crude nitro-imidazole compound at 100°C under vacuum.

Around noon, the process operator added 154 kg of solid 1-methyl-2-formyl 1-nitroimidazole into a 450 l glass lined reactor; he set the temperature of heating medium (glycol/water mixture) at 100 °C at control panel then turned on the circulation pump to reactor’s jacket. The heating was intended to melt the solid that should then be dropped into 20 litres glass distillation bowls to be used in the pilot plant distillation unit; laboratory trials had shown that those conditions could yield pure product.

The solid was almost melted after 4 hours when the operator turned on the reactor’s agitator to finish the melting. Shortly afterwards, the mixture began to boil up and a yellow fume to emerge from the closed man-hole. The operator exit quickly and 2 successive explosions were heard within seconds.

All personnel went to the assembly area in accordance with the site’s emergency plan. The company fire crew equipped with breathing apparatuses, put out small fires with the support of local fire emergency services. The policemen evacuated two nearby factory units (300 persons).

The plant was severely damaged (roof blown off, internal walls blown down, external walls were severely damaged, glass windows shattered up to 30 m…). Missiles from the pilot plant were projected up to 60 m. The head of the process reactor was projected through the roof and landed about 30 m away while the bottom portion was projected from the first floor area down to the ground floor area and embedded in the floor. The cost of the damage has been estimated at about 0.65M Euros.

Two persons were slightly injured. Although there was no damage off-site the sound of the explosion was heard some distance away. In a town 2 km away windows were shaken and the noise caused considerable public concern.

The explosion was caused by the decomposition of the product due to the heating. As the liquid level in the reactor was below the temperature probe, its exact temperature was unknown. The company had recognised the need for explosivity tests (potential explosive properties due to the presence of a nitro- group) but those had not been carried out for pilot scale. The company intended to compile the relevant data (including health and safety information) if the process developed to full production scale.

The company stopped their research and development activities on new processes; should they start again, they will have to implement a programme to ensure an adequate hazard evaluation at each stage.