An explosion at 5:05 pm involving an organic powdery product at a plant making starchy products was heard as far as the nearest residential zones 500 m from this lower-tier SEVESO-rated facility. The operator was able to quickly locate the source at the level of filters on one of the protein dryers (gluten type) placed on the 3rd floor of a drying tower 18 m high containing 10 tonnes of proteins. Fire broke out on this dryer subsequent to the explosion. The local safety systems were activated with the opening of pressure surge vents, in turn triggering the installation’s automatic fire protection systems.

The internal emergency plan was also activated, and the lone employee present at the dryer station was evacuated. The Departmental Fire Services dispatched some 30 fire-fighters to the site with 4 nozzles, one of which was mounted on a ladder. The fire was confirmed under control at 6 pm and definitively extinguished during the following hours; the emergency plan was lifted at 8:40 pm. The emergency services intervention was concluded by verifying the absence of any flashpoint in the tower. No victims were reported and no redundancies needed to be scheduled. Outside of the noise disturbance, no environmental consequences were detected. The local gendarmerie visited the site. The cost of damages was limited to cleanup and repairs to the dryer; operating losses remained insignificant (3 other dryers still available).

The damaged dryer had been placed into service in 2008, and a comprehensive safety report for the site had been anticipated for October 2010. The installation was running normally at the time of this incident, with the filter handles having been replaced 48 hours before this sequence of events; the installation was once again operational 24 hours later.

According to the operator, a plate wedged into the upper part of the pre-filter was found embedded in the extraction screw of the central part of the filter. This technical flaw would explain the difficulties encountered on the screw, which had become locked on several occasions since the previous Monday morning. A mechanical friction phenomenon between the plate and screw caused the spark and/or flashpoint that ultimately led to the explosion.