At around 9 a.m., an outbreak of fire was detected in a wood chip storage silo at a wood processing plant. The detection of low-level sparks at the outlet of the chip sorter triggered the sprinkler system on the conveyor. Then, at 9:20 a.m., a high-level detection shut down the process and initiated the installations’ sprinkler system. The site’s internal firefighters identified a fire in the 300 m³ concrete “firebreak” silo between the dryer and the sorter, and smoke was seen billowing out from the bottom of the silo. At the time of the accident, the silo was filled to one-third of its capacity. The operator turned on the silo’s sprinkler system and began to empty it. At 10 a.m., the silo had been emptied to 10% of its capacity, and an even more considerable amount of smoke was being released. The fire brigade was alerted. One of the firefighters, in an aerial basket, was overcome by carbon monoxide as he opened one of the silo’s hatches. Two other firefighters were also affected by carbon monoxide while trying to rescue him. Carbon monoxide (CO) concentrations of 300 ppm were measured at the base of the silo. The intervention by the fire brigade ended at 3:30 p.m.

The fire had originated from smouldering embers in the areas where chips can accumulate inside the silo. The fire may have smouldered several hours or even days before detection, although the temperature sensor inside the silo had failed to detect anything. Following a fire just a few months earlier (ARIA 48557), the operator lowered the drying temperature from 120 °C to 100 °C. Temperature and humidity measurements at the dryer outlet did not reveal any anomalies in the days preceding the accident. The operator explained these outbreaks of fire by the particularity of the materials being dried upstream of the silo. This dryer was being used to process a mixture of green and recycled wood having different levels of moisture content. It is possible that this mix of material led to either overheating of the already dry chips or, on the contrary, to maintaining more humid chips after exiting the dryer.

The proposed measures involved emptying the silo every 2 weeks to avoid any persistent smouldering fires. An analysis of the chips accumulated at the bottom of the silo is now conducted on a regular basis, as well as a weekly thermography measurement.