Experiments and Observation of Peat Smouldering Fires

14 February 2007

If a subsurface layer of peat is ignited, it smoulders (flameless combustion) slowly but steadily. These fires propagate for long periods of time (days, weeks, even years), are particularly difficult to extinguish and can spread over very extensive areas. Smouldering fires have an impact on the soil since the fuel burning is the combustible portion of the soil itself. An example of a peat fire is the recent wildfire that occurred this summer in Rothiemurchus, near Aviemore (Scotland). A 40 year old plantation of lodgepole pine of about 15 ha was burned accidentally. The main fire was extinguished within three days, but the peat underneath the forest continued to smoulder for more than thirty days, destroying the tree roots. Peat combustion processes remain mostly unknown to the fire community and more experimental and theoretical studies are needed. It is known that the smouldering of underground fuels is governed by the diffusion of heat and oxygen through the soil layers between the reaction site and the surface. It is highly desired to conduct more experiments on smouldering (aka glowing) combustion of peat to allow derivation of fire maps for forecast from information in the weather maps. The main factor controlling factor ignition is moisture content. This paper reports the measurements and observations from smouldering box experiments using peat from the UK. The smouldering box is a simple procedure which captures the essentials of peat ignition and burning. The test is conducted in a small box (side of ~20 cm) made of insulating side-walls and open on the top. The sample is ignited using an electrically heated coil running along one side of the peat. A successful ignition is followed by sustained smouldering combustion that consumes the sample. Temperature and spread rates are measured, and the residue is weighed and described. In our series of experiments we have found that the critical moisture content below which ignition is highly probable is 120 % ± 10 % (in dry basis). We found that this value compare well with that reported by researcher in Canada (~110%) and Indonesia (>120%), which suggests that critical moisture content depends weakly on peat origin.