In recent years leading researchers in the field have recognised that the data from these lab tests is compromised because, by necessity, the rate of oxidation is increased, giving results different to what happens in the field.

Published in June 2005 Australian Longwall Magazine

The typical outcome of laboratory work is a qualitative interpretation of the test, one that lacks quantitative assessment of the potential for spontaneous combustion in an actual coal mine. Until now. Enter Dr David Humphreys.

Humphreys has been studying spontaneous combustion for 30 years and recently completed his PhD on the subject. He is also the person who coined the term R70 as the descriptor for a commonly used lab test that rates a particular coal’s propensity for self-heating. And self-heating is, of course, what can cause a fire.

According to Humphreys, since the 1970s, 50% of the fatalities in Queensland coal mines were associated with spontaneous combustion events.

In an attempt to better understand how spontaneous heatings happen, Humphreys has developed a number of numerical modelling techniques to simulate what happens during a heating, allowing, for the first time ever, different criticalities to be taken into account.

“What is poorly understood is the degree to which changes in many parameters affect the spontaneous combustion behaviour of coal,” he said.

In fact, he said it was only the advent of powerful computing technology that had allowed for an increased understanding of what actually happens when coal self-heats.

The effects of seam thickness, particle size, temperature and numerous other parameters can be entered into the model to provide a better understanding of how a particular coal may be expected to react.

For example, Humphreys said if a coal seam were 150mm thick, any heat generated by the oxidation process would be readily lost to surrounding strata, with little chance of spontaneous combustion. It would be a different story if the same coal were 10m thick.

“As the coal seam thickness increases it could be expected that the heat losses relative to the heat generated by oxidation decrease and lead to higher and higher temperatures,” he said.

As mines get deeper, understanding the coal becomes more critical, particularly where mines are affected by temperature, as in Queensland, because with every 10C rise in temperature the rate of oxidation effectively doubles.

It is the combination of the various coal properties (particle size, heat of oxidation, rate constant and initial temperature) that determines whether self-heating will occur, and this can best be described by determining a series of critical self-heating parameters from the numerical models.

Humphreys concluded that by comparing the critical self-heating period, temperature and thickness against the in-situ ground temperature, thickness and likely exposure period, the operator would have a better assessment of the potential for spontaneous combustion occurring in any mining or stockpiling situation.

“The models are continually evolving. We are taking our best understanding of the oxidation/heat transfer process and factoring in the impact of moisture transfer between coal and the atmosphere and then estimating the criticalities,” he said.

Work is also underway to look at the effect of moisture, which has a considerable effect on coal’s potential to oxidise, Humphreys said.