The application of numerical modelling to the assessment of the potential for, and the detection of, spontaneous combustion in underground coal mines

Humphreys, D. R. (David R.) (2004). The application of numerical modelling to the assessment of the potential for, and the detection of, spontaneous combustion in underground coal mines PhD Thesis, School of Engineering, The University of Queensland.

Attached Files (Some files may be inaccessible until you login with your UQ eSpace credentials)
Name Description MIMEType Size Downloads
THE18186.pdf Full text application/pdf 24.20MB 7
Author Humphreys, D. R. (David R.)
Thesis Title The application of numerical modelling to the assessment of the potential for, and the detection of, spontaneous combustion in underground coal mines
School, Centre or Institute School of Engineering
Institution The University of Queensland
Publication date 2004
Thesis type PhD Thesis
Supervisor Dr Stewart Gillies
Dr Peter Golledge
Total pages 184
Collection year 2004
Language eng
Subjects L
291301 Process Metallurgy
640205 Other non-ferrous ores (e.g. copper, zinc)
Formatted abstract

Spontaneous combustion of coal has been a hazard to the mining industry from the very first attempts to mine coal.


The assessment of the propensity of coal to spontaneous combustion is largely based upon the results of any one of a large number of laboratory based tests. Inevitably, there are compromises in the laboratory tests to increase the rate of oxidation and self-heating from that which would occur in-situ. As a result, the assessment of the propensity of a coal to spontaneous combustion is reduced to a qualitative interpretation of the laboratory test results.


Similarly, much of the knowledge of the products of spontaneous combustion, upon which the means of detecting self-heating in underground mines is based, depends upon the qualitative extrapolation of small scale laboratory tests to the in-situ situation. There is no doubt that much valuable knowledge has been obtained from these tests which has proven invaluable in improving methods of detection. The question still arises, however, as to what stage a heating has reached when it is detectable by mine air monitoring. This has never been addressed before. 


A series of numerical models based upon the oxidation and heat transfer processes occurring in spontaneous combustion were developed to simulate self-heating of coal, under a range of conditions, such as occur in laboratory tests and in in-situ heatings. Models were developed to simulate the Adiabatic Self-Heating (R70) test, the Relative Ignition Temperature (RIT) test and various configurations of in-situ heatings. Four in-situ models were developed representing different configurations and were used for various purposes during this study. 


The four models were:

•    The ID-Long model representing a column of coal in a mass of finite depth and infinite thickness and width, with depth of the pile considered to be parallel to the airflow; 

•    The ID-Wide model representing a column of coal in a mass of finite thickness and infinite depth and width; 

•    the 2D model representing a slice of coal in a mass of infinite width; and;

•    The Quasi-3D model representing a cylindrical mass of coal suspended in air. 


Note 1: For the purposes of this thesis "pile" is used to describe any accumulation of coal whether it is a stockpile on in-situ coal seam. 


Note 2: For the purposes of discussion, the depth of the coal mass is the dimension parallel with the direction of airflow. The thickness and width are perpendicular to the depth. 


The principal properties of coal used in the modelling were the characteristic oxidation rate, Ko, and the low temperature heat of oxidation, Ho(40°C). As all of the models were based upon the same oxidation and heat transfer assumptions, it was possible to relate the simulated laboratory test to the simulated in-situ heating characteristics of coals. 


The development of in-situ heatings was examined and it was found that this could be broken into three distinct phases being: 

•    The initial incipient phase in which the coal temperature rises gradually to about 125°C.      

•    The migration phase in which the coal hotspot migrates forward to the upwind surface of the coal pile and reaches a maximum temperature of 200°C or more. 

•    The final charring phase which occurs either when the hotspot reaches the upwind surface, or forms a charline within the body of the pile and which then begins to migrate downwind at very high temperatures in excess of 500°C. 


The impact on the self-heating characteristics of coal of various parameters such as airflow flux, pile thickness and mass was also examined. In relation to airflow flux through the pile, it was found that there was little impact beyond about 3 to 5 litres/min/m^. It was also found that, for any particular coal, there are a number of Critical Spontaneous Combustion Characteristics that can be used to assess the potential for spontaneous combustion.

These are: 

•    The Critical Self-Heating Temperature which is the initial coal temperature below which the coal will not self-heat, irrespective of the thickness or mass of the coal pile. 

•    The Critical Self-Heating Period, being the shortest time in which a spontaneous heating can develop.

•    The Critical Self-Heating Thickness being the minimum thickness required for the development of a self-heating. 

•    The Critical Self-Heating Mass being the minimum mass required for a self-heating to develop. 


The Critical Self-Heating Period, Thickness and Mass are dependent upon the particle size distribution and initial temperature of the coal, whilst the Critical Self-Heating Temperature is not. The variability of these characteristics for a wide range of likely coal oxidation properties was examined and related to the corresponding R70 and RIT results that were predicted from the modelling of these tests. This provided a means of predicting the in-situ self-heating characteristics of coal based upon the results of the R70 and RIT spontaneous combustion tests. A method of assessing the potential for spontaneous combustion, based upon the Critical Spontaneous Combustion Characteristics in any circumstances and which removes the qualitative nature of previous spontaneous combustion assessment methods, has been proposed. 


The variation of gaseous products of spontaneous combustion in relation to the development of a heating was examined using the Quasi-3D model. For a wide range of conditions involving changes to airflow flux, pile mass and coal oxidation properties, the development of various indicator gases such as carbon monoxide and hydrogen and various indicator ratios was determined and related to the temperature of the heating. It was found that there is no basis for the use of traditional trigger levels of carbon monoxide (CO) make for the early detection of spontaneous combustion. For heatings in a mass of 8 times the Critical Self-Heating Mass and for a wide range of coal properties, it was found that at the close of the incipient phase, when the peak temperature had reached about 125°C and the rate of self-heating was increasing rapidly, the CO make did not exceed about 0.6 litres/min. At 200''C, the maximum CO make observed was 2.5 litres/min. By the time CO make had reached any of the traditional trigger levels, the peak coal temperature of the heating was well above 250°C and in many cases was over 400°C. 


Hydrogen makes remain low through the development of spontaneous combustion until the charring phase is reached. It is therefore concluded that hydrogen, and all other indicator gases provide no warning of the early development of spontaneous combustion.   


Although the Graham's Ratio of the off-gases from a heating can be very high, this is not observable in a mine due to the effects of dilution with other airstreams. The magnitude of the change observable in the mine ventilation system depends upon the state and size of the heating, the airflow flux through the heating and total ventilation quantity. There can be no one trigger value for Graham's or any other indicator ratio. It is concluded that the indicators of the early stages of spontaneous combustion can only be increasing trends in CO make and Graham's Ratio.

Keyword Combustion, Spontaneous
Coal mines and mining -- Safety measures
Additional Notes

Variant title: Spontaneous combustion of coal.

Document type: Thesis
Collection: UQ Theses (RHD) - UQ staff and students only
Citation counts: Google Scholar Search Google Scholar
Created: Fri, 24 Aug 2007, 18:31:40 EST