Prediction of spontaneous coal combustion tendency using multinomial logistic regression

Area of Science:

• Mining Engineering
• Geological Engineering
• Safety Science

Background:

• Spontaneous coal combustion poses significant risks to underground mine safety and operational efficiency.
• Factors influencing spontaneous combustion include gas concentrations, ventilation, and coal properties.

Purpose of the Study:

• To predict spontaneous combustion tendencies in Turkish underground coal mines.
• To identify key parameters affecting coal mine fire hazards.

Main Methods:

• Multinomial logistic regression, a multivariate statistical technique, was employed.
• Gas concentrations (methane, carbon dioxide, oxygen) and air velocity were analyzed as predictive factors.
• Coal mines were classified into 'normal situation' and 'potential combustion' hazard levels.

Main Results:

• Methane (CH4) and carbon monoxide (CO) concentrations, along with their interaction (CH4 × CO), were found to be significant factors in determining combustion risk.
• Mine I was identified as having a higher propensity for spontaneous combustion compared to Mine II and Mine III.
• The study examined the impact of variations in these factors on spontaneous combustion risk.

Conclusions:

• Gas composition and airflow are critical indicators for assessing spontaneous combustion risk in underground coal mines.
• The predictive model highlights Mine I as requiring more stringent safety protocols.
• Further investigation into factor variations can enhance mine safety management strategies.