Sustainable air pollution management in coal mining through low-cost sensors and smart monitoring platform.

Abstract

Open-pit coal mining is a significant source of air pollution, especially in developing countries, where dust and gaseous emissions from mining operations often exceed regulated limits due to limited mitigation measures and environmental controls. This study introduces a scalable, Internet of Things (IoT)-based real-time air quality monitoring system implemented in the Jharia coalfields, India, using low-cost sensors across three zones: mine site (industrial), mine office (buffer), and IIT-ISM (sensitive). The system monitored PM₁, PM_2.5, PM₁₀, CO₂, NO₂, TVOCs, temperature, and humidity over one year. ANOVA and Tukey's HSD confirmed statistically significant differences in pollutant concentrations between monitoring zones (p < 0.05), with the mine site consistently recording the highest levels. Seasonal analysis showed winter peaks due to thermal inversion, while monsoon conditions supported pollutant washout. Diurnal PM_2.5 peaks were observed at 6:00 AM and 6:00 PM, coinciding with peak operational activity. High PM_2.5/PM₁₀ (0.86-0.91) and PM₁/PM_2.5 (0.80-0.84) ratios indicated dominance of fine and ultrafine particles. Correlation analysis showed AQI strongly correlated with PM (r ≈ 1.0), while temperature showed a moderate negative correlation in winter (r ≈ -0.65). CO correlated moderately to strongly with PM (r = 0.6-0.7), while O₃ showed weak or negative correlations, indicating photochemical influences. Time-series forecasting using machine learning models identified LightGBM with higher accuracy followed by Random Forest and XGBoost. The study contributes to sustainable mining practices by integrating meteorological data into a scalable IoT-based monitoring framework. It also delivers actionable insights for environmental compliance and health risk mitigation in mining regions.