Development of statistical models to predict emission rate and concentration of particulate matters (PM) for drilling operation in opencast mines

• Published in: Air quality, atmosphere and health atmosphere and health, 2019-09, Vol.12 (9), p.1073-1079
• Main Authors: Nagesha, K. V., Kumar, Harinandan, Muralidhar Singh, M
• Format: Article
• Language: English
• Subjects: Air pollution; Artificial neural networks; Atmospheric Protection/Air Quality Control/Air Pollution; Drilling; Dust; Earth and Environmental Science; Emission; Emissions; Environment; Environmental Health; Environmental impact; Environmental impact assessment; Environmental protection; Geographical variations; Health Promotion and Disease Prevention; Impact analysis; Mathematical models; Mines; Mining; Model accuracy; Neural networks; Open pit mining; Outdoor air quality; Particulate emissions; Particulate matter; Predictions; Statistical analysis; Statistical models
• ISSN: 1873-9318; 1873-9326
• DOI: 10.1007/s11869-019-00723-7

Air pollution in mining area is one of the critical concerns because of the generation of large amount of the particulate matters (PM) in opencast mining operations. The emission of PM in the air not only deteriorates the surrounding environment but also impacts adversely to the human health. The majority of the PM particles in opencast mining comprise PM2.5 and PM10. The assessment of the limit of the PM particles is very important as it helps in environmental impact analysis (EIA/EMP) and prediction of possible dust generation (PM particles) for any project to be established. In this paper, a model was developed that is capable to predict the respirable dust particulate concentration in the ambient air at various locations near and away from the dust-generating source, especially from a drilling operation in Indian opencast mines. The modeling was carried out using three different methods, i.e., “SPSS,” “R,” and artificial neural network (ANN) methods. Results from these developed models were compared with the US Environmental Protection Agency (USEPA) model for its validity. The predicted values from the developed model showed good correlation and the least variation from the field-monitored values, indicating better accuracy, compared with other models.