Air Pollution Risk Assessment Using a Hybrid Fuzzy Intelligent Probability-Based Approach: Mine Blasting Dust Impacts

Mine blasting discharges considerable amounts of dust, which may impose environmental risks, especially to nearby ecosystems. The present study aimed to develop a hybrid risk assessment approach using the concepts of artificial intelligence, probability, and fuzzy numbers to assess the risks of blast-induced dust emissions to ecosystems surrounding surface mines. An equation was first developed using gene expression programming to predict dust emission distance. Monte Carlo based on the predictor equation was then applied to simulate the dust emission phenomena. New fuzzy-based tables were set based on the Monte Carlo results, wind analysis, and expert knowledge to quantify the risk factors in failure mode and effects analysis. The developed approach was applied to assess the blast-induced dust risks in a small-scale limestone mine, close to residential and farm areas. The gene expression programming equation resulted in the best possible model with R2 and RMSE of 0.9091 and 5.4594 for training, and 0.8754 and 7.0181 for testing, respectively. The Monte Carlo simulation indicated that the dust emission distance does not exceed 199.1 and 212.2 m with confidence levels of 90% and 99%, respectively. The results indicate that the dust risk on farms was negligible to low based on distance range of 97.21–423.75 m, which was higher compared to dust risk on humans with a distance range of 4.57–67.66 m. All cases were indicated to be safe, however, because of the non-toxicity of limestone dust and the long time interval between blasting rounds.