Prediction of blast-induced ground vibration using multivariate statistical analysis in the opencast chromite mines of the Indian State of Odisha

The production processes in open cast mining activities are primarily associated with rock blasting operations, which trigger blast-induced ground vibrations causing significant damages to the surrounding surface structures. Regular monitoring of the ground vibration levels serves as crucial input to the regulatory authorities and policymakers for designing and developing an efficient blasting system to control the extent of damages caused by the same. Thus, the present study aimed to monitor the field practiced blasting events in active open cast chromite mines in the Sukinda valley of the Indian state of Odisha to develop a new site-specific model for predicting blast-induced ground vibration. The investigation has proposed a new ground vibration prediction model using multivariate statistical analysis. The validation of the model-predicted results against the field recorded data suggests the reliable performance of the proposed model having R², RMSE, MAE and MAPE values of 0.8, 2.8, 1.925 and 0.5, respectively. Whereas, the comparison of the field recorded data with the universally employed USBM model exhibits R², RMSE, MAE and MAPE values of 0.28, 8.13, 4.72 and 0.68, respectively suggesting the better predictive capabilities of the proposed model. Hence, the present study suggests the applicability of a newly proposed predictive model for blast-induced ground vibration in the chromite mining region of Sukinda valley. Finally, one such design parameter, namely the maximum explosive charge per delay, has been optimized for different distances by limiting the PPV value within the legislative limits as per guidelines formulated by the regulatory authority of India.