A rock engineering systems based model to predict rock fragmentation by blasting

A new model for prediction of rock fragmentation by blasting is presented based upon the basic concepts of rock engineering systems (RES). The newly proposed approach involves 16 effective parameters on fragmentation by blasting with keeping simplicity as well. The data for 30 blasts, carried out at Sungun copper mine, western Iran, were used to predict fragmentation by the RES based model as well as Kuz–Ram and multiple regression modeling. To validate the new model, the fragmentations of nine production blasts were measured and the results obtained were compared with the predictions carried out by the RES, Kuz–Ram and multiple regression models. Coefficient of determination (R2) and root mean square error (RMSE) were calculated for the models to compare the results obtained. For the RES, linear, polynomial, power, logarithmic, exponential and Kuz–Ram models, R2 and RSME are equal to (0.65 and 14.51), (0.58 and 29.73), (0.54 and 21.58), (0.60 and 32.64), (0.61 and 23.80), (0.50 and 184.60) and (0.46 and 22.22) respectively. These indicate that the RES based model predictor with higher R2 and less RMSE performs better than the other models.

► An RES (rock engineering systems) based model was used to assess risk and predict rock fragmentation by blasting.
► The level of fragmentation risk obtained shows a good consistency with the fragmentation measured.
► A predictive equation with R2=0.55 was obtained for rock fragmentation.
► The results obtained were compared with regression and also with Kuz–Ram models.
► Evaluation performance was carried out, showing the merit of the model proposed over regression modeling and Kuz–Ram model.