NEW PREDICTION MODEL OF BLAST-INDUCED PPV BASED ON EMPIRICAL APPROACH IN SEDIMENTARY ROCK

Abstract

Prediction models of peak particle velocity (PPV) in blasting operations have been developed by many researchers, but all the previous research mostly referred to the concepts of waves and rocks. There has not been much development of the PPV prediction model based on an empirical approach specific to sedimentary rock blasting. This study aims to build a PPV prediction model in sedimentary rocks based on empirical and statistical approaches. The study was conducted on sedimentary rocks, which have hardness ranging from 1 MPa to 5 MPa. The analysis used the monitoring result of PPV from blasting operation and statistical methods to develop the PPV model. The PPV model based on the U.S. Bureau of Mines (USBM) as a theoretical model was compared with the PPV model based on field data as a proposed model. The PPV obtained from the USBM model did not fit the rock characteristics based on the actual data, so a new model was proposed. The new model considers the distance between the observation point and blast hole (D) and the quantity of explosive charge (Q) as independent variables. The new model has a mean absolute percentage error (MAPE) of 25.50%, 1.72% better than the USBM model. By considering the explosive charge (Q) and the distance to the observation point (D) in modeling with regression analysis, seismic velocity and rock mass density can be represented by site-specific condition constants. The site-specific constants refer to the rock constant and propagation constant, which are 5670.882 and 0.6269 in the new model that has been built.