MONITORING MICROSEISMIC OF “MO” UNDERGROUND MINING USING PHASENET, GAMMA, AND NONLINLOC PROGRAMS

• Main Author: Ruth E Lantang, Adrianna
• Format: Final Project
• Language: Indonesia
• Online Access: https://digilib.itb.ac.id/gdl/view/76588
• Institution: Institut Teknologi Bandung

The increase of seismicity in seismic areas and the emergence of seismicity in aseismic areas is a result of underground mining. Tunneling and block caving are mining activities to access and extract mineral resources. Tunneling and block caving can pose accident risks, such as tunnel collapse, rock bursts, slope instability in open pit mines, and seismic hazards. Seismic activity in mining areas varies widely from small to large-scale earthquakes associated with damage from seismic events. Microseismic is seismic wave activity that has a small size and is not necessarily felt by humans. Therefore, microseismic monitoring is needed in underground mining areas to minimize losses both in material and casualties. In this study, the data used are 50 microseismic events of the underground mine "MO" for the period November-December 2022. At the initial stage of the study, the phase selection of P and S waves was carried out using the algorithm of the PhaseNet program. The PhaseNet program shows good performance in determining the arrival times P and S waves. PhaseNet is also able to determine the arrival time of P and S waves in poor waveforms data. However, the PhaseNet program also has limitations, namely the existence of multiple picking. The existence of multiple picking can affect the accuracy and quality of the picking results, so event associations using GaMMA are needed. In this study, 4 experiments were conducted with the dbscan parameters used in the association were 0.1; 0.2; 0.5; and 1.0. And the most optimal result is 0.5. This is because the events that can be used at DBSCAN 0.5 are 41 events with an average difference in arrival time of 0.691 seconds. The 41 events used in NonLinLoc have differences in latitude, longitude, and depth that are quite far for microseismic cases, namely 0.420 km, 0.604 km, and 0.765 km. The differences in P and S wave arrival times and locations between the "MO" mine catalog data and the PhaseNet and GaMMA data are caused by problems in extracting waveforms and the assumed homogeneous velocity model.