Post-Collapse Seismicity of the Crandall Canyon Mine Using Double Difference Relocations

摘要

In 2007, the University of Utah Seismograph Stations (UUSS) temporarily expanded its seismic monitoring network in Emery County, Utah, in response to the August 6 Crandall Canyon mine collapse that trapped six workers. In spite of this additional coverage, no seismic events were located in the area during a six-day period (gap) from August 8 to August 13. The Mine Safety and Health Administration (MSHA) "Command Center Log Books" however, contain references to 16 instances of underground "bumping" and "bouncing" that occurred during the gap. When ground truth information provided by the log books is compared to the raw seismic data recorded by the UUSS network, the correlation offers insight into the chronology of mining-induced seismicity (MIS) in the aftermath of the collapse. The new research focuses on identifying and locating seismic events that occurred during the six-day gap using double-difference relocation techniques and correlating them with log book ground truth. Ten seismic events were identified in the vicinity of the mine during the six-day gap, ranging in magnitude from coda magnitude (M_c) 0.5 to 1.4. Three of these events were amenable to location by double-difference techniques. The location of one of these events is directly confirmed by ground truth information from the log books. As a result of this new research, the accuracy of Crandall Canyon double-difference locations has been improved, and the number of events correlated with ground truth information has increased. The seismic events occurring in and around the collapsed area appear in two distinct sections of the mine--east and west of the center of collapse. It was originally reported that the post-collapse events occurred in episodic clusters. The first two episodes occurred to the east, followed in time by a cluster on the west, and then by a third episode on the east. With the addition of new events to the data set, this pattern is no longer as clear, although the events still occur in clusters. Only one MIS cluster was found to occur within the bounds of the minimum estimated collapse area (Pechmann et al., 2008).