Prediction of Occurrence and Location of
Inclusive Dates: 07/01/96 - 04/30/99
Background - Rockbursts are serious worldwide problems in deep underground mines. These rockburst problems usually involve violent failures and ejection of large quantities of rock, ranging from a fraction of a cubic meter to thousands of cubic meters. The seismic energy associated with the rock ejection process can reach the equivalent of a magnitude five earthquake as designated on the Richter scale. Catastrophic rockbursts can lead to fatalities, injuries, damage to mine facilities, premature mine closures with abandonment of large reserves, and loss of production. Rehabilitation cost of rockburst damage often amounts to tens of thousands of dollars. Although systematic observation and research have improved the understanding of rockburst mechanism(s) over the years, they still occur and remain a serious problem in the mining industry.
Approach - This program represents the first of a three-phase approach to develop a systematic methodology for rockburst prediction so that early warning can be provided to mine management for appropriate actions to reduce damage and save lives. This program involves performing laboratory experiments in a controlled setting to develop a methodology by analyzing patterns of microfracture formation in rock samples under compression. The microfracturing process is monitored by measuring the time and source locations of the acoustic emission (AE). The basic approach adopted involves analysis of fractal characteristics of the microfracturing process of rock. The focus will be placed on assessing the spatial and temporal variations of fractal dimension and fractal intercept to identify patterns and correlating these patterns with the extent of microfracturing clustering so as to develop a precursor to predict failure. In the second phase, the methodology developed from the first phase will be validated using field data. A follow-up, proof-of-concept program will be proposed if the first phase is successful. The third phase will be the application of the methodology to field problems. The primary focus in this phase is development of a knowledge-based system for real-time analysis and warning. The third phase study is expected to be supported by the mining industry.
Accomplishments - Work for the first phase of the program was completed. Controlled laboratory compressive experiments of cylindrical specimens were conducted to investigate the microfracturing process that leads to formation of macrofracture planes and the subsequent sample failure. AE events associated with the microfracturing process were analyzed using two algorithms for determining the source locations of the events. One algorithm minimizes errors in source location estimation, and the other minimizes errors in estimated travel times. Both algorithms appear to qualitatively locate events in the regions in which actual fracture planes were observed. However, neither algorithm is able to describe fully the fracture planes formed. This inability can be attributed to the uncertainty associated with the velocity anisotropy in the sample and errors related to recorded arrival times. Source location prediction uncertainties may be reduced by quantifying velocity anisotropy through measuring velocity changes along predetermined fixed paths and the errors in arrival time measurements through appropriately selecting a trigger threshold amplitude. The AE microfracturing events were also analyzed for fractal characteristics (fractal dimension and prefactor ) in the temporal and spatial domains. The results indicate that some time before a sample fails, either the fractal dimension or prefactor or both peak in temporal domain and subsequently decrease. This phenomenon may be used as a precursor to provide early warning of a rockburst. Both the fractal dimension and prefactor should be used for this purpose. However, if the time interval between the peak of a fractal characteristic and the failure is too short, the early warning will not be sufficient, in practical terms. The variation of fractal dimension in the spatial domain has the potential to identify formation of the primary fracture plane and the subsequent secondary fracture planes. When microfracturing events coalesce to form the primary fracture plane, the fractal dimension drops to near two, indicating a planar geometry. This behavior is also a good indicator for rockburst prediction. In general, the available warning time associated with this indicator (spatial indicator) is relatively shorter than that for the temporal domain (temporal indicator). Consequently, the temporal indicator can be used as a tool for decision making in applying preventive measures, while the spatial indicator may be used for evacuation purposes. More work remains to be done so that this technique can be applied in the field.