Assessment of Damage and Risk from Seismic Events in Coal Mines

Seismic monitoring is typically undertaken when a mine is experiencing dynamic failure events such as coal or rock bursts, caving anomalies or rock failure associated with structural features which is causing uncertainty and difficulty during mining. The aim of the monitoring is to understand the location and nature of the rock / coal failure which is associated with dynamic events.

In many cases, the location will provide sufficient information to understand the cause and provide a management strategy. In other cases, location accuracy may not be sufficient to provide the necessary information, and as such a more detailed analysis of the nature of the failure is required to understand and manage the events. The greater detail is derived from tensor analysis and review of the S and P wave arrivals.

This project has focused on the issues surrounding recording and interpreting seismic data and understanding the mechanisms of rock failure which cause seismic events.  A key aspect of the study has been to simulate the most likely areas about a longwall panel for which dynamic events may occur. A case study of a strain burst event was simulated and compared with aspects of the seismic signature which results.

Seismic wave data is recorded on geophones and interpreted to provide information of the source event and the mechanism that created the event.

Two levels of interpretation can be obtained from the data; location of events; and moment tensor of the event.

The location of an event is resolved from triangulation from the sensors which pick up the wave form. The P wave and S wave velocity is assessed from calibration shots in the area and the effects of ray paths errors are estimated based on the known location of the shot relative to the geophone location. The triangulation is based on the geometry of the sensors and the arrival time of the wave at each sensor. Therefore, the distance of the event from each sensor is a function of the difference in arrival time and the velocity of the wave. In this way, the location of the seismic event can be estimated. This is discussed in detail within Section 5 together with its inherent limitations.

The location of events is often the default presentation of the data where the location of events is the primary requirement of the monitoring program. The plots do not provide any data on the stress changes and the nature of the failure which was associated with the event, but only represent a location where an event of some sort has happened.

The moment tensor analysis requires clear waveforms and an inversion of P and S wave data will allow derivation of:

• The nature of a seismic event in terms of the orientation of stress components;
• The nature of normal stresses in terms of explosive or implosive characteristics; and
• The potential failure surfaces, this level of analysis requires good seismic waveform transmission in the rock mass and a suitable array of sensors to describe the directionality of the stress / displacement components.

This type of analysis is invaluable for rock mechanics investigations of mining problems. It is required in circumstances for which the location plots do not provide sufficient accuracy or sufficient information to describe the nature of the failure which is associated with the seismic event.

The application of these procedures to locate and assess the rock mechanics processes that caused the seismic events is discussed with reference to the sites studied in the project.

The site at Austar was modelled to assess the rock failure modes and most likely sites of seismic activity. This data was compared to monitored data to provide a greater level of interpretation of the rock failure modes and potential burst mechanisms or seismic damage.  The risk factors noted were discussed and are presented as a guide for assessment of other sites.