Trying to predict the unpredictable
Coal burst is a sudden and destructive failure of coal walls in underground coal mines responsible for multiple injuries and death worldwide.
To date, it has caused two fatalities in Australia and several safety incidents.
It is one of the Australian coal industry’s biggest safety issues and one that is growing as mines go deeper underground in the search for resources.
Coal burst occurs suddenly with few warning signs, making it very difficult to predict.
Much like earthquakes – which can’t be predicted either – the reasons coal burst occurs is not well understood, making prediction and monitoring very difficult.
When it happens, coal burst is a violent explosion. The coal mass fragments and is ejected from the solid wall at high speed.
Monitoring and modelling of geological structures
We have developed technologies to predict and monitor the risk of coal burst around major geological structures.
Through field observation and monitoring we’ve identified key issues associated with coal burst.
For example, many coal bursts occur near a natural fault or intrusion. The closer a mine wall is to such an area, the higher the stress concentration, which can lead to coal burst.
Numerical modelling paired with comprehensive field monitoring is helping us characterise the stress concentration near major geological structures in operating coal mines.
This knowledge is helping us get closer to being able to monitor and predict the risk of coal burst by understanding the influence of different geological structures.
Solid steps forward
We’ve undertaken five Australian Coal Industry Research Program (ACARP) funded projects and further research in partnership with the University of New South Wales School of Minerals and Energy Resources Engineering.
This work has accumulated quantitative data on stress concentration near major geological structures for the first time.
Using CSIRO-developed microseismic and stress-monitoring techniques we can accurately detect fracturing inside rocks.
Our sensitive geophones and stressmeters have been able to detect even very small rock failures many hundreds of metres away from the instruments.
We have been able to validate our numerical modelling techniques, finding our modelling can predict similar stress response as found in field monitoring and also predict stress change near major geological structures.
Using real data from coal mines, we have done a case study < can we talk specifically about this case study? e.g. We have worked with Glencore in the Hunter Valley modelling geological faults around their coal mines, producing a risk assessment for coal burst>.
We are continuing to expand our understanding of coal burst risk, and are extending our work to include other cases with different geological structures.