Underground » Geology
Through previous ACARP projects C11037 (Hatherly et. al, 2003) and C15019 (Hatherly et al., 2008) a new methodology for rating rock masses was introduced. Termed the Geophysical Strata Rating (GSR), this empirical scheme was developed for the clastic rocks typically found in the coal mining districts of Australia. The GSR is calculated from borehole geophysical logging data.
This current project was directed towards continuing that work through the following objectives:
· Extend the Geophysical Strata Rating (GSR) to allow assessment of coal roof;
· Produce 3D geotechnical models of GSR based on the analysis of geophysical borehole logs and utilising coal seam picks from any available 3D seismic data or from borehole correlations;
· Demonstrate the application of the GSR in production settings; and
· Make appropriate recommendations and presentations to facilitate the uptake of the GSR as a routine tool for the analysis of geotechnical data.
In extending the GSR to accommodate coal, we have also had to consider carbonaceous material other than coal. For these carbonaceous materials, the present rating scheme for clastic rocks can be applied, provided the porosity is set to zero. For coal, previous research shows that coal brightness can be related to coal strength. We have assumed that the ash profile may be related to brightness and hence be related to the mechanical properties of the coal.
Relationships between ash content and the density, natural gamma and sonic velocity log responses are known to exist. Analysis of all three types of logs based on simple mixing laws, indicates similar apparent proportions of ash. The GSRi for coal is based on the estimate of the apparent ash. It is scaled so that stronger, high ash coals can have GSRi's greater than other carbonaceous material while the GSRi for bright coal is typically less.
In the normal GSR calculations, the defect score (the combined bed and fracture scores) is between 0 and 20. In the case of coal, the total defect score has been reduced to between 0 and 10. This is because horizontal weakness planes are often less well developed in pure coal seams and hence less dominant in their failure behaviour. Cleating normally plays the dominant role, which is governed by the GSRi estimate.
One of the advantages of a GSR analysis is that the results can be modelled to reveal spatial variations in GSR. Geologists across the entire mining sector are familiar with 2D and 3D modelling aimed at mapping variations in quality and grade. With continuous GSR data, equivalent modelling operations are possible and the results can be interrogated to reveal geotechnically significant variations in rock property. All of our results provided in this report are in the form of geotechnical models and our discussion relates to the spatial variations revealed by that modelling. We report on results from our three main sites - Crinum, Moranbah North and Newlands Northern underground mines.
At Crinum Mine, we have compared GSR results with UCS determinations from sonic logs and the mapping of geotechnical units with the roof horizons. At Crinum there is a strong correlation between roof conditions and rock strength. However manual processing is required to identify specific roof units, combined with rock strength estimates to provide a roof hazard classification. The GSR analysis was tested and able to mimic this process without manual identification of the various roof units.
In one case, the GSR analysis was able to highlight an area of poor roof performance that had not been detected prior to mining. Through preparation of plans and sections of GSR values, GSR calculations provide the means to assess specific strata influences through the lithological column as well as in broader plan view plots prepared for hazard planning. Other advantages for using the GSR include its fully computerised analysis and the ease in which data can be stored, updated and interrogated as required.
At Moranbah North Mine, our study targeted an area of known poor roof conditions over the tailgate of LW105. 2D and 3D models were developed that clearly showed the geological factors affecting immediate roof stability and caving behaviour. These included a significant thickening in strong sandstones, seam splitting and a weak rider seam all coinciding over an extended area.
The model provided detailed information of the location of heavy roof and how it coincided with the fall area along the tailgate. Previous analysis also suggested that the presence of the rider seam under these conditions can lead to particularly challenging roof conditions that require specific ground support requirements. The GSR analysis provides a detailed representation of the ground conditions that would have contributed to the falls and accurately identifies the zones in which such roof behaviour would have been triggered.
The analysis for Moranbah North was undertaken using data that was available prior to the instability. Whilst roof monitoring alone could not have been used to verify the proposed roof failure mechanisms, the geophysics-based analysis provides ample evidence to support the proposed roof failure mechanism. With such information available prior to extraction, the mine may have been able to identify hazards, improve monitoring strategies and if necessary undertake additional analysis to incorporate into the strata management plan.
At the Newlands Northern Underground, GSR and clay content analysis provided a convenient means to link geological and geotechnical characteristics of the overlying strata that contributed to weighting events and ultimately roof falls at Newlands. In this case the analysis was used to identify the key strata conditions relevant to face weighting.
In order to investigate the usefulness of GSR analysis, equivalent mid-panel sections were developed using the borehole dataset. Again, a main point to note is the difference in detail that can be obtained from the GSR analysis when compared to sections obtained from manual processing. In particular, changes in character of the overlying Marker Mudstone were identified that could not be detected unless detailed manual scrutiny of each borehole by an experienced professional was undertaken. Similarly sandstone channels were identified in the vicinity of faults which could be correlated with significant changes in the general structural nature of the coal seam and immediate roof strata that were observed underground.
Over the course of the project we have had the opportunity to study data from several other mines in the Sydney and Bowen Basins. We have not encountered problems with our formulation with the results from any of these mines. However, problems can arise with poor quality geophysical logs. Sonic logs containing spikes suggestive of unrealistically low and high velocities present one problem. Another can be noise in density and natural gamma logs. Geophysical logs provide an excellent opportunity for quantitative measurement of rock properties but if the logs are unreliable that opportunity is lost.