Ross Seedsman, Seedsman Geotechnics
Getting the geotechnical model right at the feasibility stage is the biggest issue facing Australian longwall mines.
The technical risks in planning longwall mines are the same as for all mines – geology, geotechnical and overly optimistic scheduling assumptions. The geological risks can be addressed through the JORC process but there are major weaknesses in the geotechnical component.
During pre-feasibility and feasibility, a geotechnical model needs to be formulated and then analysed to identify a range of options from which a recommendation can be made.
The recent trend has been to advocate monitoring of the geotechnical conditions in the initial driveages and the first longwalls so as to calibrate various models of behaviour. Such monitoring is simply good practice and cannot be considered to be an alternative to basic engineering assessment and design.
At feasibility stage, the investor needs to be confident that the value of the project will not be negatively impacted by the early failure of the geotechnical systems.
Major purchasing and operational decisions are required before geotechnical monitoring can begin. These include the choice of development systems and particularly the orientation and location of gateroads. The relatively frequent presence of geological faults in Australian coal fields, compared to, say, the USA, means that bolter miners will probably remain the preferred option.
Aligning gateroads to minimise the concentration of horizontal stresses above pillars near the maingate corner may not be the most important consideration. Whilst installing high levels of support in the maingate can be expensive, it is important to recognise installing support or otherwise stabilising a longwall face is even more expensive and difficult. It is possible that the key orientation decision should be based on the alignment of the longwall face itself.
The location of gateroads also needs to be considered at the feasibility stage: are narrow panels needed to control weighting cycles from massive overburden units or to limit surface subsidence, what chain pillar sizes are needed, and where should they be located if a lower seam is to be mined? And is the seam suitable for longwalling at all – could the floor be too weak?
It is perhaps noteworthy that longwall capacity has not been included in this discussion. This is because the knowledge on the loading of supports by the overburden rock mass is not adequate to allow the formulation of a geotechnical design approach. Geotechnical engineers can probably do little to stop or slow down the trend to larger and larger supports even though a critical review of recent longwall history may suggest that other aspects of the mine design, and particularly the alignment of the longwall face to joints and cleats, are more significant controls on face performance.
A commitment to improving the geotechnical model at feasibility stage will have added bonuses. It will provide a framework to better interpret the monitoring data that is subsequently obtained. Monitoring in a coal mine is not a controlled experiment. The data obtained is impacted by a huge range of geological and geotechnical variables and there is a danger that a wrong interpretation will be developed unless the ground is adequately characterised. The question then becomes: what is adequate? This can only be answered in the context of a geotechnical model.