Automated Mining Horizon Control Using Real-time Coal Seam Sensing

This report describes the outcomes of an 18 month ACARP-funded research project that focussed on improving the performance of sensing technology for automated coal mining horizon control. Achieving effective horizon control is fundamental for improving mining safety and productivity. However automated mining horizon control remains an open problem for the industry. A major issue constraining the development of horizon control capability is a lack of instrumentation that can reliably measure the coal-strata geological structure in real-time. In response to this need, this research project focussed on three major objectives:

· To identify the major factors associated with sensing system performance;

· To develop more quantitative means to understand and describe performance; and to

· To describe what is practically possible to guide the use of radar-based geological sensing.

In order to practically understand achievable sensing performance, it was essential to obtain high quality data sets to provide ground truth validation and to improve knowledge. Four extensive survey campaigns were conducted at three different mine sites, each having unique geological and structural geometry. A broad range of radar equipment was used during the survey process. A detailed data analysis process was undertaken and improvements made with regards to interface tracking algorithms. The insights gained were captured and distilled into a simple decision support tool to assist industry gain a better awareness of current sensing capability.

Specific innovations and improvements made through the research include:

· A systematic identification of the major factors that influence sensing performance;

· Introduction of statistical measures as means to quantitatively characterise variability;

· Extensive validation through practical mine site survey campaigns and data analysis; and

· Development of a practical application guide to inform system selection process.

Whilst the report shows that radar is not a universal solution for all geological scenarios, greater clarity has been obtained regarding expected sensing performance as well as the reasons why. This expected sensing performance is summarised in the form of a Practical Application Guide.  Provided that the radar system conforms to the definitions and configuration outlined in Section 16, the Practical Application Guide can provide a rapid determination of the expected radar system performance for horizon sensing where:

· Green = good performance, with clear definition of the target horizon expected for 75% or greater of operational time deployed;

· Orange = marginal performance, with clear definition of the target horizon expected for approximately 50% of time deployed;

· Red = poor performance, where the target horizon can be ascertained for no more than 25% of operational time.

The report concludes with recommendations to identify ongoing opportunities to further improve horizon sensing technology in order to achieve safer and more effective mining operations.