Geostatistical Modelling of Floor and Roof Hazard Data in the Highveld Coalfield – A New Denmark Colliery Case Study

Abstract

New Denmark Colliery (NDC) is an underground coal mine located between Standerton and Bethal in the Mpumalanga province of South Africa. Due to the nature of the coal seam and the mining method in the research study area, mining operations are in constant contact with floor and roof lithologies which introduce operational challenges. To mitigate these challenges floor and roof hazard plans are used for operational planning purposes. Existing floor and roof hazard plans are deficient of both sound theoretical and procedural formality in their construction. This lack of formality in the hazard modelling was most concerning. The research establishes a methodology for the generation of meaningful floor and roof hazard plans from empirical and theoretical applications to both interpreted and measured variables relevant to the floor and roof hazards encountered. This methodology; developed for easy operational implementation, is supported by a formal system of procedures that allow for continuous updating, validation and monitoring. The quantitative and qualitative hazard data available in the study area required bridging. For this, the author introduces the concept of applying scorecards to the hybrid data and develops the scoring logic to convert the hybrid hazard information to numeric values, usable in quantitative analyses. The estimation software in place at NDC is limited to inverse distance weighting (IDW). The research therefore sought to determine whether the application of this classical technique would suffice for the creation of hazard plans. IDW, ordinary kriging and conditional sequential Gaussian simulation were applied to measured structural variables. The estimation results were visually compared. This would then confirm the suitability of applying IDW to the research data. The research takes a turn when the hazard scorecard numbers become the focus of the research as opposed to the structural variables. Floor and Roof hazard scores are individually classified, respectively representing either iv “competent”, “moderately competent” or “incompetent” floor conditions and “normal”,” cautionary” or “high risk” roof conditions. An innovative method of defining hazard indicators sets is introduced. IDW estimation is applied to each hazard indicator set. Results are interpreted, and inflection points on the slopes of the cumulative distribution plots of the estimates are used to identify cut-off values to clearly distinguish the hazard conditions mentioned. Resulting in updated Floor Hazard and Roof Hazard Plans for NDC aligned to a formal analytical process and estimation methodology. With this in place, the research goals were accomplished through the construction and validation of reliable and easily implementable floor and roof hazard plans. Procedures for applying hazard scorecards to newly drilled boreholes and for updating the hazard models appear in the Appendices.