Multi-fractal conditional simulation of fault populations in coal seams using analogues: Method and application

Abstract

The modelling of fault populations and quantification of fault risk is a challenge for earth science and engineering applications, including minerals and coal mining, tunnel construction, forecasting of petroleum reservoir production, and selection of subterranean repositories for the disposal of toxic waste. This paper discusses a new advance in the use of stochastic fault simulation methods for the quantification of the risk of fault presence. The multi-fractal properties of a fully known fault population are used as an analogue of the properties of an undiscovered fault population. The approach is elucidated through the quantification of fault risk in a prospective coalfield at Wyong, New South Wales, Australia, and incorporates spatial patterns of available ‘hard’ and ‘soft’ geological data. The method does not find faults unequivocally; rather, the output is a map of fault probability. Simulations are found to be consistent with the available information and are statistically and spatially reasonable in geological terms. Significantly, the analogue approach provides a robust, quantified assessment of fault risk using limited exploration information.