In situ stress data management and analysis: implications for numerical modelling

Abstract

The measurement of in situ stresses is a costly and time intensive procedure, requiring significant experience, diligence and planning to obtain reliable results. Therefore, it is important to have good data management and analysis to maintain investment costs and reduce estimate uncertainty. Geotechnical design considerations rely on accurate estimates of the stress regime to predict subsurface rock behaviour. Some uses include cave deconfinement, cave preconditioning, planning drive orientation and ground support design. Two simple numerical models are presented. The first demonstrates stress redistribution around a cave zone and how it may influence optimal crusher chamber placement. The second model is of an undercut level and the stress experienced by the drawbell drives below. The models demonstrate the impact of in situ stress uncertainty, with potential consequences for design decisions by altering the stress trend and plunge. Currently, pragmatic workflows are not publicly available that holistically detail the key steps in in situ stress data analysis. In addition, rarely are end-users of the data demonstrated the uncertainty in the values provided. Instead, they are presented with a single stress average gradient and trend. This manuscript offers a workflow for those embarking on in situ stress testing campaigns and serves as a tool to communicate the complexities to potential end-users. The workflow utilises the Euclidian mean rather than scalar averaging of the stress tensor and relates 'loading methods' (hydrofracking) and 'other' (borehole breakout observations) methods of stress estimation with the full tensor resolved from relief methods (overcore and acoustic emissions). A software application within the mXrap suite is presented as a potential aid to stress data management and analysis. Implications of stress uncertainty are demonstrated and discussed concerning design, expected induced fracture orientations, and future workflow enhancements. The numerical models indicate optimal placement for a crusher chamber around a simulated cave zone was observed to be in the orientation of SHmax. Increasing the stress field plunge by 20° impacted optimal crusher placement. The second numerical model demonstrates that a rotation of the horizontal stress trend by 20° has noticeable impact on the differential stress in the backs of a tunnel where drawbells will be created, this has implications for planning the undercut veranda length.