Uncertainty in ground support design and implementation in underground mining

Abstract

In underground mining, there are a number of uncertainties in the ground support design process and during implementation of ground support designs. The minimisation of geological uncertainty is one of the rock engineering principles outlined by Bieniawski (1992) and by Stacey (2004, 2009). Generally, there are two main types of uncertainty; uncertainty due to naturally variable phenomena in time or space and that due to lack of knowledge or understanding (Baecher and Christian, 2003). McMahon (1985) listed six types of geotechnical uncertainty: risk of encountering unknown geological conditions; risk of using incorrect geotechnical criteria; bias and/or variation in estimated design parameters is greater than anticipated; human error; design changes and excessive conservatism. All of these types of uncertainty are encountered in the design of ground support. They occur in the form of uncertainties around: the design block size; loading conditions; spatial variability of ground conditions; rock mass strength; shear strength; discontinuity spacing; discontinuity orientations, etc. Uncertainty is not only present in the design of ground support but also in the implementation phase. This occurs in the form of variations in installation quality; adherence to patterns and spacing; human error in the application of the correct ground support standard; not identifying a change in ground conditions, etc. Conventional deterministic and empirical design methods do not adequately cater for uncertainties in the design of ground support. Probabilistic design methods such as the Point Estimate Method and Monte Carlo Simulation can be applied to better understand the uncertainties in the design process. Limiting uncertainties in ground support implementation can be achieved by: training of operators and supervisors; good supervision of implementation; and a well-considered and managed quality control programme. It is important to feed the results from the quality control and monitoring programmes back into the design. In this paper, the various uncertainties with ground support design and implementation will be reviewed and discussed in context of McMahon (1985) and Baecher and Christian (2003). The application of probabilistic design methods in ground support design and feedback of quality control data will be discussed. 2 Ground support design and implementation processes There are a number of well-documented design schemes for ground support design. Generally, they all require the following:  Description of the rock mass and identification of likely failure mechanisms.

Ground support installation quality controls and possible pitfalls: a case study from a critical fall of ground contract rehabilitation project

“…Specific shorter-duration contract projects rely on significantly less data or site-specific experience, but likely have more flexibility to select support products, installation equipment, and establish processes using best practice. In both cases, there are limitations associated with the design and implementation of the support (Dunn 2013). In the paper, Dunn discussed the risk associated with the support design and the implementation process.…”

“…A substantial amount of information is summarised in a minimal number of pages and details perceived to be unnecessary are excluded. Table 1 Uncertainty in ground support implementation (Dunn 2013) made where necessary e.g. variance in spin time when installing resin rebar with the mechanical bolters and jacklegs.…”

Quantifying uncertainty in mining geomechanics design

“…The steps required to minimise uncertainty and reduce errors have previously been summarised by various authors (Dunn et al 2011;Hadjigeorgiou & Harrison 2011;Hadjigeorgiou 2012;Dunn 2013Dunn , 2015; this includes ensuring that:…”

Deterministic and probabilistic stability analysis of a mining rock slope in the vicinity of a major public road — case study of the LAB Chrysotile mine in Canada

“…The results of such an approach could be conservative or not. Furthermore, using worst-case scenario inputs can be misleading, as the combined occurrence of worst-case values is expected to be low (Valerio et al 2013;Dunn 2013) Sensitivity analyses can be used to determine the impact of input parameter variability on the resulting FoS. However, these types of analyses are not well suited to address the probability of occurrence of the various scenarios.…”

“…The most widely used design acceptance criteria are still FoS based, although the use of PoF is increasingly popular (Dunn 2013;Duncan and Wright 2005). Steffen et al (2008) demonstrate the impact of uncertainty on the obtained FoS distribution.…”