An evaluation of the CUSUM and inverse velocity methods of failure prediction based on two open pit instabilities in the Pilbara

摘要

Predicting the expected time of slope collapse is an important aspect of managing open pit slope stability as it determines the appropriate actions to be taken. It is important to know when to evacuate but is also useful to know well in advance if a particular slope is creeping towards collapse or whether the deformations measured are unlikely to result in collapse. Having this type of information well in advance allows a mine to plan and execute remedial actions, such as schedule changes, slope angle changes and buttresses, that will mitigate economic as well as safety risks. While several methods of analysing slope monitoring data have been published to date, none have been able to establish themselves as the definite answer to any of these questions.This paper evaluates several previously published methods of predicting the time of slope collapse based on the monitoring data collected for slope instabilities that occurred at two of Rio Tinto's Pilbara Iron Ore operations in 2009 and 2010. The methods tested against the data are: CUSUM, inverse velocity and the slope of velocity and time multiplied velocity (SLO) method (Mufundirwa and Fujii, 2008). The paper concludes by evaluating the effectiveness of each of these methods to serve as early warning of impending failure and to predict the onset of collapse.Both instabilities were managed without injury to personnel and no loss of equipment.