Identifying “signatures” of slope failure conditions in open pit mines to support the set-up of alarms: a possible workflow

Comprehensive slope monitoring programs and early warning systems are commonly implemented in open- pit mines for the mitigation of slope failure risk. Despite the availability of massive datasets of monitoring data in modern open pit mines, often acquired by multiple sensors, negative events associated with slope failures, having an impact on both mine safety and productivity, are still occurring. A limited exploitation of the available information, due to difficulties faced by site staff in the aggregation of the different monitoring data and in their interpretation, plays a crucial role. Moreover, systematic procedures to the set-up and calibration of alarms are lacking and several key factors are often overlooked in the majority of the surface operations. The most common approach for the alarm set-up is based on engineering judgment, which is typically based on previous personal experience of the site staff. In this regard, the experience gained through the comprehensive back-analysis of monitoring data of past slope instabilities at the same site, if analyzed with the proper methods, can provide a set of essential information for the alarm set-up. With the ultimate goal of supporting the tuning of an alarming system, the initial conceptual framework of a process for the identification of “signature” parameters differently describing failures and non-failures, through the back analysis of monitoring data, is proposed in the present paper. Several cases of slope instability monitored by means of ground-based radar at undisclosed open-pit mines, some of which evolved into failure, whereas others, although showing intense phases of deformation, ultimately did not, are used to show how the proposed method can provide new insights on the characteristics of the precursory deformation in rock slope failures, thus supporting the set-up of an early warning system.