Mines have an inherent risk of geotechnical failure in both rock excavations and tailings storage facilities. Geotechnical failure occurs when there is a combination of exceptionally large forces acting on a structure and/or low material strength resulting in the structure not withstanding a designed service load. The excavation of rocks can cause unintended rock mass movements. If the movement is monitored promptly, accidents, loss of ore reserves and equipment, loss of lives, and closure of the mine can be prevented. Mining companies routinely use deformation monitoring to manage the geotechnical risk associated with the mining process. The aim of this dissertation is to review the geotechnical risk management process to optimize the geotechnical risk management analysis. In order to perform a proper analysis of slope instability, understanding the importance as well as the limitations of any monitoring system is crucial. Due to the potential threat associated with slope stability, it has become the top priority in all risk management programs to predict the time of slope failure. Datasets from monitoring systems are used to perform slope failure analysis. Innovations in slope monitoring equipment in the recent years have made it possible to scan a broad rock face in a short period with sub-millimetric accuracy. Instruments like Slope Stability Radars (SSR) provide the quantitative data that is commonly used to perform risk management analysis. However, it is challenging to find a method that can provide an accurate time of failure predictions. Many studies in the recent past have attempted to predict the time of slope failure using the Inverse Velocity (IV) method, and to analyze the probability of a failure with the fuzzy neural networks. Various method investigated in this dissertation include: Minimum Inverse Velocity (MIV), Maximum Velocity (MV), Log Velocity (LV), Log Inverse Velocity (LIV), Spline Regression (SR) and Machine Learning (ML). Based on the results of these studies, the ML method has the highest rate of success in predicting the time of slope failures. The predictions provided by the ML showed ~86% improvement in the results in comparison to the traditional IV method and ~72% improvement when compared with the MIV method. The MIV method also performed well with ~75% improvement in the results in comparison to the traditional IV method. Overall, both the new proposed methods, ML and MIV, outperformed the traditional inverse velocity technique used for predicting slope failure.
| Identifer | oai:union.ndltd.org:arizona.edu/oai:arizona.openrepository.com:10150/625550 |
| Date | January 2017 |
| Creators | Chandarana, Upasna Piyush, Chandarana, Upasna Piyush |
| Contributors | Momayez, Moe, Momayez, Moe, Slack, Donald C., Tenorio Gutierrez, Victor Octavio, Zhang, Jinhong |
| Publisher | The University of Arizona. |
| Source Sets | University of Arizona |
| Language | en_US |
| Detected Language | English |
| Type | text, Electronic Dissertation |
| Rights | Copyright © is held by the author. Digital access to this material is made possible by the University Libraries, University of Arizona. Further transmission, reproduction or presentation (such as public display or performance) of protected items is prohibited except with permission of the author. |
Page generated in 0.0029 seconds