Limit equilibrium and numerical modelling approaches in slope stability analyses with regard to risk assessment for open pit mining
Date
2016-03-15
Authors
Kanda, Mpoyi Jacques
Journal Title
Journal ISSN
Volume Title
Publisher
Abstract
The planning of open pit mines, and road and rail cuttings constitutes one of the activities usually
undertaken by geotechnical engineers. However, this endeavour faces major challenges such as the
correct design of rock slopes, the evaluation of their stability and the risk associated with them.
Two main analytical methods are used in this process: the Limit Equilibrium (LE) analysis and the
Numerical Modelling (NM) method; Slide and Phase2 programmes will be used respectively in
this regard. Previous studies have shown some discrepancies between their results in assessing the
probabilities of slope failure and the consequent economic risks. This research project aims to
understand the reasons behind these divergences and possibly to find some ways of reducing them.
To attain these objectives, a homogeneous slope model was adopted. It required a detailed
validation study depending on the analytical method, such that the settings would carefully be
calibrated to avoid any further misinterpretation of the results. For Phase2, attention was given
with regard to the number of mesh elements and their type, and for Slide, the number of slices. In
addition, for both methods, attention was given to the adequate distribution of rock mass variables,
the adequate failure criterion, etc. Deterministic and probabilistic assessments were performed to
better interpret the differences to be found from these methods and results. The response surface
methodology (RSM) facilitated the probabilistic studies, to avoid the constraints of long computer
run times and to ease the study of the influence of the rock mass parameters on the slope stability.
For the considered model, 25 slices and 1500 mesh elements were found adequate to better assess
the probability of failure (POF), while 1000 slices and 50 000 mesh elements provided results of
estimated failure volumes. Well defined distributions of rock mass variables have proven
indispensable to better assess the POF as well as the risk associated with the slope failure.
Application of the Hoek Brown criterion resulted in the LE analysis predicting higher failure
volumes than when Mohr Coulomb criterion was used. With the NM method, cares were taken not
to under or overestimate the resultant failure volume when extracting the path of the failure
surface. Deterministic assessments showed that the risk determined from NM analyses is not
always greater than that from LE analyses. The addition of rock mass parameters not taken into
account in LE analyses results does not sensibly influence the POF and the failure volume
outcomes, but can in some conditions influence the behaviour of the outcome risk of slope failure.
In essence, LE and NM methods can be relied on for probabilistic studies, or even for risk
assessments conditioned by carefully setting the models, and in case of LE being adopted to assess
the risk, it is recommended to introduce a multiplying factor for cases similar to those that have
already been analysed.
Description
A research report submitted to the Faculty of Engineering and the Built Environment, University of
the Witwatersrand, Johannesburg, in partial fulfilment of the requirements for the degree of Master
of Science in Engineering