Integrated optimization of stope boundary selection and scheduling for sublevel stoping operations

T Copland

doi:10.17159/2411-9717/2016/v116n12a7

Abstract

The decreasing availability of resources amenable to surface operations has led to increasing numbers of underground mines, with trends indicating this will continue into the future. As a result, there is a need for additional optimization processes and techniques for underground mines, with many analogous methods having already been developed for surface mining. Current methods for design and optimization of stope boundary selection and scheduling mainly involve heuristic methods which focus on a single lever. Individual optimality may be approached, but globally optimal results can be obtained only by an integrated, rigorous approach. In this paper we review previous methodologies for stope boundary selection and medium- to long-term scheduling and highlight the need for an integrated approach. Previous integrated approaches are reviewed and an improved modelling system proposed for shorter solution times and greater applicability to mining situations. Randomly generated data-sets for gold-copper mineralization are used to investigate the model performance, describing solution time as a function of data complexity.

Highlights

The decreasing availability of resources amenable to surface operations has led to increasing numbers of underground mines, with trends indicating this will continue into the future
There is a need for additional optimization processes and techniques for underground mines, with many analogous methods having already been developed for surface mining
The current industry standard for medium- and long-term planning for sublevel stoping (SLS) is to exclusively determine stope outlines and determine an appropriate sequence and schedule

Summary

Integrated optimization of stope boundary selection and scheduling

The current industry practice is to manually create stope boundaries and schedule extraction separately. The branch and bound method is an algorithmic approach that uses a LR of an integer program to evaluate subproblems to determine the optimal solution of an IP problem. A non-integer variable is chosen and its bounding integers form constraints for the sub-problem to be solved This process continues until all possible solutions are investigated and the optimal solution is found. IP is usually only limited in application due to the increased complexity of obtaining solutions, as this form of programming can model many real-life problems more accurately than LP The combination of both continuous and integer decision variables in a programming problem is termed mixed integer programming (MIP). The branch and bound technique utilizes type-two special ordered sets to determine the optimal starting and finishing locations for stope blocks in a row.

Table I

Tonn ZZn LVl t MGn g TonLim MMaxg MMing

Findings

Table III