WACEIO Projects: Mining

Australia has world class deposits of most major mineral commodities and is a major world producer of coal and many metals. With an annual turnover of around $40 billion the mining industry is the Nations largest earner. Australia also has significant reserves and production of oil and gas products. The economic viability of the mining and petroleum industries is highly dependent on efficient practices. Accurate mathematical modelling along with effective optimisation provide the opportunity to productivity gains. The broad research categories of blending, planning, scheduling, materials handling, resource management, production and optimisation of mine design are applicable to the mining and petroleum industries.

The WACEIO team (L Caccetta, S Hill, V Rehbock, L Giannini, and R Collinson) has considerable expertise and experience in these areas of research. Some examples where progress has been made are detailed below.

• Open Pit Mine Scheduling
  (WACEIO : L Caccetta and S Hill; MineMap : P Clifford; Hammersley Iron : M Slavik and S Potter; Rio Tinto : P Welgama).

  The operation and management of a large open pit mine is an enormous and complex task, particularly for mines having a life of many years. Though a number of optimisation techniques have been successfully applied to solve some important problems (for example : ore-body modelling and ore reserve estimation; the design of optimum pits; the determination of optimal blends) the fundamental problem of determining an optimal production schedule over the life of the deposit that satisfies a variety of physical and economic constraints (for example: mine extraction sequence; mining; milling and refining capacities; grades of mill feed and concentrates; minimum pit bottom widths) has, prior to our recent work, remained unresolved. The many available methods are very deficient in that they do not accurately model the operation and cannot cater for most of the constraints. Often the solution provided to industry is not feasible. The WACEIO team have developed an accurate mixed integer linear programming model and a computationally effective solution strategy using an innovative Branch and Cut approach.

  A software package implemented in C++ and containing around 30,000 lines of code has been developed and extensively tested on operating mines. The team has demonstrated that provably good solutions are obtained for practical sized problems. On gold mines our methodology produced solutions that yield an increase of at least 15% in the NPV profit. We have also developed a customized version of our software for specific iron ore mines. Again we obtain quality solutions to very large mines. This product has no competitor. Our more accurate modelling coupled with fast solution techniques provides the mining industry with powerful mathematically based tools for managing its resources. Significant economic benefits are attainable. This work certainly has strong commercial potential which we are following up on this opportunity.

• Mine Site Rehabilitation
  (WACEIO : L Caccetta, L Giannini and P Kelsey)

  Over the past decade or so mine site rehabilitation has become increasingly important. Indeed, mining leases today include conditions relating to the preservation of the surrounding environment and the rehabilitation of the mine site once the mine has been completed. There are two basic requirements of mine site rehabilitation

  1. that the site is safe, stable and non-eroding at the end of the life of the mine; and
  2. pollutants such as acid-producing waste be buried and capped with highly impermeable material.
  Clearly mine site rehabilitation entails extensive material movement and careful scheduling is required to optimise the process. The first of the above requirements gives rise to the land surface reshaping problem which is defined as finding the surface which satisfies the requirements (usually expressed in terms of wall slopes) and which minimises the material movement. We developed an integer linear programming model. A solution method using the subgradient optimisation technique has been developed and tested on producing mines. Our method provides powerful tools for mine planners. This is a considerable improvement over existing procedures which are essentially manual. This work formed part of P Kelsey's Master Thesis. The methodology developed has applications in a number of other areas such as highway design.

• An Application of Optimisation Techniques to the Truck/Loader Selection Problem
  (WACEIO : L Caccetta and S Hill; Rio Tinto : P Welgama, B Martin and N Taylor).

  An important problem in mining is that of selecting a fleet of trucks and loaders for use in extracting ore and waste throughout the life of the mining operation. The cost of the truck and loader fleet has been estimated as being up to 55% of the total cost of the operation making purchasing the correct combination of trucks and loaders critical. The problem is of obvious significance to the Australian economy, both in terms of export dollars and cost savings, given the scale of our mining and mining related industries. Initially a large purchase of trucks and loaders for the removal of ore and waste must be made. Minor changes to this fleet may occur as the mine plan changes but essentially the bulk of the fleet is purchased at the beginning of the operation. Note that fleet equipment has a lifecycle of around three to five years and so is replaced regularly. Usually though the same equipment type is purchased to keep repair costs low. Current methods for determining the best trucks and loaders to use in a mining operation largely rely on the experience of specialist consultants with computational methods restricted to the use of spreadsheets and/or simulation. Due to the complexity of the problem only a small subset of the possible combinations of trucks and loaders may be considered for selection using these methods. The application of accurate mathematical modelling and cutting edge optimisation techniques, where the optimisation is done over all possible truck and loader combinations, promises to lead to better cost savings while ensuring effective choices of equipment.

  The aim of this project is to develop effective mathematical models for determining the optimal truck/loader selection strategy for use in mining operations as well as optimisation techniques for solving the underlying problems. The complimentary strengths of the research team will allow the development of models of sufficient accuracy, with quick solution, so that mining engineers will use the resulting computational package. The results obtained to date are very positive and the team is confident that a successful commercial software package will result. The team has applied for an ARC Linkage Grant.