An EDP-Model Of Open Pit Short Term Production Scheduling Optimization For Stratiform Orebodies

The best use of available reserves to give good metal recoveries relies on short term mine production scheduling.-The aim is usually to produce a plan to guide mining in specific areas that will enable mine operations to deliver the budgeted tonnes and grade to the mill and keep within the long range mine plan. Thus, short term production scheduling is an important aspect of the open pit mine planning function. Since the 60's increased attention has been paid to effective production scheduling as a means of improving the economic viability of existing mines. Hence, mine scheduling scheduling systems have been evolving rapidly during the last two decades. These systems include manual techniques, computer based simulation, mathematical programming and more recently Computer Aided Design (CAD). Some general drawbacks of the current scheduling systems are that firstly, the costs of acquiring and using them are not at all insignificant especially for small to medium sized pits. Secondly, there is an inherent trial-and-error aspect in these systems. Thirdly, they tend to be suited for specific orebody geometries, hence lacking the necessary generality. For instance, the use of computer programs for planning open pits to mine stratiform orebodies raises problems which are not usually met when massive orebodies are considered. There for, a need exists to develop simpler but effective short term pit scheduling procedures suitable for stratiform type orebodies, considering their geometry and geology. This paper describes a PC- based model designed to assist the planning engineer in optimizing the short term production schedule in open pit mines with stratiform orebodies. The model is based on a combination of linear goal programming and deterministic mining simulation. The orebody model is made up of ore and waste mining regions where winning and and/or stripping are taking place. These shovel regions form the basic planning elements in the model. This approach does not require the laborious preparation and use of a regular block model. The computer program consists of main modules: (i) Data capture - processes the raw data into the required format for mine planning. (ii) Precedence mining constraints - checks the status of each shovel .region at the beginning of a given scheduling period with regard to geometrical mining constraints such as undercutting. (ii) LP- Optimization - provides an optimal selection of ore shovel regions which do not violet geometrical constraints for mining in the period, the objective, being to maximize the contained metal equivalent in the blended R-O-M ore and at the same time to minimize deviations in grade and tonnage from the call values. (iv) Post LP-Simulation - enables the user to smooth out the LP solution in order to adjust it to a practically executable plan and simulates mining of ore and waste for the period. Shovel allocation to selected ore and waste shovel regions, unit operations and depletion of reserves are executed in this module. Preliminary testing of the model for monthly production scheduling using data from Nchanga OPen Pit on the Zambian Copperbelt indicates that it can be a useful planning tool and gives acceptable results.