Discover in this 3 part series the technical foundation of the Advanced Simultaneous Optimization (Advanced SIMO) module in GEOVIA Whittle™. Advanced SIMO module enables mine planners to create optimal long term schedules for the operation of open pit mines.1 Advanced SIMO uses the ProberB engine that was developed by Whittle Development Pty Ltd.
By D. Place, G. Whittle, and N. Baxter
Introduction
Traditional methods used in strategic mine planning tend to optimize one parameter at a time, while Advanced Simultaneous Optimization (Advanced SIMO) considers all parameters and alternatives simultaneously, hence providing a near-optimal solution. Previous solutions that used Milawa for schedule optimization, Stockpile & Cut-off Optimization (SPCO) for cut-off and stockpile optimization, and multiple extractive blend scenarios to optimize blending are all superseded through the use of Advanced SIMO which optimizes these parameters all together to maximize profit. The traditional pit-shell optimization (Lerch-Grossman or Pseudoflow) is still required to create the optimal pit shell. Pit shells are mostly independent of the schedule and stockpiling, and therefore can be done independently of an Advanced SIMO optimization. While this statement could be argued, the slope constraints and economics remain by far the main drivers for the shape of pit shell rather than the schedule. A simultaneous optimization (SIMO) run will optimize one and only one scenario of pushback strategy. If one wants to optimize the pushback selection and strategy that optimization could be done separately using the pushback chooser, or better, with sufficient computing resource by manually evaluating the different strategies, through skin analysis for instance. From a set of input parameters (block model, economics and mining constraints, etc.), a user wants to calculate a schedule that maximizes profit through Net Present Value (NPV). The NPV is formulated as a function of those input parameters, the schedule and all the other variables. The schedule is the variable that has the most degrees of freedom and has the greatest impact on NPV. The schedule provides a description of which block to mine and when with the detail of its destinations (i.e.,if it is to be processed, stockpiled or discarded). Finding the optimal mine plan can be basically reduced to finding the optimal schedule. The optimal schedule for a mine can be described with a suite of linear equations and inequations with an objective function representing the NPV after the life of mine. The aim of the optimization is to solve all the variables and input parameters that maximize the objective function. If such a system is linear, and small enough, it can be solved using a traditional linear programming (LP) solver and an optimal solution is guaranteed. Unfortunately, in the case of most mine schedules, the system is neither small enough nor linear.
Why is simultaneous optimization better than traditional methods?
The basic assumption made when optimizing one parameter separately from others is that the parameters are independent of each other. For instance, cut-off optimization assumes that the cut-off grades can be optimized independently of the schedule. However, changing the cut-off often means that the schedule that was once optimal may not be optimal for a different cut-off. When optimizing in separate steps, a decision made at an early stage (e.g., Schedule in Figure 1) impacts later decisions, hence reducing or masking possibilities of a better solution altogether. The following sections, Steps 1 through 3, will briefly describe the logic in deriving an optimal mining schedule with SIMO.
In our next post, we will introduce the 1st step to find the optimal schedule "Aggregation to blend bins". Stay tuned!
Your may also be interested in the best practice already released:
GEOVIA Whittle Best practice
