Stochastic Pit Optimization

Most, if not all, commercial pit optimizers in the market today are exclusively implemented based on a deterministic pit optimization algorithm such as the classic Lerchs & Grossmann or the newer Pseudo Flow. Such a method uses a single, deterministic block model as input and cannot account for the stochastic nature of mineral projects.

By contrast, Stochastic Pit Optimization expects a group of block model realizations as input and thus allows the mine planner to quantify, in a single pit optimization run, geological uncertainty and risk inherent in any mineral resource project.

GeoMine-FlowPit, our ultra-fast pit optimizer, has been extended to provide our clients with a Stochastic Optimization option. The newly enhanced FlowPit can now accept a group object containing an arbitrary number of block model realizations resulting from a single block model simulation run, e.g. the SGS_Run in the screenshot below:

In the above example, the input group object "SGS_Run" contains 5 block model realizations resulting from a single Sequential Gaussian Simulation run. Such an object can also be manually constructed by dragging and dropping individual distinct block models into an empty group object in GeoMine. The result of a stochastic pit optimization run is displayed as a composite plot where the X-axis is # of Block Model Realizations and the Y-axis represents multiple corresponding values, i.e., Profit, Ore Tonnage, Waste Tonnage and Metal Quantity. The available statistics for each value include:

1. total
2. average
3. variance
4. standard deviation.

Needless to say, these statistics provide much more insight into the project being evaluated than a single average typically provided by a single deterministic pit optimization run.

Work is under the way to integrate the stochastic pit optimization process with iScheduler-OP (our Strategic and Tactical Open-pit Scheduler in GeoMine) for stochastic evaluation of life-of-mine schedules for open-pit mines.