Automated Stochastic Workflows

Recent advancement of GeoMine has led to the introduction of three automated and practical stochastic mine planning workflows.

Every mine plan is built on an estimate, and every estimate is uncertain. A single kriged block model hides that uncertainty; a set of conditionally simulated realizations exposes it. GeoMine’s stochastic workflows take that set of realizations and turn it into something a planner can act on: one consensus design that is robust across the realizations, and an NPV distribution (P10/P50/P90/Mean) that tells you how much confidence to place in it.

The same three-phase pattern is applied to all three mining methods:

Phase 1 – Optimize every realization. The relevant GeoMine optimizer (FlowPit, Stopemizer or Cavemizer) runs once per realization. Votes are accumulated from all realizations.

Phase 2 – Build the consensus design. Votes are used with user-specified threshold to produce a single consensus pit, stoping limit or caving footprint.

Phase 3 – Schedule every realization against the consensus design. The design is held fixed and the scheduler runs once per realization, producing a per-realization NPV and a self-contained HTML risk report.

The result is a design decision made with the whole realization set in view, not just the expected case.