Resource modeling experts,from the model to the plant.

Conditional simulation & plan risk

Quantify the geological, grade and geometallurgical uncertainty behind budget plans, schedules and production commitments.

Typical problems

• Production commitments rest on a single deterministic model
• No measure of how likely the plan is to be met
• Unknown dilution and misclassification at the scale you mine

How we approach it

• Plurigaussian and indicator simulation of geology
• Sequential Gaussian and multivariate (PPMT) simulation of grades
• Risk by period, phase, bench and dig polygon
• Expected dilution and probability of meeting the plan

Resource estimation & independent reviews

Build, update or audit long-term, short-term and grade-control models, with reconciliation as the test of truth.

Typical problems

• Long-term models that do not reconcile with blastholes or the plant
• Hard boundaries, capping or indicator thresholds that bias estimates
• Monthly models that are slow and hard to repeat

How we approach it

• Geological and estimation-domain review, contact analysis
• Compositing, capping, variography and estimation plans
• Indicator, ordinary, simple and multigaussian kriging
• Locally varying anisotropy (LVA) and blasthole back-testing

Reconciliation, ore control & stockpiles

Track tonnes and grade from the pit to the plant: blast movement, dig polygons, run-of-mine stockpiles and reconciliation factors.

Typical problems

• Stockpiles that feed the plant but are poorly characterized
• Mine-to-plant reconciliation gaps and drillhole/blasthole bias
• Manual workflows spread across several disconnected tools

How we approach it

• 3D stockpile block models from fleet management and topography
• Blast movement and displaced-material models
• F-factor audits and tonnage/grade traceability design
• Automated workflows refreshed with operational data

Geometallurgical modeling

Predict recovery, throughput, hardness and concentrate quality in 3D, with an honest measure of how reliable each estimate is.

Typical problems

• Sparse, unevenly sampled metallurgical test data
• Process design and plans that ignore ore variability
• Geometallurgical domains that are hard to define and defend

How we approach it

• Stochastic imputation of missing variables
• Machine-learning domains (PCA, UMAP, clustering)
• Multivariate estimation and simulation of metallurgical responses
• Process equations for throughput, energy and recovery

Drilling optimization & classification

Decide where to drill next and how to classify resources using uncertainty, not only distance to samples.

Typical problems

• Inferred resources and limited drilling budgets
• Classification that is hard to justify
• Infill campaigns that do not target the real risk

How we approach it

• Simulation-based classification and drill-spacing studies
• Infill and exploration drillhole placement optimization
• Separating interpretation risk from grade risk
• Risk maps for the periods that matter

Capability building & training

Guidelines, peer review and hands-on supervision so your team can build and defend its own models.

Typical problems

• Thin internal teams under delivery pressure
• Practices that differ between sites and modelers
• New methods that never make it into routine work

How we approach it

• Modeling guidelines and peer review
• Supervision of your team's model updates
• Courses in geostatistics, simulation and Python
• Workflows handed over on the Annapurna platform