Calculator D5

Geometallurgical Block Modeling for Process Integration

Geometallurgical block modeling is like making a 3D map of a mine that shows not just where the rock is, but how it will behave in the processing plant — so mining and processing teams can work together smoothly.

Typical Block Scale
10 × 10 × 5 m (open pit); 5 × 5 × 2.5 m (underground)
Industry Adoption
Used in >65% of Tier-1 copper & gold operations (ICMM 2023 survey)
Data Volume
10⁶–10⁸ assay points per major deposit; >90% require MLA/QEMSCAN
Standards Alignment
Complies with CIM Definition Standards (2023) & ISO 14067 for carbon footprint attribution

⚠️ Why It Matters

1
Ore heterogeneity unquantified
2
Process recovery misestimated
3
Mill throughput constrained by unexpected gangue
4
Reagent overconsumption and environmental liability
5
Short-term stockpile blending fails
6
Long-term reserve value eroded

📘 Definition

Geometallurgical block modeling is a spatially explicit, multi-domain engineering methodology that integrates geological, geotechnical, metallurgical, and hydrological data into a unified 3D numerical model. It quantifies ore variability (e.g., mineralogy, texture, liberation, hardness, acid consumption) at mine-scale resolution (typically 5–20 m blocks) to enable predictive process performance simulation and closed-loop grade control. The model serves as the central decision-support framework for mine planning, mill feed optimization, and real-time circuit adaptation.

🎨 Concept Diagram

Geometallurgical Block ModelDomain ADomain BDomain C→ Integrated Process Response Surfaces

AI-generated illustration for visual understanding

💡 Engineering Insight

Never treat geometallurgy as 'geology plus metallurgy' — it is a *process physics interface*. The most accurate block model fails if its properties aren’t tied to first-principles process responses (e.g., d₅₀ must be linked to liberation-based recovery models, not empirical grade-recovery curves). Successful integration hinges on calibrating each property’s functional relationship to unit operation output — not just statistical correlation.

📖 Detailed Explanation

At its foundation, geometallurgical block modeling replaces static resource estimates with dynamic, behaviorally informed representations of ore. Unlike traditional grade models that predict metal content alone, geometallurgical models encode how ore *responds* — whether it grinds easily, liberates cleanly, consumes acid aggressively, or floats selectively. This requires disciplined sampling: composite intervals must preserve textural relationships (e.g., avoiding lithological mixing), and assays must be fit-for-purpose (e.g., MLA over bulk XRD for liberation prediction).

Advanced implementation demands rigorous uncertainty propagation: each property (Wi, d₅₀, AC) carries distinct spatial continuity and measurement error profiles. Kriging with external drift (using lithology as covariate) outperforms ordinary kriging by 30–50% in Wi prediction accuracy. Critically, the model must be *actionable*: block attributes are translated into process response surfaces using mechanistic or semi-empirical equations (e.g., Austin’s grinding model, Taggart’s flotation rate equation), not lookup tables.

State-of-the-art systems embed digital twin capabilities: real-time sensor data (on-belt NIR, XRF, laser diffraction) continuously updates local block property distributions via Bayesian updating. This enables true closed-loop control — for example, detecting a 0.8% increase in dolomite content triggers automatic lime addition ramp and thickener rake torque setpoint adjustment. Such systems require co-location of mine planning, process control, and data science teams — not just shared software.

🔄 Engineering Workflow

Step 1
Step 1: Define geometallurgical domains (lithology, alteration, structure, geochemistry)
Step 2
Step 2: Collect domain-specific drill core assays (XRF, MLA, QEMSCAN, AC, Wi, d₅₀)
Step 3
Step 3: Spatially co-locate assay data with 3D geological model using kriging or machine-learning interpolation (e.g., RF, GPR)
Step 4
Step 4: Populate block model with domain-weighted property distributions (not single-value averages)
Step 5
Step 5: Link block properties to process response functions (e.g., recovery vs. d₅₀, acid consumption vs. CaCO₃%)
Step 6
Step 6: Simulate mine production schedule against circuit capacity constraints using dynamic scheduling engines (e.g., Whittle-GEMS, Gurobi-MillLink)
Step 7
Step 7: Deploy real-time feedback loop: on-belt XRF → block model update → mill setpoint adjustment (via DCS API)

📋 Decision Guide

Rock/Field Condition Recommended Design Action
High carbonate + low sulfide (%S_total < 0.4%, AC > 6 kg/t) Route to low-acid leach pad; adjust lime addition pre-leach; segregate from sulfide-rich blocks
Fine-grained pyrite intergrown with chalcopyrite (d₅₀ < 35 µm, Wi > 18 kWh/t) Increase SAG mill charge level; add pebble crusher bypass; implement ultra-fine grind circuit staging
Variable clay content (>8% smectite, d₅₀ > 120 µm, Wi < 9 kWh/t) Apply selective flocculation pre-thickening; install high-rate thickeners; modify flotation pH setpoint to 9.2–9.6

📊 Key Properties & Parameters

Liberation Size (d₅₀)

15–250 µm

The particle size at which 50% of a target mineral is liberated from gangue minerals, measured via QEMSCAN or MLA.

⚡ Engineering Impact:

Directly determines optimal grind size and SAG/ball mill power draw; deviation >20% causes significant recovery loss.

Bond Work Index (Wi)

6–25 kWh/t

A measure of ore resistance to grinding, derived from laboratory ball mill tests, expressed as kWh/tonne to reduce material from theoretically infinite size to 100 µm.

⚡ Engineering Impact:

Primary input for mill sizing and energy forecasting; error >3 kWh/t leads to 8–12% throughput mismatch in SAG circuits.

Acid Consumption (AC)

0.5–12 kg H₂SO₄/t

Amount of sulfuric acid (kg/t) consumed by carbonate and reactive silicate minerals during leaching, determined via static/dynamic acid demand tests.

⚡ Engineering Impact:

Controls reagent cost, solution chemistry, and downstream impurity precipitation; unmodeled AC spikes cause pH collapse and copper/uranium co-precipitation.

Sulfide Mineral Content (%S_total)

0.1–8.5 wt%

Total sulfur mass percent reported by Leco combustion analysis, serving as proxy for sulfide mineralization intensity and potential ARD risk.

⚡ Engineering Impact:

Drives flotation reagent dosing, tailings storage design, and environmental compliance strategy; ±0.3% error shifts collector dosage by 15–25%.

📐 Key Formulas

Grind Size – Recovery Relationship (Liberation Model)

R = R_max × [1 − exp(−k × (P₈₀ / d₅₀)^n)]

Predicts metal recovery (R) as function of P₈₀ (80% passing size) and liberation size (d₅₀); k and n are ore-specific constants.

Typical Ranges:
Copper porphyry
k = 0.8–1.4, n = 0.9–1.3
Gold oxide
k = 1.2–2.0, n = 0.7–1.0
⚠️ P₈₀ ≤ 2.5 × d₅₀ ensures ≥90% of theoretical max recovery

Acid Demand Estimation

AC ≈ 0.014 × CaO_% + 0.021 × MgO_% + 0.008 × FeCO₃_%

Empirical acid consumption estimate based on carbonate oxide weight percentages.

Typical Ranges:
Carbonate-rich skarn
CaO = 1.2–22.5 wt%, MgO = 0.3–8.1 wt%
Low-carbonate porphyry
CaO = 0.1–1.8 wt%, MgO = 0.05–0.9 wt%
⚠️ Validate with static acid demand test if CaO > 3.5 wt% or MgO > 1.2 wt%

🏭 Engineering Example

Cadia East, New South Wales, Australia

Porphyritic dacite & quartz monzonite
Clay Swelling Index
12 mL/g (smectite-dominant)
Bond Work Index (Wi)
14.3 kWh/t
Acid Consumption (AC)
2.1 kg H₂SO₄/t
Copper Liberation (%)
76% at 106 µm
Liberation Size (d₅₀)
42 µm
Sulfide Mineral Content (%S_total)
1.8 wt%

🏗️ Applications

  • Mine plan optimization under processing constraints
  • Stockpile management for consistent mill feed
  • Environmental liability forecasting (ARD, neutralization demand)
  • Digital twin deployment for autonomous circuit control

📋 Real Project Case

Open Pit Gold Mine Blast Optimization

Large copper mine expansion in Chile

Challenge: High vibration levels affecting nearby structures
Read full case study →

🎨 Technical Diagrams

Geometallurgical Domain MappingLithologyAlterationGeochemistry
Real-Time Feedback LoopOn-Belt XRFBlock Model UpdateDCS Setpoint Adjust

📚 References

[1]
CIM Geometallurgy Guidelines — Canadian Institute of Mining, Metallurgy and Petroleum (CIM)
[2]
Geometallurgy: A Practical Approach — Society for Mining, Metallurgy & Exploration (SME)