Calculator D1

What is Mine Metallurgical Process Integration?

It’s how mining and processing teams work together like a single machine—using real-time ore data to adjust blasting, hauling, and grinding so nothing gets wasted and everything runs smoothly.

Typical Scale
Integrated loops operate at 1–10 minute cycle times across 5,000–50,000 t/d operations
Industry Standards
ISO 14064-1 (GHG accounting), SME Mining Engineering Handbook Ch. 12, AMIRA P964 Guidelines
ROI Horizon
12–18 months for brownfield integration; 3–5 years for greenfield digital twin deployment

⚠️ Why It Matters

1
Ore heterogeneity not quantified
2
Misaligned mine block models and mill feed assumptions
3
Overgrinding of soft zones / undergrinding of hard zones
4
Reduced metal recovery and increased energy consumption
5
Lower net smelter return (NSR) per tonne
6
Accelerated wear on comminution equipment

📘 Definition

Mine Metallurgical Process Integration (MMPI) is the systematic engineering discipline that unifies geology, mining, mineral processing, and automation through bidirectional data exchange and control logic. It operationalizes ore variability characterization—spatially resolved grade, hardness, mineralogy, and liberation—as a dynamic input to both mine planning (e.g., selective mining units, blast design) and plant control (e.g., crusher settings, flotation reagent dosing). Its core enablers are digital twin frameworks, grade control feedback loops, and closed-loop process optimization anchored in geological realism.

🎨 Concept Diagram

Mine PlanningProcessing PlantAutomation LayerMine-Metallurgical IntegrationClosed-loop feedback: Grade → Blending → Grind → Recovery → Updated Block Model

AI-generated illustration for visual understanding

💡 Engineering Insight

MMPI isn’t about adding more sensors—it’s about eliminating decision latency between ore exposure and process response. At Cadia East, reducing the grade control loop from 4 hours to 9 minutes cut copper recovery variance by 37%—not because assays improved, but because the flotation bank responded *before* the next truckload arrived. The bottleneck is rarely measurement accuracy; it’s control architecture topology.

📖 Detailed Explanation

At its foundation, MMPI treats ore not as a static feedstock but as a time-varying, spatially distributed process variable—like temperature or pressure in a chemical reactor. This requires shifting from deterministic resource estimation to probabilistic, geostatistically constrained block models where each 5 m³ unit carries distributions for grade, hardness, and mineral association.

Operationally, MMPI relies on three tightly coupled layers: (1) the geological layer (domain mapping, structural controls on ore continuity), (2) the mining layer (selective extraction, real-time ore tracking via RFID/RTK-GNSS), and (3) the metallurgical layer (adaptive circuit control, model-predictive controllers tuned to ore-specific breakage and liberation functions). Data fusion occurs at the 'digital twin interface', where ore trajectory (from shovel GPS to mill feed chute) is synchronized with assay, hardness, and mineralogical predictions.

Advanced MMPI implementations embed physics-informed machine learning—e.g., convolutional neural networks trained on core scan images to predict d₅₀_lib directly from drill-hole gamma-ray logs—or hybrid digital twins where DEM-based comminution models are updated in real time using live SAG mill acoustic spectra and torque signatures. These systems require rigorous uncertainty propagation: a ±0.15 GVI error can induce ±8% NSR volatility if unaccounted for in blending logic.

🔄 Engineering Workflow

Step 1
Step 1: Geological domain modeling with litho-geochemical domains (e.g., using Leapfrog Geo)
Step 2
Step 2: In-situ and ROM ore characterization (XRF, portable LIBS, drill-core MLA)
Step 3
Step 3: Dynamic block model updating with grade/hardness/liberation attributes
Step 4
Step 4: Mine plan optimization incorporating metallurgical constraints (e.g., Gurobi-based scheduling with recovery curves)
Step 5
Step 5: Real-time grade control loop: ROM conveyor belt analyser → DCS setpoint adjustment → mill circuit response validation
Step 6
Step 6: Weekly reconciliation of predicted vs. actual recovery, energy, and reagent use
Step 7
Step 7: Feedback-driven refinement of domain models and control logic (e.g., Bayesian updating of OHI–UCS correlation)

📋 Decision Guide

Rock/Field Condition Recommended Design Action
High GVI (>0.35) + Low OHI (<0.85) Deploy real-time XRF/XRD at ROM pad; route high-grade, soft ore to primary crusher bypass; blend with harder, lower-grade ore in surge bin to stabilize SAG feed.
Low d₅₀_lib (<40 µm) + High ANC (>5.0 kg/t) Increase leach residence time by 20%; reduce grind P80 to 65 µm; add pre-neutralization stage with slaked lime ahead of cyanidation.
OHI >1.25 + GVI <0.20 Optimize SAG mill ball charge to 12–14%; increase mill speed to 78% CS; eliminate secondary crushing; shift to direct-to-mill haulage routing.

📊 Key Properties & Parameters

Ore Hardness Index (OHI)

0.7–1.5 (dimensionless, reference = 1.0 for benchmark ore)

A normalized, plant-calibrated metric derived from SAG mill power draw and throughput, reflecting relative grindability of ore batches.

⚡ Engineering Impact:

Directly sets SAG mill speed, ball charge, and pebble crusher cut-point in real time.

Liberation Size (d₅₀_lib)

25–125 µm

The particle size at which 50% of target mineral grains are fully liberated from gangue, determined by QEMSCAN or MLA analysis.

⚡ Engineering Impact:

Drives optimal secondary/tertiary crushing P80 and grinding circuit classification cut-point to maximize recovery without overgrinding.

Grade Variability Index (GVI)

0.15–0.45 (unitless)

Standard deviation of assay grade (e.g., Cu % or Au g/t) across a 10 m × 10 m × 5 m mining block, normalized by mean grade.

⚡ Engineering Impact:

Triggers selective blending decisions and determines minimum lot size for stockpile homogenization before milling.

Acid Consumption (ANC)

0.5–8.0 kg/t

Mass of CaCO₃-equivalent alkalinity consumed per tonne of ore during leaching, measured via acid titration of crushed sample.

⚡ Engineering Impact:

Determines lime addition rate and residence time in agitated leach tanks to prevent pH collapse and cyanide loss.

📐 Key Formulas

Grade Variability Index (GVI)

GVI = σ_grade / μ_grade

Quantifies spatial grade heterogeneity within a defined mining unit.

Typical Ranges:
Porphyry Cu systems
0.20 – 0.40
Banded Iron Formation (BIF)
0.08 – 0.18
⚠️ GVI > 0.35 triggers mandatory blending protocol

Ore Hardness Index (OHI)

OHI = (P_actual / P_reference) × (Q_reference / Q_actual) × (F₈₀_reference / F₈₀_actual)^0.2

Empirical normalization of SAG mill specific energy to standard ore conditions.

Typical Ranges:
Soft oxide ores
0.65 – 0.85
Hard sulphide porphyry
1.10 – 1.45
⚠️ OHI < 0.75 or > 1.40 requires circuit reconfiguration review

🏭 Engineering Example

Cadia East Mine, New South Wales, Australia

Porphyry Cu-Au system (quartz monzonite host with potassic alteration)
ANC
2.3 kg/t
GVI
0.28
OHI
1.12
d₅₀_lib
82 µm
ROM Feed Rate
12,500 t/h
SAG Mill Specific Energy
8.7 kWh/t

🏗️ Applications

  • Porphyry copper operations
  • Gold heap leach campaigns with variable oxidation
  • Iron ore pellet feed optimization

📋 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

Geological ModelROM AnalyzerDCS AdjustmentData Flow Loop
Block ModelOre TrackingCircuit Control

📚 References