Calculator D4

Ore Tracking and Traceability System Design

A system that follows each batch of mined ore from the pit to the processing plant, like a barcode scanner for rocks, so engineers know exactly what grade and type of material is being fed into crushers and mills.

Typical Scale
15–40 Mt/y throughput; 200–500 traceable batches/hour
Industry Standards
CRIRSCO Code, ISO/IEC 17025 (lab accreditation), ISA-95 Level 3 MES integration
Data Volume
12–35 TB/month per major operation (GNSS, assay, sensor streams)

⚠️ Why It Matters

1
Inconsistent ore feed composition
2
Unstable mill throughput and recovery
3
Excessive reagent consumption
4
Premature wear of grinding media and liners
5
Reduced metal recovery and increased unit operating cost
6
Failure to meet concentrate specification or smelter penalties

πŸ“˜ Definition

An Ore Tracking and Traceability System (OTT) is an integrated engineering framework that links geological, geotechnical, and operational data across the mine-to-mill value chain. It enables deterministic assignment of spatially resolved ore attributes (e.g., grade, hardness, mineralogy) to discrete material batches via real-time positioning, sampling, and digital twin synchronization. Its core function is to close feedback loops between upstream grade control decisions and downstream metallurgical performance.

🎨 Concept Diagram

Ore Tracking & Traceability SystemGeological Model & Block GradesReal-Time Sensor Data (GNSS, Payload, Spectro)Process Control System (DCS/MES)Digital Twin Engine (LP Solver + Physics Model)Fig. 0: Core system architecture (bidirectional data flow)

AI-generated illustration for visual understanding

πŸ’‘ Engineering Insight

Traceability isn’t about tracking every tonne β€” it’s about tracking *enough* tonnes, *accurately enough*, *fast enough* to move the needle on recovery. A 0.3% improvement in copper recovery requires <15-minute latency and Β±0.4 m positional fidelity β€” not sub-centimeter GPS. Over-engineering traceability wastes CAPEX; under-engineering it guarantees metallurgical drift.

πŸ“– Detailed Explanation

At its foundation, ore traceability relies on deterministic mapping: assigning a known geological domain (e.g., a 5Γ—5Γ—2.5 m block) to a physical material unit (e.g., one shovel bucket). This requires synchronized timestamps, calibrated GNSS antennas on equipment, and unambiguous material identity β€” typically enforced via RFID tags on haul trucks or optical barcodes on conveyor belts.

As systems mature, traceability shifts from static assignment to dynamic inference. When direct measurement gaps exist (e.g., no assay for a given shovel pass), Bayesian updating combines prior block model estimates with real-time proxies β€” such as gamma-ray spectrometry (for K/U/Th), dielectric permittivity (for clay content), or acoustic emission during crushing (for brittleness). These proxies are calibrated against reference assays using partial least squares regression.

The most advanced implementations embed traceability into digital twin architecture: a live, physics-informed simulation of the entire ore flow path. Here, traceability data feeds not just control logic but also predictive maintenance models (e.g., liner wear rate vs. OCF), environmental compliance reporting (e.g., arsenic co-location), and regulatory audit trails compliant with ISO 22000 and CRIRSCO reporting standards.

πŸ”„ Engineering Workflow

Step 1
Step 1: Define traceability boundary (pit face β†’ mill feed bin) and assign unique batch IDs per shovel cycle
β†’
Step 2
Step 2: Integrate GNSS, payload, and sensor telemetry from haul trucks and conveyors into centralized data lake
β†’
Step 3
Step 3: Apply spatial-temporal alignment to link assay composites (blasthole, face, shovel bucket) with physical material movement
β†’
Step 4
Step 4: Compute dynamic ore blend ratios using linear programming (LP) solver constrained by grade, hardness, and throughput targets
β†’
Step 5
Step 5: Dispatch real-time setpoints to crushers, SAG mills, and flotation banks via OPC UA interface
β†’
Step 6
Step 6: Validate traceability fidelity via mass balance reconciliation (Β±1.2% tolerance) and isotopic tracer audits
β†’
Step 7
Step 7: Update geostatistical block model weekly using reconciliation residuals and process response data

πŸ“‹ Decision Guide

Rock/Field Condition Recommended Design Action
High GVI (>22%) + Low OCF (<0.9) + Latency > 45 min Deploy in-pit XRF analyzers on shovels + dual-belt weighfeeders with synchronized RFID tagging
Moderate GVI (12–18%) + OCF 1.3–1.7 + Latency < 20 min Implement automated composite sampling on primary crusher discharge conveyor with 2-min cycle time
Low GVI (<10%) + High OCF (>1.8) + Latency < 10 min Use historical block model interpolation with real-time gamma-ray density correction on secondary crusher feed

📊 Key Properties & Parameters

Grade Variability Index (GVI)

8–25% for copper porphyry, 15–40% for gold oxide heaps

Dimensionless metric quantifying the coefficient of variation of assay grade within a defined block or blast round (Οƒ/ΞΌ Γ— 100%)

⚡ Engineering Impact:

Drives frequency and density of blasthole sampling and dictates minimum batch size for reliable process control

Ore Competence Factor (OCF)

0.6–2.2 (unitless)

Empirical ratio of rock mass rating (RMR) to specific energy index (SEI), used to predict comminution response

⚡ Engineering Impact:

Determines optimal crusher setting and SAG mill ball charge configuration for target P80

Batch Traceability Latency

4–120 minutes (depending on haul distance and lab throughput)

Time elapsed between ore extraction at shovel and validated geochemical/metallographic assignment in the process control system

⚡ Engineering Impact:

Limits responsiveness of real-time circuit adjustments; >30 min latency prevents closed-loop control of flotation reagent dosing

Spatial Assignment Accuracy

0.15–0.8 m horizontal, 0.2–1.2 m vertical

Root-mean-square error (RMSE) of GPS-RTK or total station positioning applied to individual haul truck loads or conveyor belt segments

⚡ Engineering Impact:

Directly limits resolution of ore blending models; >0.5 m RMSE degrades reconciliation accuracy below 92%

πŸ“ Key Formulas

Grade Reconciliation Error (GRE)

GRE = |(Ξ£Q_i Γ— g_i)_measured βˆ’ (Ξ£Q_i Γ— g_i)_model| / Ξ£Q_i Γ— 100%

Quantifies discrepancy between measured plant feed grade and predicted block model grade

Typical Ranges:
Open-pit copper
0.8–2.4% absolute
Underground gold
1.5–4.1% absolute
⚠️ ≀1.5% absolute for Grade Control Certification (JORC/NI 43-101)

Minimum Effective Batch Size (MEBS)

MEBS = (3 Γ— Οƒ_g)^2 / (Ξ”g_target)^2 Γ— Q_hour

Smallest material volume required to detect a target grade change Ξ”g_target with 99.7% confidence

Typical Ranges:
SAG mill feed control
85–220 t
Flotation bank dosing
12–45 t
⚠️ Must be ≀5% of hourly feed rate to enable responsive control

🏭 Engineering Example

Escondida Mine, Chile

Copper-Molybdenum Porphyry (Diorite/Granodiorite)
GVI
19.3%
OCF
1.42
Reconciliation Accuracy
98.4%
Batch Traceability Latency
18.7 minutes
Spatial Assignment Accuracy (RMSE)
0.29 m
Mill Recovery Improvement (vs. pre-OTT)
+1.7% Cu

πŸ—οΈ Applications

  • Mine-to-mill optimization
  • Grade control reconciliation
  • Metallurgical circuit tuning
  • Regulatory reporting (CRIRSCO, JORC)
  • Predictive maintenance scheduling

πŸ“‹ 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 β†’

🎨 Technical Diagrams

ShovelHaul TruckCrusherMill Feed BinFig. 1: Material flow with traceability handoff points
GNSSAssayPayloadLab IDFig. 2: Multi-sensor fusion timeline (ms resolution)

πŸ“š References

[2]
ISA-95 Enterprise-Control System Integration β€” International Society of Automation
[3]
Geostatistics for Mine Planning and Grade Control β€” Society for Mining, Metallurgy & Exploration (SME)