🎓 Lesson 19 D5

Ore Blending Optimization Using Haul Data

Ore blending optimization using haul data means mixing different types of mined material during transport to ensure the processing plant receives a consistent, target-grade feed.

🎯 Learning Objectives

  • Calculate blend ratios required to achieve a target grade given assay values and tonnages of two or more ore sources
  • Design a haul-based blending strategy that respects truck payload limits and cycle time constraints
  • Analyze GPS and payload sensor data to identify deviations from planned blend targets
  • Explain how dilution and ore loss impact blend quality and downstream metallurgical recovery
  • Apply linear programming concepts to formulate a simple constrained blend optimization problem

📖 Why This Matters

In modern open-pit mines, inconsistent feed grade causes costly fluctuations in concentrator throughput, reagent consumption, and final concentrate quality. A 1% deviation in copper grade can reduce annual revenue by millions—yet many mines still rely on manual, reactive blending. Haul data (GPS location, payload weight, assigned shovel, real-time assay tags) provides the highest-resolution, lowest-latency input for proactive blending—turning haulage from a cost center into a precision control system.

📘 Core Principles

Blending relies on three foundational layers: (1) Geometallurgical modeling—linking spatial orebody domains to grade, hardness, and metallurgical response; (2) Operational fidelity—ensuring haul data (truck ID, start/end time, payload mass, GPS coordinates, assigned blast/face) is time-synchronized and validated against shovel dispatch and lab assay databases; and (3) Control logic—using weighted averaging, moving-window statistics, or model-predictive control to compute real-time blend deviations and trigger corrective actions (e.g., rerouting trucks, adjusting shovel assignment). Advanced implementations integrate digital twins that simulate blending outcomes under varying fleet availability and grade uncertainty.

📐 Weighted Average Blend Grade

The fundamental calculation for verifying whether a blended stream meets target grade. Used continuously in haul dispatch systems and stockpile management dashboards.

Weighted Average Grade

G_blend = Σ (m_i × G_i) / Σ m_i

Calculates the grade of a mixture composed of multiple ore streams, weighted by their respective masses.

Variables:
SymbolNameUnitDescription
G_blend Blended grade % Target or resulting grade of the combined material stream
m_i Mass of ore stream i tonnes (t) Net payload mass of material hauled from source i
G_i Grade of ore stream i % Assay-determined grade (e.g., Cu%, Fe%, Au g/t) of source i
Typical Ranges:
Copper porphyry: 0.3–1.2% Cu
Iron ore (hematite): 58–65% Fe
Gold oxide heap leach: 0.8–2.5 g/t Au

💡 Worked Example

Problem: Truck T-104 hauls 125 t from Zone A (assay: 0.82% Cu); Truck T-105 hauls 118 t from Zone B (assay: 1.45% Cu); Truck T-106 hauls 132 t from Zone C (assay: 0.61% Cu). Target blend grade = 0.90% Cu. Does this 3-truck batch meet target?
1. Step 1: Compute total blended tonnage = 125 + 118 + 132 = 375 t
2. Step 2: Compute weighted sum = (125 × 0.82) + (118 × 1.45) + (132 × 0.61) = 102.5 + 171.1 + 80.52 = 354.12 %·t
3. Step 3: Compute weighted average grade = 354.12 / 375 = 0.9443% Cu
4. Step 4: Compare to target: 0.9443% > 0.90% → batch is 4.9% above target; may require dilution with low-grade material in next cycle
Answer: The result is 0.944% Cu, which exceeds the target of 0.90% Cu by 0.044 percentage points (4.9% relative error). This deviation falls outside typical allowable tolerance of ±0.03% Cu for primary crusher feed.

🏗️ Real-World Application

At Rio Tinto’s Yandi iron ore operation (Pilbara, WA), haul trucks equipped with RTK-GPS and integrated load cells transmit payload and origin-face ID every 15 seconds to the Mine Production System (MPS). Assay data from blast-hole samples is geo-referenced and interpolated to faces using conditional simulation. MPS calculates real-time blend grade per stockpile bin using 2-hour rolling windows and automatically redirects trucks exceeding grade tolerance (>±0.1% Fe) to alternate stockpiles—reducing crusher feed grade standard deviation by 37% and improving downstream sinter consistency.

📋 Case Connection

📋 Canadian Gold Mine: Steep Ramp Optimization in Narrow Vein Underground

Excessive truck cycle times and premature tire/brake wear due to suboptimal ramp gradient (15%) combined with tight hori...

📋 South African Platinum Mine: Waste Dump Reclaim Optimization

Inefficient haulage routing and underutilized fleet capacity during waste dump reclamation, resulting in excessive diese...

📚 References