Calculator D4

Sensor-Based Feed Forward Control in Crushing Circuits

Using real-time sensor data from crushers and conveyors to automatically adjust feed rate, crusher settings, or upstream equipment *before* problems like overload or poor product size happen.

Typical Scale
Deployed on gyratory (primary) and cone (secondary/tertiary) crushers ≥1,200 mm diameter
Industry Adoption
Used at >27 Tier-1 operations globally (2023 ICME survey); highest ROI in ROM blending circuits with >30% grade variance
Standards Alignment
Complies with ISA-106 (Mining Automation) and ISO 22400-2 (KPIs for Smart Manufacturing)

⚠️ Why It Matters

1
Ore grade and hardness variability across mine faces
2
Uncompensated changes in crusher power draw and throughput
3
Excessive fines generation or oversize bypass
4
Increased liner wear and unplanned downtime
5
Reduced mill throughput due to inconsistent crusher product
6
Lower overall plant recovery and higher unit processing cost

📘 Definition

Sensor-based feed forward control (SFFC) is an advanced process automation strategy in mineral processing that uses predictive inputs—such as real-time ore hardness estimates (via XRF, NIR, or gamma transmission), belt load mass flow, particle size distribution (PSD) from laser scanners, and moisture content—to proactively tune crusher operating parameters (e.g., closed-side setting, eccentric speed, feed rate) without waiting for downstream feedback. It integrates multi-sensor fusion, dynamic material property estimation, and model-predictive control (MPC) logic to maintain target throughput, energy efficiency, and product gradation under variable ore feed conditions.

🎨 Concept Diagram

Ore FeedCrusherProductOHI SensorLaser PSDMoistureSensor-Based Feed Forward Control

AI-generated illustration for visual understanding

💡 Engineering Insight

Feed forward control fails not from poor models—but from unmodeled transport delays. A 4.2-second lag between primary crusher discharge and secondary feed conveyor laser measurement will destabilize any MPC unless explicitly embedded as a Smith predictor. Always measure and compensate for physical signal transit time—not just computational latency.

📖 Detailed Explanation

At its core, sensor-based feed forward control replaces reactive operator intervention with anticipatory machine action. Instead of waiting for amperage spikes or vibrating screen carryover to trigger a response, SFFC uses upstream measurements—like ore hardness inferred from gamma transmission through the feed chute—to predict how the crusher will behave *before* the rock enters the crushing chamber. This shifts control from correction to prevention.

The engineering rigor lies in sensor fusion fidelity: gamma sensors estimate bulk density and effective atomic number (Zₑff), which correlate to quartz/feldspar ratio and thus comminution resistance; NIR captures clay-bound water and oxidation state, informing moisture-related plasticity. These signals are fused using Kalman filtering to suppress noise and resolve conflicting indications—e.g., high gamma density + low NIR absorption may indicate dense hematite-rich ore rather than wet claystone.

Advanced implementations embed digital twin physics: a real-time DEM (Discrete Element Method) model, parameterized with live OHI and moisture, simulates particle breakage trajectories inside the crusher cavity. This allows the controller to anticipate not just power demand, but also liner impact zones and fragmentation bias—enabling prescriptive adjustments to eccentric throw angle or mantle profile wear compensation, all before measurable wear occurs or product specification drifts beyond tolerance.

🔄 Engineering Workflow

Step 1
Step 1: Sensor calibration & cross-validation (gamma/NIR vs lab BWI, laser vs sieving)
Step 2
Step 2: Real-time OHI and PSD model deployment on edge PLC with <100 ms latency
Step 3
Step 3: Dynamic crusher power model (based on OHI, CSS, speed, feed rate) loaded into MPC controller
Step 4
Step 4: Feed forward action logic configured: e.g., 'If OHI increases 0.15 over 30 s → initiate CSS ramp at 0.3 mm/min'
Step 5
Step 5: Closed-loop validation test: inject synthetic hardness step-change; verify PSD D80 drift < ±3 mm within 90 s
Step 6
Step 6: Integration with mine dispatch system to align crusher setpoints with upcoming ROM blend forecast
Step 7
Step 7: Weekly performance audit: track % time in auto-feed-forward mode, energy/kT deviation, and liner wear delta vs baseline

📋 Decision Guide

Rock/Field Condition Recommended Design Action
OHI > 1.9 + PSD D80 > 140 mm + Moisture > 9% Reduce CSS by 2–4 mm; increase eccentric speed by 5–8 rpm; activate pre-screening bypass if available
OHI < 0.85 + Mass Flow > 95% design capacity + PSD D80 < 55 mm Increase CSS by 3–6 mm; reduce eccentric speed by 4–6 rpm; engage surge bin holdback to smooth feed
Moisture > 11% + Laser PSD shows >25% particles < 6 mm Activate vibratory deck heaters; divert 15–20% feed to dry screening loop; reduce crusher speed to limit paste formation

📊 Key Properties & Parameters

Ore Hardness Index (OHI)

0.7–2.4 (unitless, 1.0 = median BWI of 12 kWh/t)

Dimensionless index derived from real-time gamma transmission or NIR spectral response, calibrated to Bond Work Index (BWI) equivalents.

⚡ Engineering Impact:

Directly scales crusher power demand prediction and dictates optimal CSS adjustment rate.

Mass Flow Rate

800–3,200 t/h (primary crushing), 400–1,600 t/h (secondary/tertiary)

Real-time volumetric feed rate corrected for density and belt speed, measured via nuclear densitometers + belt scale integration.

⚡ Engineering Impact:

Primary actuator signal for feed conveyor speed and hopper gate opening control; deviation >±5% triggers cascade override.

PSD D80 (Laser Scanner)

45–180 mm (primary), 12–65 mm (secondary)

80th percentile particle size (mm) estimated continuously via 3D laser profilometry on the discharge conveyor.

⚡ Engineering Impact:

Drives closed-side setting (CSS) correction via MPC to maintain target circuit P80 while minimizing recirculating load.

Moisture Content

3.2–12.8 %

Bulk moisture (% w/w) inferred from microwave attenuation or capacitive sensors mounted on feed chute.

⚡ Engineering Impact:

Adjusts crusher choke feed setpoint and screen deck inclination to prevent blinding and maintain screening efficiency.

📐 Key Formulas

Ore Hardness Index (OHI)

OHI = (μ_gamma × k₁) + (R_NIR × k₂) + b

Empirical calibration combining gamma attenuation coefficient (μ_gamma, cm⁻¹) and NIR reflectance ratio (R_NIR, 1000/1500 nm band) to estimate BWI-equivalent hardness.

Typical Ranges:
Dacite/Porphyry
1.2 – 1.8
Banded Iron Formation
1.7 – 2.3
⚠️ OHI > 2.1 triggers automatic feed reduction and liner inspection alert

Predicted Crusher Power (kW)

P_pred = K × OHI × F × (1/CSS)^0.42 × N^0.68

Physics-informed empirical model linking predicted power to OHI, mass flow (F, t/h), CSS (mm), and eccentric speed (N, rpm).

Typical Ranges:
Gyratory Primary (13 m dia)
2,800 – 4,100 kW
HP500 Cone (Secondary)
520 – 860 kW
⚠️ Alarm if |P_pred − P_actual| > 4.5% for >60 s

🏭 Engineering Example

Cadia East, New South Wales, Australia (Newcrest Mining, now Newmont)

Porphyritic Dacite / Quartz Monzonite
OHI
1.62
PSD D80
68 mm
CSS Setpoint
52.3 mm
Mass Flow Rate
2,480 t/h
Moisture Content
5.3 %
Crusher Power Draw
3,140 kW

🏗️ Applications

  • ROM blending optimization in porphyry copper circuits
  • Hard-rock gold operations with erratic quartz veining
  • Iron ore export facilities requiring strict size compliance

📋 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

OHI SensorLaser PSDMoisture SensorSensor Fusion Node
OHI → CSS RampPSD → Speed AdjustMPC ControllerFeed Forward Action Logic

📚 References

[1]
Guidelines for Sensor-Based Ore Characterisation and Control — Australian Centre for Geomechanics (ACG)
[2]
ISA-106: Mining and Mineral Processing Automation — International Society of Automation
[3]
Crushing and Grinding Circuits: A Practical Handbook — Society for Mining, Metallurgy & Exploration (SME)