📋 Case Study

Mine Planning & Scheduling Case Study 2

Commercial application

🏗️ Project Overview

Open-pit copper mine in northern Chile; 120 Mt/y throughput capacity; 25-year life-of-mine; operated by a multinational mining company with integrated processing and rail logistics.

🎯 Challenge

Inconsistent haul truck utilization due to static, annual mine plans that failed to account for real-time geotechnical variability, equipment downtime, and fluctuating ore grade targets—resulting in 18% schedule slippage and $24M/year in underutilized fleet costs.

🔧 Design Approach

Adopted a dynamic, stochastic short-term scheduling framework integrating geostatistical ore body modeling, discrete-event simulation of haul cycle times, and mixed-integer linear programming (MILP) optimization updated weekly using live GPS fleet telemetry and blast fragmentation data.

📐 Design Diagram

AI-generated project design illustration

📐 Key Calculations

Optimal Haul Cycle Time Adjustment Factor

T_cycle_adj = T_base × (1 + 0.003 × ΔSlope% + 0.012 × Moisture_content_% - 0.008 × RQD)

Result: 1.14 (unitless)

Corrects nominal cycle time for site-specific haul road conditions and rock mass quality, improving truck dispatch accuracy by 22%.

Stochastic Ore Grade Variance Penalty Weight

λ_grade = σ²_grade / (Target_grade² × Annual_production_t)

Result: 0.0072 t⁻¹

Quantifies risk cost of grade deviation per tonne, enabling MILP to prioritize selective mining blocks that balance recovery and schedule adherence.

Fleet Utilization Efficiency Index

UEI = (Actual_ore_hauled_t / (Fleet_capacity_t × Operating_hours)) × 100%

Result: 86.3%

Benchmarked against industry best practice (≥85%), confirming feasibility of proposed dispatch logic before implementation.

📊 Results

Metrics: Schedule adherence improved from 71% to 94%, Truck utilization increased from 62% to 86%, Ore grade variance reduced by 31% relative to target, NPV uplift: $112M over LOM

Implementation of the dynamic planning-scheduling loop reduced schedule slippage to <3%, increased net smelter return per tonne by 4.7%, and enabled real-time re-optimization during unplanned pit wall instability events without production loss.

💡 Lessons Learned

•Integration of real-time geotechnical data into scheduling requires standardized API interfaces between survey, geotech, and fleet management systems.
•Stochastic optimization gains are negated without cross-functional operational discipline—scheduler training and KPI alignment across mining, maintenance, and geology teams were critical success factors.

✅ Key Takeaways

1Dynamic, data-driven mine scheduling is not just a software upgrade—it demands concurrent process redesign, governance protocols, and performance-linked accountability across technical functions.