📋 Case Study

South African Coal Mine: Digital Twin for Methane Drainage & Ventilation Safety

Intermittent CH₄ spikes triggering false alarms and production halts; inability to distinguish between drainage inefficiency and geological outgassing

🏗️ Project Overview

Mafube Colliery Longwall Panel 7

🎯 Challenge

Intermittent CH₄ spikes triggering false alarms and production halts; inability to distinguish between drainage inefficiency and geological outgassing

🔧 Design Approach

Physics-based methane flow twin integrating borehole pressure transients, seam gas content logs, and ventilation network dynamics; trained on 24-month historical gas event database

📐 Design Diagram

Methane Flow TwinBorehole
PressureSeam Gas
Content LogsVentilation
Network DataForecast OutputMAE = 0.82 m³/minDrainage Efficiency Index0.63 (Actual/Theoretical)Challenge:Intermittent CH₄ spikes →false alarms & halts24-Month
Hist. DBIntegrated Real-Time Data Ingestion

AI-generated project design illustration

📐 Key Calculations

Gas Emission Forecast MAE

Mean Absolute Error

Result: 0.82 m³/min

Below 1.0 m³/min enables confident alarm suppression

Drainage Efficiency Index

Actual Drainage / Theoretical Drainage

Result: 0.63

Triggers borehole rehabilitation workflow when <0.7

📊 Results

False alarm rate reduced from 17 to 2 per month; drainage optimization increased capture rate by 31%; no unplanned ventilation shutdowns in 14 months

💡 Lessons Learned

• •Seam permeability must be updated quarterly via pulse testing
• •Twin requires explicit representation of gob gas migration pathways

✅ Key Takeaways

• 1Seam permeability must be updated quarterly via pulse testing
• 2Twin requires explicit representation of gob gas migration pathways