🎓 Lesson 11
D5
Case Study: Pillar Recovery Sequence Failure Review
Pillar recovery sequence failure happens when removing support pillars in underground mines goes wrong, causing sudden collapses that endanger people and stop production.
🎯 Learning Objectives
- ✓ Analyze pillar stability using the pillar strength-to-stress ratio (PSR) criterion
- ✓ Design a safe pillar recovery sequence by applying the 'backfill-assisted staged retreat' methodology
- ✓ Explain how stress shadowing and abutment pressure influence pillar failure timing
- ✓ Calculate the critical pillar width using the modified Salamon–Gibson empirical formula
- ✓ Apply mine-wide seismic monitoring data to validate sequence performance
📖 Why This Matters
In 2018, a pillar recovery sequence failure at the Mponeng Gold Mine (South Africa) triggered a 3.2-magnitude rockburst, halting production for 47 days and injuring 12 workers. Such events underscore that pillar recovery isn’t just excavation—it’s the most hazardous phase of underground mining. Getting the sequence wrong doesn’t just cost money; it risks lives, violates regulatory requirements (e.g., MSHA Part 46, SANS 10209), and can trigger cascading instability across entire mine panels.
📘 Core Principles
Pillar recovery relies on three interdependent geomechanical principles: (1) progressive stress redistribution—removing pillars shifts load onto adjacent pillars and hangingwall, increasing their stress state; (2) time-dependent rock mass response—creep and brittle fracture evolve over hours to weeks post-extraction; and (3) sequence dependency—the order of pillar removal determines whether stress is absorbed gradually (safe) or concentrated abruptly (failure). Modern practice treats pillars not as static supports but as engineered 'stress buffers' whose geometry, orientation, and removal timing must be validated via numerical modeling (e.g., UDEC, Phase2) and microseismic monitoring. Failure typically initiates at pillar corners due to tensile fracturing, then propagates along bedding or joint planes—making geological structure mapping non-negotiable.
📐 Critical Pillar Width Calculation
The modified Salamon–Gibson formula estimates minimum stable pillar width based on depth, rock mass rating, and panel geometry. It accounts for both compressive strength and structural weakening, making it preferred over simple area-to-depth ratios in modern design.
Modified Salamon–Gibson Pillar Width
W_min = k × f_RMR × √(W_p × L_p)Estimates minimum stable width of a rectangular pillar under given depth and rock mass quality.
Variables:
| Symbol | Name | Unit | Description |
|---|---|---|---|
| W_min | Minimum stable pillar width | m | Smallest width that maintains stability under expected stress conditions |
| k | Depth factor | dimensionless | Empirical coefficient scaling with mining depth (k = 0.001 × depth_in_m) |
| f_RMR | Rock Mass Rating correction factor | dimensionless | Adjusts for rock mass quality; f_RMR = 0.01 × (RMR − 50) + 1.0 |
| W_p | Panel width | m | Distance between main haulage entries or abutments |
| L_p | Pillar length | m | Dimension parallel to mining direction |
Typical Ranges:
Deep hard-rock gold mining (>1000 m): 40 – 65 m
Shallow massive sulfide deposits (<500 m): 18 – 32 m
💡 Worked Example
Problem: Given: mining depth = 1,200 m, RMR = 58, panel width = 45 m, pillar length = 30 m. Calculate minimum stable pillar width (W_min).
1.
Step 1: Compute depth factor k = 0.001 × depth = 0.001 × 1200 = 1.2
2.
Step 2: Compute RMR correction factor f_RMR = 0.01 × (RMR − 50) + 1.0 = 0.01 × (58 − 50) + 1.0 = 1.08
3.
Step 3: Apply formula W_min = k × f_RMR × √(panel_width × pillar_length) = 1.2 × 1.08 × √(45 × 30) = 1.296 × √1350 ≈ 1.296 × 36.74 = 47.6 m
4.
Step 4: Compare with actual pillar width (assumed 25 m): 25 m < 47.6 m → design is unstable; requires sequencing or backfill
Answer:
The calculated minimum stable pillar width is 47.6 m; the existing 25 m pillar is critically undersized and must be recovered only after partial backfilling or in smaller sub-sequences.
🏗️ Real-World Application
At Vale’s Onça Puma Nickel Mine (Brazil), pillar recovery in the C Zone was halted after microseismic event clustering exceeded 50 events/day (>2.0 magnitude) following removal of two adjacent pillars. Engineers resequenced using a 'checkerboard + central pillar retention' pattern, installed hydraulic backfill in previously extracted zones, and extended pillar dwell time to 14 days before next removal. Post-implementation, seismicity dropped to <5 events/day, and full recovery proceeded without incident—validating the importance of adaptive, data-driven sequencing over rigid pre-planned schedules.
📋 Case Connection
📋 Underground Limestone Mine Tunneling with Hybrid TBM
Highly variable ground conditions—including intact limestone (UCS 80–120 MPa), fault zones with clay-filled shear zones...