🎓 Lesson 21
D5
Diagnosing and Correcting Scale-Up Failures
Scale-up failure happens when a blasting design that works well in small-scale tests doesn’t perform as expected in full-size mine operations.
🎯 Learning Objectives
- ✓ Calculate optimal burden and spacing using Kuz-Ram and Langefors models
- ✓ Analyze blast design parameters for geometric and energetic similarity
- ✓ Explain how rock mass rating (RMR) and joint spacing influence scale-up fidelity
- ✓ Apply dimensional analysis to diagnose mismatch between lab-scale and field-scale energy density
- ✓ Design a scaled validation test plan incorporating bench height, stemming length, and delay timing
📖 Why This Matters
In mining, a poorly scaled blast can cost millions: excessive oversize forces secondary crushing, poor fragmentation reduces truck payload efficiency, and overbreak damages adjacent walls—compromising slope stability. Over 60% of unplanned downtime in open-pit operations traces back to scale-up failures—not equipment breakdowns. Mastering scale-up diagnostics isn’t theoretical; it’s the difference between profitable production and costly rework.
📘 Core Principles
Scale-up in blasting rests on three pillars: (1) Geometric similarity—maintaining proportional relationships among burden (B), spacing (S), and bench height (H); (2) Kinematic similarity—preserving consistent particle velocity ratios via scaled charge weight and delay timing; and (3) Dynamic similarity—ensuring equivalent specific energy (J/m³) and stress wave attenuation across scales. Crucially, rock mass heterogeneity (e.g., joint frequency, RMR < 40) breaks similarity assumptions—making empirical correction factors essential. The Kuznetsov–Rammler (Kuz-Ram) model links fragmentation to energy input and rock properties, while Langefors’ formula adds confinement and stiffness effects. Failure occurs when any pillar is violated—especially under variable weathering or seismic velocity gradients.
📐 Kuz-Ram Fragmentation Prediction
The Kuz-Ram model estimates the fragment size distribution (x₅₀) based on explosive energy, burden, and rock properties. It is widely used for preliminary scale-up validation and benchmarking against field fragmentation surveys.
Kuz-Ram x₅₀
x₅₀ = K × (Bⁿ / PF)Predicts the 50th percentile fragment size (mm) based on burden (m), powder factor (kg/m³), rock factor K (empirical), and exponent n (rock competency).
Variables:
| Symbol | Name | Unit | Description |
|---|---|---|---|
| x₅₀ | Median fragment size | mm | Size at which 50% of fragments by mass are smaller |
| K | Rock factor | dimensionless | Empirically calibrated value (6–25) reflecting rock strength and structure |
| B | Burden | m | Distance from free face to first row of holes |
| n | Exponent | dimensionless | Typically 0.6–0.9; lower values indicate more competent rock |
| PF | Powder factor | kg/m³ | Mass of explosive per unit volume of rock broken |
Typical Ranges:
Medium-hard granite (RMR 60): K = 10–14, n = 0.75–0.85
Soft sedimentary rock (RMR 35): K = 6–9, n = 0.6–0.7
💡 Worked Example
Problem: Given: ANFO density = 0.85 g/cm³, VOD = 4,000 m/s, burden B = 4.2 m, rock density ρ = 2.65 g/cm³, rock factor K = 12 (medium-hard granite), exponent n = 0.8, powder factor PF = 0.35 kg/m³.
1.
Step 1: Compute specific energy E = (VOD² × ρ × 10⁻⁶) / 2 = (4000² × 2650 × 10⁻⁶)/2 ≈ 21.2 MJ/m³
2.
Step 2: Calculate x₅₀ = K × (Bⁿ / PF) = 12 × (4.2⁰·⁸ / 0.35) ≈ 12 × (3.32 / 0.35) ≈ 12 × 9.49 ≈ 113.9 mm
3.
Step 3: Compare with target x₅₀ = 80–100 mm for primary crusher feed; result (114 mm) exceeds upper limit → indicates need to reduce burden or increase PF.
Answer:
The predicted x₅₀ is 114 mm, which exceeds the target range of 80–100 mm. To correct, reduce burden to ~3.7 m or increase powder factor to 0.39 kg/m³.
🏗️ Real-World Application
At the Escondida copper mine (Chile), a 2021 scale-up trial of a new emulsion explosive failed—fragmentation improved in 2-m-diameter test holes but produced 35% oversize in 16-m benches. Root cause analysis revealed unaccounted joint persistence (>2.5 m spacing) and reduced confinement at scale, lowering effective stress wave reflection. Engineers corrected by increasing stemming length by 25%, reducing burden from 4.8 m to 4.1 m, and introducing millisecond delays tuned to P-wave velocity (3,200 m/s). Post-correction x₅₀ dropped from 142 mm to 91 mm—within specification—and dig rate increased 18%.
📋 Case Connection
📋 Bioethanol Fermentation Bioreactor Scale-Up with Inhibition Kinetics
Ethanol inhibition caused premature cessation at large scale despite matching nominal conditions
📋 Nitric Acid Absorption Tower Design for Tail-Gas Treatment
Incomplete absorption of NO and NO₂ due to slow liquid-phase oxidation kinetics and poor gas distribution