🎓 Lesson 17
D5
QRA Fundamentals: Dispersion Modeling & Frequency Estimation
Quantitative Risk Assessment (QRA) for dispersion modeling estimates how far and how fast hazardous materials (like toxic gases from blasting or process failures) spread in the air, and how often such events might happen.
🎯 Learning Objectives
- ✓ Calculate downwind concentration profiles using Gaussian plume dispersion models
- ✓ Estimate release frequency for common blasting-related hazardous scenarios (e.g., misfire-induced NO₂ release, ANFO decomposition)
- ✓ Analyze consequence contours (e.g., ERPG-2, IDLH) against site layout to identify emergency evacuation zones
- ✓ Explain the impact of meteorological stability class and source height on dispersion outcomes
- ✓ Apply industry-standard failure rate data (e.g., from OREDA or CCPS) to construct basic fault trees for blast-related gas release events
📖 Why This Matters
In mining and explosives handling, unintended releases—such as nitrogen dioxide (NO₂) from misfired ANFO, hydrogen sulfide from ore processing, or dust clouds from overbreak—can endanger personnel, nearby communities, and infrastructure. QRA-based dispersion modeling isn’t theoretical: it directly informs blast exclusion zones, shelter-in-place protocols, real-time monitoring placement, and regulatory compliance (e.g., EPA RMP, OSHA Process Safety Management). Getting this wrong can mean delayed evacuations—or unjustified operational shutdowns.
📘 Core Principles
Dispersion modeling rests on two pillars: (1) atmospheric physics—governed by turbulence, wind speed/direction, and stability class (Pasquill-Gifford categories A–F)—and (2) source characterization—release rate, duration, height, temperature, and buoyancy. Frequency estimation relies on probabilistic methods: initiating event identification (e.g., detonator failure, water infiltration into emulsion), failure rate databases, and logical modeling (fault/event trees). Together, they convert engineering design choices (e.g., blast pattern geometry, venting strategy) into spatially resolved risk maps. Crucially, QRA distinguishes between instantaneous (puff) and continuous (plume) releases—and mining scenarios often involve short-duration puffs with complex terrain effects.
📐 Gaussian Plume Model (Steady-State)
The Gaussian plume model estimates ground-level concentration (C) at distance x downwind and y meters laterally from a continuous elevated point source under steady meteorological conditions. It assumes constant wind speed, flat terrain, and no chemical transformation—making it suitable for first-pass emergency planning near open-pit blasts or surface facilities.
💡 Worked Example
Problem: A misfire in a 20-ton ANFO blast releases NO₂ at an effective stack height of 15 m (including buoyant rise) for 30 seconds. Average wind speed = 3.5 m/s; Pasquill stability class = D (neutral); release rate = 4.2 kg/s. Calculate C at x = 500 m, y = 0 m (centerline).
1.
Step 1: Determine σ_y and σ_z at x = 500 m using Pasquill-Gifford curves: for Class D, σ_y ≈ 36 m, σ_z ≈ 18 m.
2.
Step 2: Apply Gaussian plume formula: C(x,y,0) = (Q / (2π u σ_y σ_z)) × exp[−½(y/σ_y)²] × exp[−½(H/σ_z)²], where H = 15 m.
3.
Step 3: Plug values: Q = 4.2 kg/s, u = 3.5 m/s → C = (4.2 / (2π × 3.5 × 36 × 18)) × exp[0] × exp[−½(15/18)²] ≈ 0.000148 × 0.606 ≈ 8.97×10⁻⁵ kg/m³ = 89.7 mg/m³.
Answer:
The centerline concentration is 89.7 mg/m³ at 500 m — exceeding ERPG-2 (10 mg/m³ for NO₂) by nearly 9×, indicating immediate respiratory hazard and mandatory evacuation within that radius.
🏗️ Real-World Application
At the Bingham Canyon Mine (Utah), a 2019 QRA study modeled NO₂ dispersion following a large-scale misfire in a copper porphyry bench. Using AERMOD with local 5-year meteorological data and calibrated terrain elevation, engineers identified a 1.2-km downwind zone where concentrations exceeded ERPG-2 during stable nighttime inversions. This led to revised blast timing protocols (banning night blasts during Class F stability) and installation of automated NO₂ sensors at 0.8 km and 1.5 km downwind—reducing false alarms by 73% and cutting average emergency response time from 4.2 to 1.8 minutes.
🔧 Interactive Calculator
🔧 Open Process Safety & Risk Management Calculator📋 Case Connection
📋 Ethylene Oxide Sterilization Unit QRA & Emergency Response Optimization
Inadequate off-site consequence modeling; evacuation radius underestimated by 400%