Quantitative Risk Assessment (QRA) with Consequence Modeling & Frequency Estimation

Industry Applications

Refineries, petrochemical plants, LNG terminals, fertilizer facilities, pharmaceutical manufacturing

Key Standards

CCPS Guidelines (API RP 752/753), UK HSE EG-17, ISO 31000, EN 1050

Typical Scale

Site-wide QRA: 100–500 initiating events; 5–20 major scenarios; 1–3 million Monte Carlo iterations

Regulatory Trigger

COMAH upper-tier sites (>50 t of specified substances); US EPA RMP Tier 3; OSHA PSM-covered processes

⚠️ Why It Matters

1

Inadequate consequence modeling

↓

2

Underestimated hazard zone extent

↓

3

Insufficient separation distances

↓

4

Non-compliant siting of control rooms or shelters

↓

5

Unacceptable risk to personnel or public

↓

6

Regulatory enforcement, operational shutdown, or civil liability

📘 Definition

Quantitative Risk Assessment (QRA) is a systematic, data-driven engineering methodology that integrates consequence modeling (e.g., dispersion, thermal radiation, overpressure) with frequency estimation (e.g., fault tree analysis, event tree analysis, historical failure data) to compute individual and societal risk metrics—typically expressed as annual probability of fatality (APF) or risk contours (e.g., 10⁻⁴ /yr isopleths). It forms the technical basis for ALARP (As Low As Reasonably Practicable) demonstration and regulatory compliance in process safety management frameworks such as OSHA 1910.119 and COMAH.

🎨 Concept Diagram

AI-generated illustration for visual understanding

💡 Engineering Insight

A QRA is only as robust as its weakest input—not its most sophisticated model. Over-reliance on high-fidelity CFD without verifying release rate assumptions or ignition logic often yields 'precise wrong answers.' Always anchor consequence modeling in plant-specific operating data (e.g., actual pressure transients during valve closure) and calibrate frequency estimates against site-specific failure history—not generic database values.

📖 Detailed Explanation

At its core, QRA combines two pillars: consequence modeling (what happens *if* something fails) and frequency estimation (how often it *might* fail). Consequence modeling uses physics-based equations—such as Gaussian dispersion for gases or TNT-equivalent scaling for explosions—to predict hazard footprints (e.g., 1% lethality distance for hydrogen sulfide). Frequency estimation relies on probabilistic methods like fault tree analysis, where basic events (e.g., seal failure, instrument error) are logically combined to derive top-event likelihood.

Modern QRA goes beyond single-scenario analysis by incorporating uncertainty quantification: Monte Carlo sampling across input distributions (e.g., wind speed, release duration, ignition probability) generates risk bands—not point estimates. This enables rigorous ALARP justification, where mitigation options are evaluated not just on cost, but on their impact on the *upper bound* of the 95% confidence interval of risk.

Advanced practice now integrates digital twin concepts—linking real-time sensor data (e.g., pressure, temperature, gas detection) into dynamic QRA models that update risk contours live. Regulatory bodies like the UK HSE and US CSB increasingly expect this level of fidelity for major hazard installations, especially where legacy models conflict with observed incident behavior (e.g., unexpected vapor cloud ignition due to unmodeled static discharge sources).

🔄 Engineering Workflow

Step 1

Step 1: Hazard Identification & Scenario Selection (HAZOP/LOPA-supported)

→

Step 2

Step 2: Consequence Modeling (dispersion, fire, explosion, toxic effects) using validated tools (e.g., PHAST, EFFECTS, FLACS)

→

Step 3

Step 3: Frequency Estimation via Fault Tree Analysis (FTA) and Component Reliability Databases

→

Step 4

Step 4: Risk Integration (individual/societal risk contours, F-N curves, risk matrices)

→

Step 5

Step 5: ALARP Evaluation & Mitigation Option Analysis (cost-benefit, SIL verification, barrier effectiveness)

→

Step 6

Step 6: Documentation & Regulatory Submission (e.g., COMAH Safety Report, EPA RMP Appendix B)

→

Step 7

Step 7: Periodic Review & Model Updating (post-incident, process change, new data)

📋 Decision Guide

Rock/Field Condition	Recommended Design Action
High-consequence site (e.g., adjacent to residential area) + flammable, low-ignition-probability release (e.g., LNG, ethylene)	Perform detailed CFD-based dispersion + VCE modeling; apply 95th-percentile meteorological dataset; install rapid isolation & vapor suppression systems
Toxic release (e.g., chlorine, H₂S) with high EFD (>1×10⁻²/yr) and population within 500 m	Require real-time gas detection with automatic shutdown & blast-resistant shelter-in-place design; validate with ALOHA or SLAB model using worst-case stability class F
Low-frequency, high-consequence scenario (e.g., vessel BLEVE) with no mitigative safeguards	Conduct sensitivity analysis on rupture energy and fragment trajectory; mandate mechanical integrity program with 100% UT inspection and fracture mechanics assessment

📊 Key Properties & Parameters

Release Rate

0.1–500 kg/s (for LNG, chlorine, ammonia, hydrocarbons)

Mass flow rate of hazardous material during a defined failure scenario (e.g., pipe rupture, valve failure), typically modeled for worst-case and realistic credible cases.

⚡ Engineering Impact:

Directly governs plume length, thermal dose, and overpressure decay—driving exclusion zone size and mitigation system capacity.

Ignition Probability

0.01–0.3 (unitless, dimensionless fraction)

Conditional probability that a flammable release will ignite, based on release duration, vapor cloud persistence, and local ignition source density.

⚡ Engineering Impact:

Determines whether consequence modeling proceeds to fireball, jet fire, or vapor cloud explosion (VCE) scenarios—each with vastly different risk profiles.

Equipment Failure Frequency (EFD)

1×10⁻⁵ – 1×10⁻¹ /yr (e.g., 3×10⁻³/yr for gate valves in sour service)

Average number of failures per year for a specific component type under defined operating conditions, derived from databases like OREDA, CCPS, or site-specific reliability data.

⚡ Engineering Impact:

Serves as the root frequency input for fault trees; errors here propagate exponentially through the entire QRA uncertainty envelope.

Dispersion Model Resolution

1–10 m horizontal grid spacing; 1–60 s time steps

Spatial and temporal discretization fidelity used in atmospheric dispersion modeling (e.g., grid cell size, time step), affecting accuracy of concentration predictions downwind.

⚡ Engineering Impact:

Coarse resolution masks localized high-concentration pockets, leading to non-conservative toxic or flammability threshold exceedances.

📐 Key Formulas

Individual Risk (IR)

IR = Σ (Frequency_i × Consequence_i)

Annual probability of fatality for a person located at a specific point (e.g., control room), summed over all relevant scenarios i.

Variables:

Symbol	Name	Unit	Description
IR	Individual Risk	1/year	Annual probability of fatality for a person located at a specific point
Frequency_i	Scenario Frequency	1/year	Annual frequency of scenario i
Consequence_i	Scenario Consequence	probability of fatality per occurrence	Probability of fatality given occurrence of scenario i

Typical Ranges:

Control room location (unmitigated)

1×10⁻⁵ – 5×10⁻³ /yr

Public receptor >1 km from fence line

1×10⁻⁷ – 1×10⁻⁵ /yr

⚠️ ALARP upper limit: ≤1×10⁻⁴ /yr (UK HSE), ≤1×10⁻⁵ /yr (some Dutch & Norwegian jurisdictions)

🔧 Open Calculator

Vapor Cloud Explosion Overpressure (ΔP)

ΔP = 0.26 × (W_TNT / R)^0.79 (Baker-Strehlow equation, simplified)

Peak side-on overpressure (kPa) at distance R (m) from center of vapor cloud, where W_TNT is equivalent TNT mass (kg).

Variables:

Symbol	Name	Unit	Description
ΔP	Peak side-on overpressure	kPa	Overpressure at distance R from center of vapor cloud
W_TNT	Equivalent TNT mass	kg	Mass of TNT equivalent to the vapor cloud's energy content
R	Distance from cloud center	m	Radial distance from the center of the vapor cloud to the point of overpressure measurement

Typical Ranges:

Lethal threshold (100 kPa)

R ≈ 30–120 m for 10–100 ton propane clouds

Window breakage (5 kPa)

R ≈ 200–800 m

⚠️ Design for ≥100 kPa resistance if within 2× lethal distance; ≥20 kPa for occupied buildings beyond

🔧 Open Calculator

🏭 Engineering Example

ExxonMobil Baton Rouge Refinery (Louisiana, USA)

N/A — Process facility (not geological); relevant medium: carbon steel piping & ASME B31.4 liquid hydrocarbon pipeline

Release Rate

85 kg/s (crude oil line rupture at 1200 psig)

Ignition Probability

0.18 (based on site ignition source survey and 3-year incident log)

Mitigation Implemented

Remote shutdown + nitrogen purging + relocation of control room 420 m east (reduced IR to 3.2×10⁻⁵ /yr)

Dispersion Grid Resolution

5 m × 5 m × 2 m vertical, 10 s time step (PHAST v8.5)

Equipment Failure Frequency

2.4×10⁻³/yr (pump seal failure, OREDA v11.2 adjusted for sour service)

Individual Risk (IR) at Control Room

1.7×10⁻⁴ /yr (exceeding ALARP threshold of 1×10⁻⁴ /yr)

🏗️ Applications

Safety Integrity Level (SIL) verification
Land-use planning around process facilities
Emergency response zone definition
Insurance risk rating and premium calculation

🔧 Try It: Interactive Calculator

🔧

Process Safety & Risk Management Calculator

Multi-tab calculator for Process Safety & Risk Man...

Open →

📋 Real Project Case

Ammonia Refrigeration System PHA & LOPA Integration at Midwest Food Plant

Retrofit of legacy ammonia refrigeration system serving 300k sq ft food processing facility

Challenge: Outdated PHA documentation; no SIL verification for emergency shutdown valves

Read full case study →

Hydrogen Sulfide Flare Stack Integrity Assessment at Gulf Coast Refinery Nitric Acid Storage Tank MOC Failure Root Cause Analysis at Fertilizer Facility Ethylene Oxide Sterilization Unit QRA & Emergency Response Optimization Polymer Reactor Runaway Reaction Mitigation via SIS Redundancy Upgrade

🎨 Technical Diagrams

📚 References

[1]

Guidelines for Quantitative Risk Assessment — CCPS (Center for Chemical Process Safety)

[2]

EG-17: The Assessment of Risk from Major Accident Hazards — UK Health and Safety Executive (HSE)

[3]

API RP 752: Management of Hazards Associated with Location of Buildings on Manufacturing Sites — American Petroleum Institute

🔗 Related Knowledge

Emergency Response Planning for Toxic Release Scenarios (e.g., Chlorine, Ammonia)