Quantitative Risk Assessment combines event frequencies and consequence modelling to estimate risk from major accident scenarios. It is often used for facility siting, land-use planning, occupied building decisions and comparison of major design options. QRA can provide a powerful system-level view, but the numbers are only as reliable as the scenario set, frequency data and consequence assumptions behind them.
Scenario identification
A QRA begins with a structured set of loss-of-containment scenarios. Sources may include equipment inventories, isolatable sections, leak-size categories and hazard studies. The scenario set should represent the facility without creating unnecessary duplication. Important low-frequency high-consequence events should not disappear simply because a generic leak-frequency model focuses on common equipment failures.
Frequency analysis
Release frequencies may come from industry databases, equipment counts or fault-tree analysis. Isolation, detection, ignition and escalation probabilities can modify the pathway to consequences. Where site-specific data are used, the basis should be statistically meaningful. Small datasets can give a false sense of precision.
Consequence modelling
Fire, explosion, dispersion and toxic models estimate physical effects. The appropriate model depends on fluid behaviour, congestion, weather and release conditions. Consequence endpoints are then combined with occupancy or population data to estimate risk. Consistency between source terms and consequence models is essential.
Individual risk
Individual risk expresses the annual probability of fatality for a hypothetical person at a location, assuming the defined exposure basis. Risk contours can support facility layout and land-use decisions. Interpretation should consider the occupancy assumptions behind the result. A contour is not a prediction that an incident will occur at that point.
Societal risk
Societal risk considers events that can affect multiple people and is often represented using FN curves. It is influenced by population distribution as well as event frequency and consequence. Two facilities with similar individual risk can have different societal risk if one is close to a densely occupied area.
Uncertainty and sensitivity
QRA results can change with leak frequencies, ignition probabilities, weather, occupancy and model choice. Sensitivity analysis helps identify which assumptions drive the result. This is particularly important when a design decision sits close to a risk criterion. The objective should be informed engineering judgement rather than treating one calculated number as absolute truth.
Occupancy and exposure
Risk to personnel depends on where people are and how long they remain exposed. Control rooms, workshops, accommodation and transient work areas may have very different occupancy. Temporary project activities can also change exposure. The QRA basis should state whether occupancy is current, forecast or conservative so results are interpreted correctly.
Using QRA for decisions
QRA is most useful when it informs a decision: comparing layouts, locating occupied buildings, evaluating risk-reduction options or demonstrating compliance with a defined criterion. It is less useful when produced as a stand-alone number without connection to engineering choices. Sensitivity studies can help show where design changes provide the greatest reduction in overall risk.
How QRA fits with other studies
HAZID, HAZOP and bowties provide detailed hazard and barrier understanding. QRA aggregates many scenarios to show overall risk. The methods complement each other: qualitative studies explain how events happen, while QRA helps quantify where cumulative risk is concentrated and whether additional design measures are justified.
How Pro Engineers can assist
Pro Engineers supports process safety risk studies from hazard identification through consequence assessment and risk review. For related semi-quantitative scenario assessment, see our LOPA services.
Event trees and escalation
QRA often uses event trees to represent what happens after a release. Detection, isolation, immediate ignition, delayed ignition and no ignition can lead to different outcomes. Fire, explosion and dispersion consequences may therefore arise from the same initiating leak. The probabilities assigned to these branches can materially influence risk, so the basis should be consistent and clearly documented.
Equipment counts and leak frequencies
Generic leak-frequency data are commonly applied to counts of valves, flanges, vessels, pumps, compressors and piping. The quality of the equipment count matters because errors propagate directly into scenario frequency. The analyst should also understand whether the frequency database represents comparable service, pressure and equipment types. Site-specific performance can be useful where enough credible data exist.
Occupied building decisions
QRA can support decisions about control rooms, workshops and other occupied buildings. The analysis should consider both external event exposure and building vulnerability. Occupancy assumptions need to reflect how the building is actually used, including shift patterns and transient personnel. Where risk is sensitive to occupancy, operational measures may complement physical relocation or strengthening.
Risk reduction options
When risk is above or close to a criterion, the model can be used to test alternatives. Smaller isolatable inventories, faster isolation, improved detection, relocation of occupied areas or changes to layout may reduce risk. The most useful sensitivity studies are those tied to realistic engineering choices rather than arbitrary numerical changes.
Relationship with LOPA
QRA and LOPA use quantitative information at different levels. LOPA evaluates one defined scenario and its independent protection layers. QRA combines many scenarios to estimate cumulative risk across locations or populations. Results should be consistent where the same initiating events and barriers appear in both analyses.
Communicating QRA results
Risk contours and FN curves can appear precise, but they depend on assumptions and model choices. Reports should explain the dominant contributors, uncertainty and sensitivity rather than presenting only final plots. Decision-makers need to understand what drives the result and which risk controls provide meaningful improvement.
Keeping the QRA current
A QRA should be revisited when major inventories, isolation philosophy, occupancy, layout or consequence assumptions change. Brownfield modifications can alter several scenarios at once, so management of change should identify whether the existing model still represents the facility. Periodic updates help keep risk contours and major contributors aligned with the plant that is actually operating.
The updated model should retain traceability to the underlying hazard studies and current operating basis.