A quantitative risk assessment can look complete while still containing weak assumptions, missing scenarios or results that cannot be traced back to source data. An independent QRA review tests whether the study is technically defensible before it is used for a statutory submission, investment decision, plant-layout change or risk-acceptance decision.
This checklist explains what a competent reviewer should verify. It is intended for project owners, EHS leaders, engineering teams, insurers and regulators reviewing a study prepared by an internal team or a third-party consultant. For information about commissioning a complete study, see Elion’s Quantitative Risk Assessment service.
1. Confirm the study basis and decision objective
A review should begin with the purpose of the QRA. A study prepared for environmental clearance may require a different reporting emphasis from one used for occupied-building siting, emergency planning or comparison of design alternatives.
- Is the facility boundary and study scope clearly defined?
- Are applicable acceptance criteria and regulatory references stated?
- Are plot plans, process descriptions and population data current?
- Are assumptions, exclusions and data gaps visible to the decision-maker?
2. Validate hazardous inventories and scenario selection
The scenario register is the foundation of the model. The reviewer should reconcile inventories, temperatures, pressures, compositions, storage conditions and release points against PFDs, P&IDs, equipment data sheets and site observations.
- Have credible loss-of-containment sizes been considered?
- Are loading, unloading, transfer, storage and process operations covered?
- Are isolation time, detection, shutdown and inventory segmentation justified?
- Are toxic, flammable and explosive outcomes screened consistently?
3. Audit the event-frequency basis
Frequency calculations should be reproducible. Generic failure data must be appropriate to the equipment and operating context, and any credit for safeguards must be supported by independence, reliability and proof-testing assumptions.
- Can each initiating frequency be traced to a recognised source?
- Are event trees complete and conditional probabilities explained?
- Has double counting between scenarios been avoided?
- Are ignition probabilities, weather distributions and occupancy factors suitable for the site?
4. Check consequence-model inputs and model selection
A reviewer should verify that the software model matches the physical behaviour of the release. This includes discharge, evaporation, dense or passive dispersion, jet fire, pool fire, flash fire, vapour-cloud explosion and toxic exposure as applicable.
- Are hole size, release direction, duration and surface conditions reasonable?
- Are weather classes, roughness, temperature and humidity representative?
- Are radiation, overpressure and toxic endpoints stated with their basis?
- Have congestion, confinement and terrain limitations been recognised?
5. Reconcile risk contours, F-N curves and ALARP conclusions
Individual-risk contours and societal-risk F-N curves must be generated from the same documented scenario set. A visual contour alone is not enough: the reviewer should be able to trace material risk contributors and confirm that population and occupancy inputs are suitable.
- Do the numerical results agree with the maps and tables?
- Are dominant risk contributors identified?
- Are onsite and offsite populations treated consistently?
- Does the ALARP demonstration compare reasonably practicable options rather than simply state that risk is acceptable?
6. Test whether recommendations are actionable
Recommendations should identify the risk driver, proposed control, responsible discipline and priority. Vague instructions such as “improve safety” do not support implementation. Where a safeguard is credited in the model, its performance requirement should be clear enough to maintain during operation.
Common warning signs in a QRA report
- Copy-pasted assumptions or site names from another facility
- Risk contours without a documented scenario-frequency register
- No reconciliation between model inputs and process documents
- Unexplained software defaults or acceptance criteria
- Recommendations that do not address the dominant contributors
- No revision trail after design or inventory changes
When an independent QRA review is valuable
Independent validation is particularly useful before regulatory submission, financial close, major expansion, land-use or occupied-building decisions, and acceptance of a consultant’s final report. It can also be used after an incident or management-of-change review to test whether the existing QRA remains valid.
Expected review deliverables
A structured review normally produces a comment register graded by technical significance, a marked-up report, a calculation and input check, and a close-out record showing how each issue was resolved. The objective is not to reproduce the entire study; it is to establish that the model, assumptions, results and decisions form a traceable chain of evidence.
Elion can perform an independent technical review or a complete QRA assignment. Contact the QRA team with the study report, facility type and decision objective for a scoped review.