Physical Address
304 North Cardinal St.
Dorchester Center, MA 02124
ISO/TR 26369:2009 — Fire Safety Engineering — Assessment, Verification and Validation of Calculation Methods
A Framework for Evaluating the Suitability and Accuracy of Fire Models Used in Performance-Based Design
1. The Role of Calculation Methods in Fire Safety Engineering
ISO/TR 26369:2009 addresses a fundamental need in performance-based fire safety engineering: ensuring that the calculation methods and computer models used to demonstrate fire safety performance are fit for purpose. As building codes worldwide increasingly permit performance-based alternatives to prescriptive fire safety requirements, the reliance on fire models — ranging from simple hand calculations to complex computational fluid dynamics (CFD) simulations — has grown dramatically.
The scope of ISO/TR 26369 encompasses all forms of calculation methods used in fire safety engineering, including analytical equations (plume correlations, flame height calculations), zone models (CFAST, B-RISK), field models based on CFD (FDS, FireFOAM), evacuation models (Pathfinder, STEPS), and structural fire resistance calculation methods. Each category presents different challenges for verification and validation.
The effort invested in verification and validation should be proportional to the risk associated with the design decision. A fire model used to justify a 50% reduction in exit width demands far more rigorous validation than a model used for qualitative smoke visualization.
| Model Category | Examples | Typical Applications | Validation Approach |
|---|---|---|---|
| Hand calculations | Alpert correlations, McCaffrey plume | Detector activation, smoke layer height | Compare with experimental database |
| Zone models | CFAST, B-RISK | Two-layer smoke filling, tenability | Benchmark against full-scale tests |
| CFD models | FDS, FireFOAM, OpenFOAM | Smoke transport, sprinkler activation | ISO 9705 room corner test; grid sensitivity |
| Evacuation models | Pathfinder, STEPS, buildingEXODUS | Egress time, occupant flow, bottlenecks | Drill data validation; behavioral sensitivity |
| Structural fire models | SAFIR, ABAQUS (thermal) | Steel protection, concrete spalling | Furnace test replication; material verification |
2. Framework for Assessment, Verification, and Validation
ISO/TR 26369 establishes a three-part framework for evaluating calculation methods: assessment, verification, and validation. While these terms are sometimes used interchangeably, the Technical Report carefully distinguishes them.
Verification confirms that the method correctly implements its underlying mathematical and physical principles. For software models, this means checking that the code solves the governing equations correctly and that numerical algorithms are properly implemented. Verification is typically performed by comparing outputs with analytical solutions, performing code-to-code comparisons, and conducting numerical convergence studies such as grid sensitivity analysis for CFD models.
A model extensively validated against experimental data can still produce incorrect results if applied outside its verified range. Always check both the verification status and the validation envelope before applying a fire model to a specific design scenario.
Validation determines whether the model adequately represents real-world fire behavior by comparing predictions with experimental data. ISO/TR 26369 provides guidance on selecting appropriate validation datasets representing the physical phenomena relevant to the intended application. Validation is scenario-specific: a model validated for pre-flashover compartment fires may not be valid for post-flashover conditions.
Assessment is the overarching process of evaluating whether a method is appropriate for a specific design application, considering verification and validation status, user qualifications, input data quality, uncertainty treatment, and documentation of assumptions.
3. Practical Implementation and Quality Assurance
ISO/TR 26369 provides practical guidance for implementing the framework in engineering practice. The quality of fire modeling outputs depends not only on the models but on how they are applied, documented, and reviewed.
Model selection guidance helps engineers choose the appropriate level of sophistication. Simple hand calculations may suffice for prescriptive compliance, while CFD may be necessary for complex geometries or performance-based alternatives. The simplest adequate model should be used, as more complex models introduce additional uncertainties that must be carefully quantified. Engineers should also consider the computational resources available, the time constraints of the project schedule, and the level of expertise required to properly operate the selected model. A common mistake is selecting a model based on familiarity rather than suitability for the specific fire scenario being analyzed.
Equally important is the quality of input data. Fire models are highly sensitive to input parameters such as fire heat release rate, material thermal properties, and ventilation conditions. Engineers should use conservative values or ranges for input parameters where precise data is unavailable, and clearly document all assumptions and data sources in the modeling report. Sensitivity analysis should be performed on key input parameters to understand their influence on modeling results and to identify which parameters warrant the most effort in characterization. This systematic approach to input data quality directly impacts the reliability of fire modeling predictions and the defensibility of design decisions.
Engineering firms that implement formal verification and validation protocols based on ISO/TR 26369 report 40-60% fewer peer review comments on their fire modeling work, as systematic documentation provides clear evidence of technical rigor.
Uncertainty and sensitivity analysis are essential. All fire models involve uncertainties from input parameters, numerical approximations, and physical sub-models. ISO/TR 26369 requires these uncertainties be characterized and quantified through sensitivity studies.
Documentation requirements include: a clear statement of the design question, description of the selected model, detailed input data with sources, verification evidence, validation evidence, uncertainty analysis, and clear presentation of results with caveats about limitations. Well-documented modeling studies facilitate peer review, enable reproduction by other engineers, and provide defensible evidence in regulatory approval processes or legal proceedings related to fire safety performance.
Peer review is strongly recommended. Independent review by qualified engineers not involved in the original modeling can identify errors or inappropriate assumptions that might otherwise go undetected.
Frequently Asked Questions
Q1: What is the difference between verification and validation?
Verification answers Did we build the model correctly? — checking that the implementation is error-free. Validation answers Did we build the right model? — checking that predictions match real-world behavior.
Verification answers Did we build the model correctly? — checking that the implementation is error-free. Validation answers Did we build the right model? — checking that predictions match real-world behavior.
Q2: Does ISO/TR 26369 apply to BIM-integrated fire modeling?
Yes. The principles apply regardless of platform. Additional verification steps are needed to confirm correct data transfer between BIM and fire modeling platforms.
Yes. The principles apply regardless of platform. Additional verification steps are needed to confirm correct data transfer between BIM and fire modeling platforms.
Q3: How should engineers handle models with limited validation data?
ISO/TR 26369 recommends a conservative approach: use bounding analysis, employ multiple models for cross-checking, increase safety factors, and clearly document limitations.
ISO/TR 26369 recommends a conservative approach: use bounding analysis, employ multiple models for cross-checking, increase safety factors, and clearly document limitations.
Q4: Are there certification schemes for fire modeling software?
There is no formal ISO certification, but several organizations maintain qualified product lists. Engineers should verify their chosen model has been validated against benchmarks relevant to their specific application.
There is no formal ISO certification, but several organizations maintain qualified product lists. Engineers should verify their chosen model has been validated against benchmarks relevant to their specific application.
