CFD Fire Modelling: When and How to Use It

Around the world, more and more buildings now include large atria, deep basement car parks, malls and tall mixed-use towers. These are exactly the spaces where prescriptive smoke control rules are hardest to apply with confidence. Computational fluid dynamics (CFD) fire modelling lets us see how smoke actually moves through a space, whether occupants can get out safely, and whether the smoke control system does what it is designed to do. In this post we explain what CFD fire modelling is, when a project is likely to need it, what we check, and how a sound study is put together.

What is CFD fire modelling?

How FDS and PyroSim work, in plain words

The most widely used tool is FDS (Fire Dynamics Simulator), developed by NIST, which we usually drive through the PyroSim interface. The building is divided into a three-dimensional grid of small cells. For each cell, the software calculates how heat, smoke and gases flow over time, taking account of the design fire, openings, fans and make-up air. The result is a time-based picture of temperature, visibility, gas concentrations and smoke layer movement.

CFD compared with hand calculations

Not every project needs CFD. For a simple, regular space, a zone model or hand calculation (such as the smoke-filling and descent approach we cover in our smoke descent analysis post) can be enough. CFD earns its place when the geometry is complex, when there are multiple connected levels, when airflow from fans or openings dominates, or when we need to show local conditions along escape routes rather than an average layer.

When does a project need a CFD smoke study?

Most building codes are largely prescriptive. Examples include NFPA 101 (Life Safety Code) and NFPA 92 in many jurisdictions, the National Building Code of India (NBC 2016 Part 4), the UAE Fire and Life Safety Code of Practice, and the Saudi Building Code SBC 801 (Fire Code). They set requirements for exits, compartmentation, smoke control and fire protection systems. In practice, CFD studies tend to be requested in the following situations. Authorities and peer reviewers often ask for them, though whether one is required always depends on the project and the authority having jurisdiction.

Atria and malls

Atria connect several floors, so smoke from a fire at the base can threaten all of them. NFPA 92 (Standard for Smoke Control Systems) covers smoke management in large-volume spaces, and NFPA 101 (Life Safety Code) contains specific provisions for atria. CFD is commonly used to check that the atrium smoke exhaust keeps the smoke layer above occupied levels for long enough. Our smoke management course covers the design principles.

Basement car parks

Jet fan ventilation systems are often designed using simple air-change calculations. A CFD study shows whether the system actually clears smoke, protects the routes firefighters use to approach, and avoids dead zones where smoke builds up.

High-rise and mixed-use towers

In tall buildings, stair pressurisation, lift shafts, stack effect and the smoke extract system all interact. CFD, often used alongside airflow network tools such as Ventus, helps us check that these systems work together. See our Ventus smoke control course.

Performance-based alternatives

When a design cannot reasonably meet a prescriptive requirement (for example an extended travel distance or an open, unusual layout), a performance-based approach may be proposed for the authority to accept. NFPA 101 Chapter 5 sets out a recognised performance-based option, built on defined design fire scenarios and stated performance criteria. CFD fire modelling, paired with evacuation analysis, is the usual evidence for such a case. Codes such as NBC 2016 Part 4, the UAE Fire and Life Safety Code and SBC 801 also have routes for authorities to consider alternative or engineered solutions, so check the applicable code and agree the approach with the authority early. For NBC context, see our NBC 2016 life safety course and free NBC 2016 downloads.

Tenability criteria: what are we checking?

The four main measures

A smoke study tests conditions along escape routes against tenability criteria, normally:

• Visibility: can occupants see exit signs and find their way?

• Temperature: is the air temperature, and the radiant heat from the smoke layer, survivable?

• Toxic gases: carbon monoxide and other products, sometimes assessed as a fractional effective dose (FED).

• Smoke layer height: does the layer stay above head height for the time needed?

Where acceptance values come from

Acceptance values should always be taken from a cited source, such as NFPA 92 and its annex, the SFPE Handbook of Fire Protection Engineering, or CIBSE Guide E, and agreed with the authority or peer reviewer before modelling starts. We do not quote numbers from memory. A report that presents results without stating its criteria and their source is weak, whatever the results show.

ASET vs RSET: linking smoke to evacuation

The core question is simple. Is the available safe egress time (ASET) longer than the required safe egress time (RSET), with an adequate margin?

• ASET comes from the CFD model: the time until any tenability criterion is exceeded at a given location.

• RSET comes from evacuation analysis: detection, alarm, pre-movement and movement time. We use Pathfinder for the movement part. See our Pathfinder evacuation analysis course.

The comparison should be made location by location, not just for the building as a whole. A common mistake is a comfortable margin overall but a tight one at a critical stair entry.

A typical CFD smoke study workflow

1. Inputs

• An architectural model, simplified sensibly. Detail that does not affect airflow only adds run time.

• The design fire: a heat release rate curve (growth, peak and any sprinkler-controlled plateau), plus fuel properties such as soot and CO yields, with justification for each.

• Ventilation: exhaust rates, make-up air routes and their timing, door positions and any pressurisation.

2. Mesh and sensitivity

Cell size must be fine enough to resolve the fire plume. FDS guidance relates this to the fire's characteristic diameter. A mesh sensitivity check shows that the conclusions do not depend on the grid. We cover this in PyroSim Module 1.

3. Scenarios and results

We run the agreed scenarios, typically different fire locations and failure cases. Results are taken from slice files, device measurements and Smokeview animations, all read against the tenability criteria at key times.

4. The report

A good report states its assumptions, the design fires and their basis, the mesh details and sensitivity check, each scenario, results against the stated criteria, the ASET/RSET comparison, and clear conclusions and limitations.

Common mistakes that get CFD reports rejected

• An unrealistically small design fire, or one with no justification.

• A mesh too coarse to resolve the plume, with no sensitivity run.

• Missing or unrealistic make-up air, which starves the exhaust.

• Sprinkler effects assumed without justification.

• Only one scenario modelled, with no failure cases.

• Attractive images presented without acceptance criteria or sources.

Learn it or outsource it?

Learn it. Start with our free interactive smoke management training. Our Interactive Training courses are free, with an optional paid certificate. For hands-on skills, we run paid PyroSim, Pathfinder and Ventus workshops. See our training and workshops.

Outsource it. As a global, cloud-based fire and life safety engineering firm, we carry out CFD fire modelling, smoke control design and evacuation analysis for clients worldwide, working to NFPA, NBC 2016, UAE Fire and Life Safety Code and Saudi SBC 801 requirements. See our fire engineering services and projects.

Send us your drawings for a CFD smoke study proposal: contact us, email info@fireandsmokesystems.com or call +91 96624 42111.

FAQ

Is CFD fire modelling mandatory?

Usually not by default. NFPA 101, NBC 2016 Part 4, the UAE Fire and Life Safety Code and SBC 801 are mostly prescriptive. CFD is typically used to support performance-based solutions or to demonstrate that complex smoke control designs work. Whether it is required depends on the project, the applicable code and the authority having jurisdiction.

Which software is used for CFD fire modelling?

Mostly FDS through the PyroSim interface, with Pathfinder for evacuation (RSET) and Ventus for airflow and pressurisation networks.

How long does a CFD smoke study take?

It depends on the number of scenarios, the model size and the review cycle. We confirm a programme once we have seen the drawings.

What does a CFD fire modelling report include?

The design fire and assumptions, mesh details and sensitivity check, visibility, temperature and CO results at key times, smoke layer height, an ASET/RSET comparison, and conclusions against stated tenability criteria.