Smoke doesn't follow prescriptive rules. It drifts along ceiling voids, pools in atriums, and behaves differently depending on ventilation conditions, room geometry, and what's burning. A code written in words can't fully anticipate that. What it can do  and what SBC 201 does is allow engineers to demonstrate that a building is safe through analysis rather than prescriptive compliance alone. CFD fire modelling is often how that analysis gets done.

What CFD Fire Modelling Actually Is

CFD is computational fluid dynamics, a branch of fluid mechanics that uses numerical methods to simulate how gases move and interact. In fire engineering, it's used to model how heat and smoke develop and travel through a built space from an ignition source.

The standard tool in the sector is FDS  Fire Dynamics Simulator  developed by the National Institute of Standards and Technology in the US and used widely in Canada and internationally. FDS can model combustion, smoke transport, sprinkler activation, detector response, and structural heating. It's not a visualisation tool or a rough approximation; it solves the equations governing thermally driven fluid flow, and its outputs have been validated against experimental fire data over many years.

For a fire safety engineer, this means building a digital model of a space, defining a fire scenario with specific fuel type, growth rate, and location, then running the simulation and reading what the model produces: smoke layer heights over time, temperature at various elevations, visibility at occupant head height, carbon monoxide concentrations. Those outputs feed directly into the tenability and egress assessment required under SBC 201.

Where CFD Fits in the SBC 201 Compliance Process

SBC 201 permits performance-based compliance as an alternative to its prescriptive provisions. Under this path, the design team has to show that the building achieves the same level of safety the prescriptive code would provide -- even if the specific design is different. CFD is typically how the fire side of that argument gets made.

The process involves selecting design fire scenarios -- the credible worst-case fires the building's systems need to handle -- then using CFD to model fire development and smoke spread, and pairing that with egress analysis to compare available safe egress time (ASET) against required safe egress time (RSET). If ASET exceeds RSET by an appropriate margin, the performance-based case holds. If not, something in the design needs to change.

CFD doesn't stand alone in this. It's one input into a broader engineering argument that also covers egress modelling, system reliability, uncertainty, and life safety objectives. The simulation is evidence, not the conclusion.

What Separates a Credible Study from a Weak One

An AHJ reviewing a performance-based SBC 201 submission will look carefully at the fire modelling, and they'll know what to look for. The quality of the inputs matters more than most people assume.

Scenario selection is where studies most often go wrong. A study that models a single fire in a convenient location, with heat release rates set to produce reassuring results, won't survive serious scrutiny. A credible study identifies the scenarios that genuinely test the design, applies conservative assumptions with published justification, and considers ventilation failure modes and system reliability.

Mesh resolution matters too. CFD simulations are discretised -- the space is divided into cells, and the accuracy of the output depends on how fine that grid is. A coarse mesh runs faster but produces less reliable results, particularly near fire sources and in narrow corridors. The choice of mesh should be justified, not just assumed.

Documentation has to explain all of it: the software version, the domain size, the mesh resolution, the fire scenario parameters and where they came from, and the uncertainty inherent in the results. A study that presents only the outputs without showing its working will not satisfy a technically confident reviewer.

Projects in Canada Where CFD Modelling Is Commonly Used

Large open-plan spaces are the most common application: atriums, airport terminals, shopping centres, warehouses. The smoke management challenges in these spaces are significant, the geometry doesn't map cleanly onto prescriptive provisions, and the consequences of getting it wrong are serious.

Heritage buildings come up regularly, because compartmentalisation can't always be added without destroying historic fabric. Underground parking structures have ventilation challenges that CFD helps quantify. Mixed-use developments with unusual occupancy stacking often need modelling to demonstrate that egress provisions are adequate across all occupancies simultaneously.

What these projects share is that the prescriptive code doesn't give a clear path forward, and engineering analysis is the only way to show the building is safe.

Our CFD Modelling Work

Our fire protection engineers have carried out CFD studies on projects across Canada, using FDS and associated post-processing tools. We've worked alongside design teams, AHJs, and peer reviewers to develop performance-based SBC 201 submissions that are technically rigorous and clearly documented.

We put particular effort into scenario selection and input justification, because those are the parts that get challenged. And we write the reports so that an architect, a developer, or a building official can follow the argument - not just another fire engineer.

Conclusion

 SBC 201's performance-based path still requires demonstrating equivalency to the prescriptive level of safety. A simulation alone doesn't do that -- it's one part of a wider engineering argument that has to address life safety objectives, system reliability, and uncertainty. The modelling provides evidence; the engineer interprets it and makes the case. To find out more about how we deliver this work, see our CFD fire modelling service page.

FAQs

How accurate is FDS?

FDS has been validated against a substantial body of experimental data and is widely accepted by regulators in Canada and internationally. Accuracy depends heavily on input quality -- the fire growth curve and ventilation conditions matter most. Engineers working to accepted practice use conservative assumptions and run sensitivity analyses to account for what's uncertain.

How long does a study take?

For a mid-complexity project, four to eight weeks is a reasonable range, covering scenario development, simulation runs, post-processing, and the report. Projects with more fire scenarios, iterative design changes, or peer review requirements take longer.

Does having a sprinkler system mean we don't need CFD?

Not necessarily. Sprinklers change fire development significantly, and FDS can model their activation and effect. Whether CFD is needed depends on the compliance path and the design challenges specific to your project. If you're taking a performance-based approach under SBC 201, you probably need modelling of some kind -- but a consultant can give you a clearer answer once they understand the project.