CFD (Computational Fluid Dynamics) is used in wind engineering to simulate how wind moves around buildings, through urban areas, and across complex terrain — entirely within a computer. It lets engineers predict pedestrian wind comfort, wind loading on facades, and airflow patterns before anything is built. Below, we answer the most common questions about how CFD works in practice, which standards it satisfies, and when to bring it into your project.
What can CFD reveal that traditional wind tunnels cannot?
CFD can analyse wind behaviour across entire city districts, masterplans, and large urban areas where a physical wind tunnel model simply is not practical. While wind tunnel testing remains highly accurate for individual buildings and smaller developments, it requires a physical scale model — which sets a hard limit on how large or complex the study area can be.
With CFD, there is no such limit. A simulation can cover a full urban district with hundreds of buildings, multiple street orientations, and varying terrain heights. The Rotterdam city-wide wind study is a good example: the computational domain covered a diameter of 8,000 metres, with a mesh of more than 583 million cells — roughly 20 to 30 times more than a typical single-building study. That kind of scale is simply not achievable in a physical wind tunnel.
CFD also gives you something wind tunnels cannot: complete spatial coverage. Instead of measuring at a fixed set of sensor points, CFD produces continuous flow data across the entire model. This means you can generate detailed wind maps at pedestrian height, identify problem zones that were not anticipated, and extract multiple thematic outputs from a single simulation run — pollutant dispersion, urban ventilation potential, and wind energy yield, for example.
The two methods are complementary rather than competing. For wind engineering projects involving individual high-rise buildings or facade pressure testing, a wind tunnel often remains the right choice. For large-scale area assessments, CFD is the method that makes the study possible at all.
How does a CFD simulation for a building project actually work?
A CFD wind study for a building project follows three main steps: building a 3D model of the site and its surroundings, generating a computational mesh that divides the air volume into millions of cells, and running the simulation across multiple wind directions before processing the results into usable maps and reports.
Step 1: Building the 3D model
The model is constructed using geodata sources such as building footprint databases, elevation models, and city 3D datasets. Existing and permitted buildings are included in the model. Details smaller than roughly one metre are typically left out — they have negligible influence on the results but would significantly increase computation time. Vegetation is included where it forms larger clusters that meaningfully affect airflow. A surrounding buffer zone of at least 250 metres is added around the core study area to ensure wind enters the model realistically.
Step 2: Creating the computational mesh
The air volume above and between the buildings is divided into a grid of cells — the mesh. Cell sizes vary: very fine near building surfaces (as small as 0.25 metres) to capture the velocity gradient in the atmospheric boundary layer accurately, and much coarser at the outer edges of the domain (up to 64 metres). Five prismatic cell layers are extruded from wall surfaces to handle near-wall airflow correctly. The total cell count depends directly on the scale of the project.
Step 3: Running simulations and post-processing
The simulation runs for each relevant wind direction using a steady-state solver. Meteorological input — wind speed distributions, wind roses, and directional frequencies — comes from measured datasets such as KNMI data. Once the raw simulation results are available, they are processed into colour-coded maps and visualisations. Ideally, the specific outputs needed are agreed in advance with designers and decision-makers so that one simulation run can produce multiple useful thematic maps.
Which wind engineering standards does CFD satisfy?
CFD can satisfy both the Dutch NEN 8100 standard and the international Lawson criteria, depending on where your project is located and what the permit authority requires. The method itself is not the limiting factor — what matters is that the simulation inputs, boundary conditions, and post-processing follow the requirements of the applicable standard.
For projects in the Netherlands, NEN 8100 is the standard used in permit procedures. It assesses pedestrian wind comfort by measuring the probability that wind speed at eye height (1.75 metres) exceeds 5 m/s, and wind danger by measuring the probability that it exceeds 15 m/s. Meteorological input must follow NPR 6097:2006 for permit-relevant studies, using wind data from the 1963 to 2002 measurement period. CFD results processed against this dataset are directly usable in permit applications.
For projects in the UK and other international markets, the Lawson criteria are the standard reference. The underlying simulation methodology is the same; the post-processing and classification framework differs. CFD studies following the Lawson criteria are accepted by planning authorities in cities including London and Leeds.
Beyond comfort and danger assessments, CFD also supports wind loading studies for facades and structures, referenced against Eurocode EN 1991-1-4. The full range of applications spans pedestrian comfort, structural loading, pollutant dispersion, and urban climate analysis — all from the same simulation infrastructure.
What types of wind problems is CFD best suited to solve?
CFD is best suited to wind problems that involve large areas, complex urban geometry, or the need for spatially continuous data rather than point measurements. It is particularly useful when the study needs to inform design decisions at an early stage, when the project area is too large for a physical wind tunnel, or when multiple wind-related questions need to be answered from a single study.
The types of problems CFD handles well include:
- Pedestrian wind comfort assessments around new buildings, towers, or area developments — identifying locations where wind speeds at street level exceed comfort or safety thresholds
- Large-scale masterplan studies covering multiple city blocks or entire districts, where the interaction between buildings at neighbourhood scale drives the wind conditions
- Facade wind loading for structural and cladding design, generating pressure distributions across building surfaces
- Pollutant and odour dispersion studies, tracing how airborne contaminants spread from sources such as traffic corridors or industrial exhausts
- Urban ventilation and heat stress analysis, identifying areas with insufficient airflow for cooling or air quality management
- Rooftop wind energy potential, calculating expected energy yield per square metre of rotor area across a building or district
CFD is less suited to situations where highly detailed pressure measurements on a specific facade are needed for structural certification — in those cases, a physical wind tunnel test often produces more defensible results for the permit authority. The two methods work best when chosen based on the specific question being asked, not as a default preference for one over the other.
How long does a CFD wind study take to complete?
A standard CFD wind comfort study for a single building or modest development typically takes one to three weeks from receiving the necessary inputs to delivering the final report. Large-scale area studies covering entire districts can take significantly longer, depending on the size of the model domain, the number of wind directions simulated, and the complexity of the post-processing required.
The main time drivers in a CFD study are:
- Model preparation: Building the 3D geometry and mesh is often the most time-intensive step, particularly when geodata quality is inconsistent or the study area is large
- Simulation runs: Each wind direction requires a separate simulation. A full study typically covers 12 wind directions, and some projects run separate simulations for different seasons
- Post-processing and reporting: Converting raw simulation data into maps, classifying results against the applicable standard, and writing a report that holds up under regulatory scrutiny all take time
Turnaround time is also heavily influenced by how quickly you can provide the inputs — site geometry, surrounding building data, and a clear brief about the study scope. Projects where those inputs arrive complete and well-organised move significantly faster than those where the brief evolves during the study.
When should CFD be introduced into the design process?
CFD should ideally be introduced at the earliest stage where building volumes and site layout are sufficiently defined — typically during the schematic design phase, well before permit application. Introducing wind analysis early gives designers the ability to adjust building orientation, massing, and street layout based on the results, rather than discovering wind problems after the design is largely fixed.
Wind problems are significantly harder and more expensive to solve at the detailed design stage. If a pedestrian wind comfort assessment reveals that a corner of a building creates dangerous downwash conditions, the most effective remedies — adjusting building height, adding a setback, or reorienting the facade — are only practical when the design is still flexible. Late-stage discoveries typically leave only less effective options: canopies, screens, and landscaping that mitigate but rarely fully resolve the problem.
A useful rule of thumb: if your building exceeds roughly twice the height of its immediate surroundings, a wind study is worth commissioning as soon as you have a stable massing model. For area developments and masterplans, wind analysis is most valuable when it can directly inform the urban design framework — street orientation, building cluster arrangement, and the placement of public spaces relative to the prevailing wind direction.
For permit applications in the Netherlands, NEN 8100 assessments are often required before a permit is granted. Starting the study early enough to allow for design iterations — rather than treating it as a box-ticking exercise at the end — is the most effective way to use the results.
How Actiflow helps with CFD in wind engineering
We have been applying CFD to wind engineering projects since 2005, working with architects, developers, structural engineers, and municipalities across the Netherlands, Belgium, the UK, and beyond. Our team uses advanced virtual wind tunnel technology — validated CFD algorithms running on our own high-performance computing cluster — to produce results that are accurate, defensible, and directly usable in permit procedures.
Here is what working with us looks like in practice:
- We advise you on which study type and which standard applies to your project — NEN 8100 for Dutch permit procedures, Lawson for international projects
- We build the 3D model from geodata and your design files, so you do not need to prepare a specialist input format
- We deliver colour-coded wind maps and clear graphics that you can present directly to clients, planners, and permit authorities
- We cover the full range of project scales — from a single high-rise to city-wide area assessments like the study we carried out for the city of Rotterdam
- We are known for speed and flexibility: for regular clients, we can set everything aside and start the next day if the deadline demands it
- We provide a single point of contact from intake to final report, so nothing gets lost between the technical work and the deliverable you actually need
Curious how we can help with CFD in wind engineering? Contact us — we would be happy to discuss your project and help you find the right engineering solution. You can also find out more about our team and background on our about us page.
