This project supported a major data centre client in London by using computational fluid dynamics (CFD) to assess the interaction between wind, heat rejection equipment, and rooftop plant performance. The client required assurance that the external roof plant arrangement could maintain resilient cooling performance across a range of wind conditions, while also understanding whether any localised risks could affect operational reliability.

The assessment focused on a large rooftop plant area containing several critical systems, including dry air coolers (DACs), air-cooled chillers, hybrid coolers, and the exhaust discharge from diesel rotary uninterruptible power supply (DRUPS) units. These systems were represented in the CFD model to capture their combined influence on local airflow, heat plume behaviour, and intake conditions. Particular attention was given to the possibility of warm air recirculation, where discharged heat from one item of equipment could be drawn back into nearby cooler intakes, reducing available thermal performance margin.

The methodology began with the development of a representative three-dimensional model of the rooftop environment, including the main building geometry, plant layout, equipment heights, discharge directions, intake locations, and surrounding obstructions. Key heat-rejection sources and exhaust streams were included with appropriate thermal and flow boundary conditions, enabling the model to resolve how plumes developed and dispersed above the roof.

A range of wind scenarios was then simulated to reflect the sensitivity of the roof plant to changes in wind direction and speed. This was important because rooftop plant performance is rarely governed by still-air conditions alone. Crosswinds, sheltered zones, taller structures, and exhaust momentum can all influence whether heated discharge is carried away safely or redirected towards sensitive intake locations.

The CFD analysis examined several performance indicators, including airflow patterns across the roof, temperature rise at equipment intakes, plume trajectories from heat rejection systems, and the extent of any recirculation zones. Results were reviewed both globally, to understand the overall roof airflow behaviour, and locally, to identify whether particular items of plant were more exposed to wind-driven effects.

The study found that, under typical operating conditions, most rooftop equipment operated with limited cross-interaction. Heat plumes generally dispersed without creating sustained adverse intake conditions. However, the analysis also identified some localised areas where wind direction, equipment height, or exhaust discharge could modestly increase intake temperatures. These effects were not considered severe enough to require additional physical mitigation, but they were useful in defining practical operational responses.

Where elevated intake conditions were indicated, temporary operational measures, such as fan speed adjustments in line with manufacturer guidance, were considered sufficient to maintain performance margins. This helped the client avoid unnecessary capital works while still having a clear risk-based response strategy.

From a risk-management perspective, the CFD assessment provided confidence that intake criteria could be maintained through routine operational controls and monitoring. The likelihood of sustained adverse intake conditions was assessed as low, and the consequences were considered manageable through pre-defined operational procedures.

The project also provided wider value for data-driven operations. By identifying roof areas most influenced by wind effects, the results supported more informed control set-points, alarm thresholds, inspection priorities, and maintenance planning. Overall, the analysis gave the client high-level assurance that the rooftop plant arrangement and control philosophy were suitable for resilient cooling operation, while providing clear guidance for ongoing performance monitoring and operational risk management.