Natural Ventilation design

Millennium Tower, Nairobi

Battle McCarthy

  • Introduction: Overview of objectives and outline of the approach adopted.
  • Wind Data: The major environmental force to which the design must respond.
  • Proposed Design:Position, size and shape of the key natural ventilation features.
  • Modelling Methodology
  • Modelling Results:
    1. Wind Pressure on the Facade: An overall model of the building and its environment forming the base for al further studies.
    2. Wind Scoop on the Roof: Optimising the size and shape of the openning and the internal geometry.
    3. Air Exhaust Outlet: Optimising the size and position of the air exhaust outlet.
    4. Wind flow adjacent to the Fins:Quantifying wind pressure adjacent to the fins
    5. Wind Shadow. Investigating the effect of the negative pressure shadow cast by the top of the flue on the opening to the wind scoop.


    This webpage provides an overview of studies into the thermal performance of the Millennium Tower in Nairobi, by Battle McCarthy.

    The work aimed to maximise the potential for natural ventilation of the proposed building by the study of internal conditions for variations of the building form. This has enabled recommendations to be made for the positions and sizes of the key elements of the natural ventilation design.

    The study of natural ventilation within the proposed building determined the optimum size, shape and configurations for the following key design features:
    The exhaust air outlet at roof level The wind scoop at roof level The inlets to supply shafts at the base of the tower Flue infiltration at its base The typical office floor supply and exhaust arrangement

    Computational Fluid Dynamics CFD has been used to study and optimise design .

    Architect's Model of the building (left) and Proposed natural ventilation design (Cross section, Right)

    tower model proposed design-vertical section

    Proposed natural ventilation design (plane of the standard floors)

    proposed design-plane

    Back to the Top

    Wind Data

    Weather data used was as recorded at Nairobi Kenyatta Airport in 1995 but with air temperatures as recorded at a weather station closer to the proposed site. The wind rose included below clearly shows the predominance of winds from the NE, which is significant for the positioning of inlets and outlets. The generally constant wind direction will allow the use of fixed inlets and outlets devices.

    Wind rose for Nairobi and accompanying vertical wind speed profile


    Back to the Top

    Modelling Methodology

    For this project the building and its surrounding environment have been considered as an integrated object. A virtual wind tunnel experiment was carried out to estimate the wind pressure over the façade of the building and study the external flow pattern around the building. This was followed by a number of detailed modelling studies, each focusing on a single design issue. Findings from one area of study would contribute to subsequent studies where considered appropriate.

    The predicted wind pressure distribution determines the driving force acting on the openings of the building envelope. These results provided input data for a bulk air model, which enabled prediction of air movement inside the building, that is, between the vertical shafts, offices, and the flue. Two dominant wind directions, NE and E, were considered. An exponential function was used to determine the vertical wind velocity profile from ground level to a height of 150 m.

    The detailed CFD modelling studies consisted of four parts:

    1. Optimising the rooftop wind scoop
    2. Optimising the exhaust air outlet
    3. Determining the extent of the negative pressure shadow cast by the exhaust flue
    4. Determinig the pressure around the fins on each side of the building

    Back to the Top

    Pressure on the Building Facade

    A computer model was created representing the whole building within its immediate environment. Results provided the designer with a greater insight into the interaction of the proposed building with the wind. The information provided included flow pattern, velocities and pressure. Two dominant wind directions, NE and E, were considered.

    The following images show CFD output from this study.

    Velocity vectors coloured by speed (m/s) and pressure contours on a vertical plane (Pa), NE wind

    velocity pressure

    Velocity vectors showing the wind flow over the roof and wind scoop, NE wind


    Back to the Top

    Optimising Design of The Wind Scoop

    A further CFD study provided a quantitative analysis leading to a wind scoop design which increased the air flow into the building for a given wind speed.

    Three designs with different size and shape of opening and internal geometry were compared. The criteria for evaluation was the volumetric flow rate into the connecting vertical duct. The graph included below shows that an opening with a curved internal profile (represented by the white line with yellow triangular markers) allows the greatest volumetric air flow rate into the connecting duct. A CFD image from this study is shown adjacent to the graph.

    Speed profiles across the horizontal section of the connecting duct (Left) and Velocity vectors coloured by pressure along the central vertical plane of the wind scoop (Right)

    wind scoop wind scoop

    Back to the Top

    The Exhaust Air Outlet

    CFD modelling showed that a wall positioned infront of the exhaust air outlet would reduce the air flow speed from the flue. Without the wall, the air flow from the flue is greater.

    Original Design with adjacent wall (Left) and improved design (Right)

    millflue1 .millflue2

    Back to the Top

    Pressure around the fins

    CFD modelling was used to study the pressure adjacent to the fins on each side of the building. Results indicated that the pressure difference between the two sides was much smaller than that between the study floor and the wind scoop. This will ensure that air moves down to this level from the wind scoop and not directly across the space.

    Flow around building and fins. Velocity vectors coloured by pressure

    Details one: The fins on the SE facade


    Details two: The fins on the NW facade

    Back to the Top

    Negative Pressure shadow cast by the Exhaust Air Outlet

    Whilst the major driving force for air flow into the building is the wind pressure at the wind scoop, the pressure difference through the building will be affected by the pressure at the exhaust air outlet. CFD was used to quantify the negative pressure region adjacent to the exhaust air outlet for the dominant NE wind and to determine the extent to which this region extends towards the wind scoops.

    Flow pattern over the roof and around the wind scoops.



    Pressure contours on the level of the thermal flue.


    This study found that the the negative pressure region extends along the full length of the perimeter exhaust flue and suggested that the lower edge of the opening to the wind scoop should be at least 2 meters above rooflevel to ensure positive pressure at the wind scoop for the two dominant wind directions.

    Back to the Top

    © Battle McCarthy