Who Can See the Wind?

Natural ventilation is becoming popular for green buildings, but doing it properly requires carefully modelling the interaction of thermal and aerodynamic forces.

In its simplest form, natural ventilation is as simple as opening a window or door; it permits some form of air exchange with the outdoors. At the other end of the spectrum, natural ventilation implies an engineered balance of driving forces and pressure losses to move air through a building at predictable minimum flow rates, to provide adequate ventilation for air quality, for thermal comfort and to manage heat loads.

Humans have a long history with natural ventilation. It has been used to ventilate all-types of buildings from hospitals, schools and homes, to electrical sub-stations and industrial facilities.

Natural ventilation requires the management of the two principal driving forces: the buoyancy force associated with a temperature difference and the kinetic force associated with wind movement. In each case, pressure differences are set up between the indoor and outdoor environment, which drives the airflow. The restraint on natural ventilation is associated with pressure losses as air moves through openings (expansion and contraction of flow area), turns corners and passes through screens or filters.

Some buildings are difficult to ventilate naturally in a “managed” way. Very tall buildings can be difficult to ventilate naturally because the combination of wind and stack effect pressures can lead to adverse flow directions — people in one location receiving the stale air from those in another. Wind pressures can also lead to uncomfortable conditions. It is possible to naturally ventilate super-tall buildings but this feature must be designed-in from the beginning. Some jurisdictions (e. g. Chicago) require that all residences have operable windows, even though in some cases, these windows are not useful to the occupant.

In general, natural ventilation works best when the outside air temperatures are just below what would generally be considered comfortable indoor conditions. However, the outdoor temperature that will lead to acceptable indoor temperatures depends greatly on the internal heat loads (from human occupants, equipment, lights, solar gain, etc.) and the flow rate that can be achieved.

The American Society of Heating, Ventilating and Air Conditioning Engineers (ASHRAE) Standard 55 (2004) has a discussion on thermal comfort that includes a chart highlighting the acceptable indoor temperatures in naturally ventilated buildings. The key is that occupants must be given some measure of control over their environment.

Balancing driving forces and pressure losses

Design of natural ventilation in a complex building, or where natural ventilation is the only form of ventilation, can involve intuitive experience as well as hard science. In each case, the form of the building will be important and various features of the building’s architecture can enhance natural ventilation or work against it.

Designing natural ventilation is the art of balancing driving forces and pressure losses to achieve a desired minimum air flow rate. The required minimum flow rate need not be constant. In general there are three different criteria for identifying the minimum flow rate:

(1) the human biological requirement established in codes and by organizations (e. g. ASHRAE, NBCC, CIBSE) and sometimes simply referred to as 20 cfm/person (10 L/s/person);

(2) the flow required to maintain human or equipment temperature limits. ASHRAE 55 suggests that acceptable temperatures range from 17 to 31 C depending on the outdoor temperature;

(3) the flow to maintain contaminants (e. g. carbon dioxide concentrations) at a maximum allowable limit.

Various tools are available to assess natural ventilation and the use of one tool over another depends on the driving forces and critical nature of the flow.

To estimate thermal conditions, an energy model can be used. Wind pressures can be estimated using computational fluid dynamics (CFD) modelling. The power of CFD is it permits one to couple flow inside and outside the space. However, the use of CFD to predict wind pressures on complex buildings or in built-up surroundings can lead to errors. Wind tunnel modelling is the best tool to assess wind pressures.

Flows within the space can be assessed using a physical scale model (typically a water tank where brine serves as a proxy for heated air and the flow model is inverted), or alternatively again by using CFD models.

The use of natural ventilation systems can enhance the human environment with a minimal input of energy, but the design of robust natural ventilation systems requires a and analysis of different physics, as well as various tools to ensure the design’s viability.

Computational fluid dynamics

Computational fluid dynamics (CFD) is a method of solving the fundamental equations describing fluid flow using computational techniques. In executing a CFD simulation, the space to be simulated is divided up into small volumes called cells. This process is called gridding. Historically these cells have been hexahedrals (boxes), but they can be many different shapes. Generally, the smaller the size of these cells, the more accurate the solution. The equations of motion for fluid flow (conservations of mass and momentum) are then applied to these cells along with equations representing turbulence, energy, contaminants, radiation, moisture, etc.

The equations in each cell are numerically coupled to the equations in the adjacent cells, and the equation for upward momentum is coupled to the equation for temperature when buoyancy is involved. These equations for the sometimes millions of cells cannot be solved analytically: a computer solves the equations set iteratively.

The results from a CFD simulation include data at each cell for each of the solved parameters. These parameters can then be manipulated to generate predictions of other variables. As an example, air temperature, humidity, solar radiation and air speed can be combined into a measure of thermal comfort if one makes an assumption about what the occupant might be wearing (e. g. trousers and shirt vs. suit) and their activity level (e. g. sitting vs. walking briskly).

One of the great challenges in CFD modelling is accurately representing turbulence. In some simulated environments turbulence is reasonably well understood and represented (e. g. pipe flow). However in other environments turbulence plays a key role and is difficult to accurately predict (e. g. separation around a bluff body such as a building).

It is possible to generate a fully coupled CFD simulation of a flow field both around and within a naturally ventilated building where the flow through openings is predicted. However there is a risk that the simulation will yield incorrect total volumetric flow predictions through the building because the results rely on an accurate pres-review sure prediction at the building openings. Depending on the location of the opening, the shape of the building and the simulation set-up, the prediction may be significantly different from the actual answer.

The simulations presented in this article coupled wind tunnel testing and CFD modelling at the boundary of the building. In both cases, the pressures measured in a wind tunnel were imposed as input conditions to the CFD simulations.

Stack effect

Stack effect is a phenomenon present in all vertical shafts that are at different temperatures from outdoors. This includes chimneys and buildings. A building’s shafts include the vertical HVAC risers, elevator and stairwell shafts. The temperature difference sets up a scenario where there is a density difference indoors to out.

This pressure difference provides a driving force for air movement. If there are openings in the building faade (even if they are very small cracks) then there will be air flow
in at the bottom and out at the top for a heating- climate scenario.

Stack effect flows in shorter buildings and chimneys (e. g. 5 storeys) are a few Pascals, compared to wind pressures which can be measured in 10s of Pascals. Hence it is possible for a slight wind to overwhelm the natural stack effect. A robust design of natural ventilation in a building that uses stack effect as a driving force will be configured so that at worst the wind effects will be benign and if possible they will assist. It is important that all wind conditions and directions be considered because the “prevailing” wind is not always one that exists more than 50% of the time.

Natural ventilation at a U. S. high school

The Lynnwood high school located just north of Seattle is slated to open in the fall of 2009. Designed by Bassetti Archi- tects, the school was originally conceived as a building where the classrooms would be all naturally ventilated.

Supplemental heating was to be provided in the winter, and no fans would be installed. The classrooms are configured with inlets and windows in the side wall and a chimney for relief. The building was arranged north-south and the majority of the winds come from either the north or south. Temperatures in the region are considered moderate, with dry bulb temperatures above 22C only 3% of the time, including the summer. For 37% of the time the temperature is less than 7C. These lower temperatures, provide an ideal set of conditions for natural ventilation.

The design of the classrooms involved energy modelling, CFD simulations and wind tunnel testing.

In the end, all classrooms were identified as being at risk if the vents on the chimneys remained on one side only. Therefore the design was adjusted so that operable dampers were installed on both the north and south sides of the building. Control would be based on wind direction and speed. CFD modelling was used to confirm the in-room ventilation distribution. CFD modelling also assisted with the design of the inlet air box located beneath the windows. The objective of this work was to minimize pressure losses while developing multiple flow paths for different operating modes (heating, cooling, pre-heating).

The design work for the school demonstrated that many of the classrooms could be ventilated without requiring a fan. Active control of a damper was all that was required. Some classrooms, due to their function and location around the building perimeter could not be guaranteed to always be ventilated. They have supplementary fans just in case.

University in Saudi Arabia with solar chimney

The King Abdullah University of Science and Technology (KAUST) is being built near Rabigh, Saudi Arabia and will open in the fall of 2009. The west coast of Saudi Arabia can be a hot and humid environment during a large part of the year.

The use of solar chimneys was to provide a means whereby air movement would always occur through a semi-outdoor set of courtyards.

While shading and maximizing the availability of wind flow was used to enhance comfort in these semi-outdoors spaces, there are times when there will be little to no wind driven flow. Consequently, a pair of solar chimneys was devised to ensure that air movement would exist in specific areas of the campus. The chimneys create a stack effect by absorbing solar radiation and re-releasing it into the passing air stream.

The chimneys are arranged so that they have two layers of glass. The outer one is as transparent as possible to allow sunlight in. The inner layer is intended to be as absorptive as possible. Solar energy is absorbed and then released throughout the day into the air surrounding the inner absorptive glass. Not only does the inner glass absorb, it also serves as a thermal mass. The chimneys are approximately 60 metres tall and 8 x 16 square metres in area.

While the chimney is predicted to draw up to 400,000 cfm under ideal high solar conditions, the top of the chimney is arranged so that wind driven pressures will also move air through the chimney no matter what the wind direction. Dampers at the top are available to control flow in the event that too much air is driven through the chimney.

For the solar towers, different levels of wind tunnel testing, CFD modelling, energy modelling and screening calculations were performed. The figure on p. 35 shows one component of the CFD modelling that was used to test the absorption of solar energy and re-release it from the glass back into the air stream.

The form and function of the chimneys at KAUST work together. Architecturally, the overall transparency of the elements, coupled with the diagrid structural members, creates a visual landmark for the campus. As a demonstration of passive design, they show that the environment can be harnessed to enhance human comfort. Mechanical systems are not required.

Duncan A. Phillips, Ph. D., P. Eng., is a Senior Specialist with Rowan Williams Davies & Irwin (RWDI), consulting engineers, scientists, experts in environment & building physics, Guelph, Ont. www.rwdi.com