October - December 1997 Issue


Reducing Indoor Air Pollution and Saving Energy, Too

To reduce energy use, owners and designers of buildings seal up leaks, maximize insulation, and install ventilation systems meant to provide occupants with clean, conditioned air. However, those ventilation systems don't always work as planned. In some cases, airborne contaminants and heat generated indoors by building materials and equipment--and even normal human "effluents" such as carbon dioxide--become trapped. The impact on the comfort, health, and productivity of occupants can be serious. Numerous "sick buildings" have required evacuation and repair, even state-of-the-art new construction such as the headquarters of the US Environmental Protection Agency.

The traditional approach to ventilation does pose potential problems. The goal in conventional "mixed-flow" ventilation is to mix up all the air in the room, producing an even temperature throughout the space and diluting any pollutants. Typically, ventilation air is injected at high velocity at or near the ceiling. If mixing is not sufficient, stagnant areas can occur in the "occupied zone" nearer the floor. And while mixing does dilute pollutants, it also sweeps them away from their sources and distributes them throughout the room, exposing all occupants. In addition, mixed-flow ventilation requires large quantities of air that must be heated or cooled by energy-consuming devices.

During the past three years, Professors Qingyan Chen and Leon R. Glicksman and their coworkers in the Building Technology Program and the Energy Laboratory have been studying a radically different approach to ventilation. The approach, called displacement ventilation, seeks to prevent mixing rather than cause it. Cool, clean air enters a room at a low level at low velocity, flooding the floor much as water would. Heat and contaminants emanating from people and equipment tend to occur in isolated "plumes" hovering above individual sources. In the absence of mixing, those plumes of warm, contaminated air rise naturally toward the ceiling, where they are vented. As they rise, the plumes entrain the surrounding air, lifting the cool ventilation air to the breathing level of the occupants. Because the clean air is supplied directly to the lower, occupied area rather than mixed throughout the whole room, less air is required. Moreover, the entire room need not be as cold to provide a comfortable air-conditioning effect. The result is reduced energy use. (Most commercial buildings require cooling even in winter. However, conventional heating systems can be used simultaneously as needed.)

Displacement ventilation originated in Scandinavia and is now used widely in Europe. However, in the United States it has been installed at only a few experimental sites. Today's US building regulations are based on mixed-flow ventilation and require more air than is typically delivered by displacement ventilation systems. Moreover, there is little pressure to change those regulations, largely because of serious concerns about comfort. With displacement ventilation, fresh air flows directly into the occupied zone. If the air is too cold, moving too fast, or fluctuating in velocity, people get cold feet. Also, the air at an occupant's feet is cooler than that at his or her head. If the difference is too large, the occupant is uncomfortable. Discomfort may be more of a problem in US buildings than in European buildings because the former generally have higher cooling loads than the latter.

Recognizing the potential benefits of displacement ventilation, the MIT researchers are working to examine and ensure its effectiveness using two coordinated approaches: they are developing new computer models that can simulate conditions in a room with displacement ventilation, and they are gathering data in a full-size experimental room.

Designing a displacement ventilation system for a given space involves many decisions. For example, what type and how many diffusers should be used to distribute the air, and where should they be located? What should the temperature, velocity, and humidity of the input air be? The optimal specifications depend on the size and configuration of the room; the number, location, and types of heat and pollution sources (people, equipment, lights, furniture, and so on); and many other factors. An appropriate computer model would enable a designer to analyze the ventilation needs of a specific space and to define a system that would ensure good air quality, comfort, and energy efficiency.

To identify a suitable model, Professor Chen and his colleagues examined standard "computational fluid dynamics" (CFD) models that describe turbulent flows like those that occur in rooms with displacement ventilation systems. However, the standard models proved inadequate. In general, they are designed for smaller-scale applications such as airflows in fans or over aircraft wings. In addition, they cannot accurately simulate the behavior of buoyant plumes--a behavior critical to the flows of air, heat, and pollutants in a room with displacement ventilation.

Professor Chen's group has now developed a CFD model that incorporates a new, simplified technique to simulate the turbulent airflows in buoyant plumes. For a given situation and ventilation system, the model predicts distributions of air velocity, temperature, and contaminant concentrations throughout a room. The new model can be put on a personal computer and in less than an hour's time produce a good first prediction of air circulation patterns--a marked contrast to traditional CFD models, which require large-capacity computers and extensive computing time.

To validate the model under a variety of well-defined conditions, Professors Chen and Glicksman have built a special facility at MIT--a full-size room with a controllable environment. The room is about 5 m long, 3.5 m wide, and 2.5 m high. It is built inside another room, enabling the researchers to select and maintain a constant wall temperature, change the building materials, and add windows as needed to simulate different external situations. The researchers can use a variety of diffusers with different designs and can control the properties of the ventilation air. Instruments located throughout the room measure temperature, humidity, velocity, and velocity fluctuations. To observe patterns of air circulation directly, the researchers create a plane of light cutting across the room by shining a powerful slide projector through a narrow slit or by using a rapidly oscillating laser beam. They then inject puffs of smoke into the room and capture images of the smoke's movement using video recording equipment placed perpendicular to the plane of light. Finally, they can inject a tracer gas at various locations to mimic a pollutant being given off by a person or machine. The compositions of samples collected at various locations show how the tracer gas disperses throughout the room.

The space outside the experimental room houses the controls--a variety of computer displays that allow the user to select conditions and observe effects in on-screen diagrams of the room. The controls are designed both for ease of operation and for use as a teaching facility.

To simulate specific situations, the researchers "furnish" the experimental room in various ways. For example, the figure below shows the room as a small two-person office. Sitting at tables are two "occupants"--square models with heat sources inside. In front of them are computers and monitors that generate heat. Typical office lights are overhead, and one wall contains an external window. In another setup, the room represents a section of a large office that contains two occupants sitting on opposite sides of a partitioned cubicle. And in another, the room is a quarter of a large classroom. Six "students" sit at tables facing the front of the room. In each case, the setup is altered to reflect summer and winter conditions, with the latter including a heat source and a cold exterior window.

In a series of tests, the researchers operated the experimental room and the new CFD model under similar conditions. In most cases, the model's predictions match the experimental results well. The calculated and measured temperatures and velocities throughout the room agree closely. Predictions of velocity fluctuations are not as good, in part because velocities are so low that measuring fluctuations is difficult. Predictions of the tracer gas concentrations are reasonably accurate, though large discrepancies occur at some locations. Smoke-visualization tests show the incoming air falling gently to the floor and then slowly drifting upward. Plumes of smoke appear above people and equipment, swirling and rising like smoke coming out of a chimney. In general, the studies confirm that a properly designed displacement ventilation system can indeed provide both clean air and comfort.

The researchers are continuing to use the model to simulate different situations. They are now performing tests in a privately owned facility that replicates a large industrial workshop. They are also using the model and parallel experiments to measure the airflow in fume hoods, including those used in biology laboratories at MIT. Results show that the hoods use airflows higher than required to ensure removal of contaminants--a significant loss of energy. In addition, they are looking into health-care applications. If displacement ventilation systems were used in hospitals and shelters, germs from patients with communicable diseases would rise to the ceiling rather than being spread to other patients.

Meanwhile, they are expanding the capabilities of the model. They are adding equations that describe chemical interactions among airborne contaminants. Thus modified, the model will be able to predict what reaction products will form and where substantial buildup may occur. And they have developed computer codes that calculate energy use associated with a given simulation. Already they are analyzing the energy use and costs associated with designing, installing, and operating displacement ventilation systems versus conventional mixed-flow ventilation systems. Ultimately, they plan to develop a set of design guidelines that they hope will reduce the amount of airflow required under today's building regulations--a change that could produce substantial energy savings nationwide.

Qingyan Chen is assistant professor of building technology. Leon R. Glicksman is professor of building technology and director of MIT's Building Technology Program. This research was supported by the American Society of Heating, Refrigerating, and Air-Conditioning Engineers; Halton, Inc.; Trox GmbH; and the National Science Foundation. Further information can be found in references.