Practical methods for improving estimates of natural ventilation rates.

This paper discusses four concepts that have been found useful in improving estimates of ventilation rates in residential buildings. These concepts are improved methods for describing leakage distribution and wind pressures: 1. Separation of large, well defined "local" leakage sites from the background building leakage. 2. Changing surface pressure coefficients to account for the effect of upwind obstacles. 3. Making wind pressures (in terms of pressure coefficient and wind shelter) continuous functions of wind direction. 4.

Measuring subfloor ventilation rates.

This paper reports on ventilation measurements taken beneath a suspended timber floor of a BRE/DoE energy and environment test house. Sulphur hexafluoride was introduced into the subfloor void at a constant rate and the resulting concentration measured. Wind speed, wind direction, and internal, external and subfloor temperatures were also recorded. A range of air brick locations were used for each run which lasted two to three days.

Volume control of fans to reduce the energy demand of ventilation systems.

The fan and the ductnetwork is designed for 100% ventilation rate. Because the fan energy is the main important energy consumption in systems all over the year it is worthwhile to control the systems correctly. By reducing the air volume rate the pressure drop in the ductnetwork drops nearly with the second power.

Comparing predicted and measured passive stack ventilation rates.

BRE have experimental data for the flows found in Passive Stack Ventilation (PSV) ducts from a test house in Garston. These data cover different duct diameters, number of bends and roof terminals, all measured over a variety of weather conditions. In the first part of this paper the data are analyzed to separate temperature and wind effects, and to see how well they fit well to the expected model of duct flow. The second part gives a comparison of the same data with predictions from the single zone ventilation model BREVENT.