The LESO building is a three storey, medium-sized office building on the campus of the Swiss Institute of Technology in Lausanne. In this building component leakages have been carefully determined followed by extensive measurements of the boundary conditions as well as the air flows. This paper first gives some basic concepts of the evaluation and the sensitivity analysis. Then, the measured data are compared with results from simulations performed with the COMIS multizone air flow program.

One of the options to increase the energy efficiency of buildings in the cooling season, is to extract heat from the building envelope during the night by natural or forced ventilation. The exploitation of this technique by architects and designers requires the development of guide lines and a predesign tool showing how the potential cooling power depends on the influence of opening sizes and positions and on the interaction with the thermal mass.

The paper describes measurements made on large doors - 10 to 20 m2 in 2 buildings in Narvik. The air change was measured with the tracer gas (SFg). The method of constant concentration or decaying concentration of the tracer gas was used. The dosing, measuring and calculation of the air change was made with a Briiel & Kjaer gas analyser type 1302 and computer. Use of the decaying method was best with short opening times. The opening of the door in 5 to 7 minutes gives an air exchange of 500 m³ to 1300 m³ or an air change from 0.2 to 1.0.

Air infiltration continues to play a major role in the ventilation of houses, despite modern trends to increased airtightness of the building envelope. In colder climates, stack effect is the principal driving force for this natural air exchange. The neutral pressure level divides the envelope areas subjected to stack effect pressures driving infiltration from those subjected to pressures driving exfiltration. The neutral pressure level is therefore important to our understanding of stack driven air exchange and our ability to model it.

A new model concept has been developed to model the three dimensional energy and mass transfer in an imperfectly mixed fluid. The model permits to predict the dynamic behaviour of the volumetric concentration of heat flow, mass flow and fluid flow. A laboratory test installation has been built to analyse the model capabilities to predict the dynamic behaviour of the air flow pattern within a ventilated space in order to control the energy and mass transfer in the ventilated space.

Preliminary work has indicated that thermography can be used to determine air leakage pathways from or to buildings. Accurate measurements have now been taken using temperature controllable environmental chambers.These results reinforce the potential useof thermography for this application. In conjunction with the physical measurements a simulation has been carried out using computational fluid dynamics.

It has been shown that thermal imaging can give an indication of air flow rates through small cracks. Using a finite difference analysis package it is possible to determine the surface temperature of an air transfer grille when subjected to airflow rates at higher temperatures than the grille surface. This paper will address this technique by presenting the results of the finite difference analysis package for a specific grille.

Multizone models are a common tool for calculating air and contaminant exchange within rooms of a building and between building and outside. Usually a whole room is then modelled by one calculation node with the assumption of homogeneously mixed conditions within this room. In real cases, however, temperature and contaminant concentrations vary in space. The exchange to the neighbouring nodes via the flow paths is then a function of the local values of these variables.

Improvements to ventilation systems for the purpose of saving energy may also affect the provision of good air quality. Measurement of ventilation effectiveness may be used to determine whether or not good fresh air distribution and satisfactory contaminant removal has been achieved in a specific case. However, for such measurements to be useful, it is necessary to establish recommended values of the parameters and to check the reliability of the measurement procedures. This paper is concerned with the second of these problems.

This paper describes a laboratory model for the testing and validation of tracer gas measurement techniques. Previous attempts at experimental validation have often been limited to two zones, or a particular measurement strategy, or a particular range of flows. The model consists of four zones, each of 1m³ internal volume. The zones are connected so that all possible inter-zone flow paths exist. The flow down each path is driven by a pump and monitored by a flow meter. A control panel enables any combination of interzone flows to be set, within the capacity of the pumps.