Air flow in buildings is a complex flow and pressure distribution problem that makes quantification difficult. However, certain parameters have recently become easy to quantify – specifically the air pressure relationships within buildings. The measured building air pressure field can be used with network analysis to solve the building flow and leakage regime creating an analytical macro model of the building flow and leakage regime. The response of the analytical model can be further tuned by perturbing both the building air pressure field and the analytical model. Building analysis typically focuses on flows and requires that all flow paths into and out of a control volume be defined. The flow path resistances need to be characterized. Determining all air flow paths and determining the flow path resistances directly is difficult. As such, estimates of these flow path resistances are commonly used. These estimates are based on limited field data and laboratory measurements. The literature provides some component values that vary by orders of magnitude and their application is often unable to predict building flow fields (ASHRAE, 1997). Standard building analysis develops the building pressure field from the flow field. This paper argues that developing the flow field from the building pressure field is more powerful. Determining the characteristics of the building pressure field directly is considerably easier than determining flow path resistances. It allows closing of the gap between the mathematical sophistication of available multi-cell air flow models and the necessary input information defining the building boundary conditions. This approach allows the pressure response of the building to be used to ‘‘tune’’ the models extending the range of their applicability and accuracy.

Control of airflow is essential to several important performance aspects of the building system. Air carries moisture which impacts a material’s longterm performance (serviceability) and structural integrity (durability), behavior in fire (smoke spread), indoor air quality (distribution of pollutants and microbial reservoirs) and thermal energy. Typical case studies are presented to illustrate how each of the above characteristics is affected when unintended airflow occurs as a result of poor construction. In some cases, there was simply a lack of understanding of the consequences of ignoring potential leakage paths and the interaction of the mechanical conditioning systems with the building structure. Rehabilitation of a troubled building requires that these interactions be understood. In general, the approach to developing that understanding is not involved.

This paper reviews work that has led to the development of a number of multi-zonal airflow models (network models). At present, the network analysis and perturbation methods cannot be used to solve the interstitial flow, pressure and resistance regimes. Network analysis and perturbation may suggest that such flows exist, but the complexity and workmanship dependence of the interstitial flow, pressure and resistance regime requires direct measurement. In other words, at present, the boundary conditions of the interstitial regime can be defined analytically using traditional methods, but the pressures and flows within the interstitial spaces cannot be predicted with certainty using analytical means.Difficulties in obtaining the detailed information on mechanical systems and the leakage areas of building assemblies make the traditional approach of measuring airflows and constructing models using leakage areas impractical for diagnostic purposes. Additionally, a pressure difference across an assembly alone, where interstitial pressures are not considered, is not enough to describe performance of the building envelope. The interstitial air pressures are usually small and until recently have been beyond measurement.Research on building envelope durability and indoor air quality has shown the significance of these small, but persisting interstitial air pressure fields. To enhance the capability of network models, a relational model was developed. This approach permits the measured building air pressure field to be used with the network analysis. Furthermore, the response of the analytical model is calibrated by comparing the effect of a specific perturbation on both the building air pressure field and the analytical model. This paper is written in two parts, the first reviewing issues in prediction of airflow in the buildings and building envelope, the second analyzing the actual data of measured pressure response of buildings. sake of clarity. When some relations are deemed less important than the others,

An extensive investigation of moisture problems in wood frame houses revealed that one of the most common problems was the formation of mould and mildew on inside wall surfaces of exterior corners (Van Poorten, 1983). The majority of the problem houses were of recent construction (around 1975), electrically heated yet "naturally ventilated." Clearly, inadequate ventilation was one of the contributing factors.It was not possible to explain this occurrence solely in terms of thermal bridging, two-dimensional heat flow, hindered air circulation or lack of radiant heat gain. It was then suggested that additional cooling of wall surfaces resulted from defects in the sheathing which allowed wind to cool the wall construction in corners where a rapid change in wind pressure occurred. It was also suggested that in future housing, such problems could be largely avoided by providing airtight sheathing at building corners by moving the air barrier, customarily located on the room-side of the wall insulation, to the weather side where it would perform two functions (Timusk, 1983).A full-scale comer of a wood frame wall was built and evaluated for wind cooling effects in the laboratory. According to prediction, significant additional cooling did take place when defects in wall sheathing did exist and when the corner was exposed to a typical wind pressure gradient.The effectiveness of an externally-located air barrier was further demonstrated in a prototype house. Its installation and inspection was facilitated by locating it where it is not penetrated by joists, partition walls and electrical wires.

Concern has been expressed that recent increases in thermal insulation levels of residential basements in areas of seasonal frost may increase the risk of damage to foundations due to frost heave and adfreezing of the soil to the foundation wall. An hypothesis is presented which states that the adfreezing bond strength is affected by the direction of soil moisture movement in response to thermal gradients. This hypothesis may be used to explain why there have been no reported adfreezing problems associated with heated, insulated residential basements. Key words: adfreezing, basement, frost heave, insulation, moisture.