Probabilistic model PROMO for evaluation of air change rate distribution Krystyna Pietrzyk 1 and Carl-Eric Hagentoft 2 1 Swedish Testing and Research Institute, Department of Energy Technology, Building Physics, Borås, SE-50115, Sweden 2 Chalmers University of Technology, Department of Building Technology, Building Physics, Gothenburg, SE-41296, Sweden ABSTRACT: The probabilistic model (PROMO) applied to the problem of air infiltration in low-rise buildings is presented. The PROMO model allows the estimation of the effect of variations of climatic conditions on air exchange in a building. In PROMO, experimental data are used in order to evaluate the parameters and types of the distributions of temperature, wind speed, and wind direction. Those distributions are employed to evaluate the distributions of air change rate caused by air infiltration and can be used to estimate probability of inadequate ventilation. FORM (First-Order Reliability Method) technique is used for probabilistic approximations. For validation of the PROMO model, the probability density functions of the air change rate for the test house situated near Gothenburg are estimated from the results of full-scale measurements of pressure differences across the envelope and compared to those calculated by PROMO from the climatic parameters measured at the site. The agreement of the results obtained in these two ways is very good. KEYWORDS: air infiltration, air change rate, probabilistic model, reliability, FORM INTRODUCTION The main subject of the paper is a study of climate-induced air infiltration in low-rise buildings. The probabilistic model PROMO, allowing the estimation of probability density functions of air change rate has been developed and validated with the help of experimental data. It can be used to evaluate the contribution of air infiltration to ventilation and energy balance and the influence of that contribution on the reliability of air exchange or reliability of energy performance of a building. In the context of air exchange, air infiltration constitutes a complement to the design of ventilation. This complement is often disregarded, or considered in a very simplified manner. The tendency of relying to larger degree on natural ventilation calls for better modelling techniques for air infiltration. An attempt to create such a modelling technique is presented. A modelling approach proposed seems very well suited to handling the natural driving forces governing the rate of air infiltration. It also gives a possibility to quantify the probability of insufficient or excessive air exchange (Pietrzyk, 2000). Physical relationships for air infiltration and ventilation in buildings, which form the background to the probabilistic modelling can be found in the AIVC publications (for ex. in Liddament, 1986), or in (Etheridge & Sandberg, 1996). FORM technique is used for the probabilistic approximations of air change rate (Pietrzyk & Hagentoft, b, Pietrzyk 2000) and reliability analysis (Pietrzyk, 2000, c). The description of FORM can be found in (Haldar, 2000, Ditlevsen, 1996) and also in (Pietrzyk & Hagentoft, a). PROBABILISTIC MODEL PROMO The probabilistic model PROMO has been created under the following assumptions: a low- rise building with single ventilation and temperature zone, the steady-state conditions of air