Evaluation of fire engineering design correlations for Externally Venting Flames using a medium-scale compartment-facade fire experiment

In a fully developed under-ventilated compartment fire, flames may spill out of external openings (e.g. windows); Externally Venting Flames (EVF) pose a significant risk of fire spreading to adjacent floors or buildings. The main scope of this work is to comparatively assess a range of fire engineering design correlations used to describe the external dimensions of the EVF envelope. The predictive accuracy of each correlation is evaluated through comparison with experimental data obtained in a medium-scale compartment-façade fire arrangement, using a variety of fire load levels. A series of medium-scale fire compartment experiments is performed, employing a ¼ scale model of the ISO 9705 room equipped with an extended façade. An extensive sensor network is used aiming to monitor the dynamic behaviour of a broad range of important physical parameters (e.g. gas and surface temperatures, heat flux, fuel mass loss, gas concentrations). A dedicated image processing tool is developed to allow estimation of the EVF envelope main dimensions (e.g. height, width, projection). An “expendable” fuel source (n-hexane liquid pool fire) is utilized to effectively simulate realistic building fire conditions. Digital camera imaging is used to determine the main geometrical characteristics of the EVF envelope. Comparison of fire engineering design correlation predictions with experimental data reveals that correlations for the estimation of EVF height err on the safe side in under-ventilated fire conditions; increasing the heat release rate results in more conservative EVF height and projection predictions. It is shown that EVF projection and width strongly depend on both excess heat release rate and height. In addition, the necessity to derive appropriate criteria for the identification of the EVF projection is demonstrated. The obtained extensive set of experimental data can be used to validate CFD models or evaluate the accuracy of other available fire design correlations.