Review and comparative assessment of engineering correlations for the fire-induced air inflow rate through a compartment opening

Abstract

In a building fire, the initial fire compartment is rarely airtight, since it is quite common for some opening (e.g. door, window) to be present, either by chance (e.g. a window left open) or due to the failure of the glazing in a high temperature environment. The opening provides essentially a physical connection of the fire compartment to either the ambient environment or an adjacent room. The developing fire-induced buoyant flow is significantly affected by the presence of an opening, since an open vent provides a means for ambient air to enter the compartment (air inflow through the bottom section of the opening) and for gaseous combustion products, or even flames, to exit the compartment (plume outflow through the top section of the opening). In the context of building fire safety design, it is commonly required to estimate the characteristics of the fire-induced opening flow (or “vent flow”), since the ambient air inflow rate is the main parameter affecting fire ventilation conditions and can thus determine whether over- or under-ventilated fire conditions will prevail. As a result, the air inflow rate through an opening has a significant impact in several important fire development characteristics, such as fire spreading rate, compartment temperature, flashover or backdraft hazards etc.There are several works available in the open literature that focus on developing appropriate fire engineering correlations to accurately estimate the fire-induced opening air inflow rate. These correlations are commonly used in simplified pre-flashover fire simulation tools (e.g. two-zone models). However, despite the importance of this parameter, a general consensus has yet to be achieved. The majority of the available correlations are semi-empirical in nature, where a general analytic opening flow model is combined with experimental results obtained in relevant fire tests. In addition, there are certain empirical parameters appearing in various correlations (e.g. flow discharge coefficient) or other parameters that require prior experimental observations, which may not always be available (e.g. thermal discontinuity height). Motivated by this observation, this work is aimed at evaluating a broad range of available fire engineering correlations for the estimation of the fire-induced air inflow rate through an opening; the correlations are validated and comparatively assessed by comparing the respective predictions with experimental data obtained in eight large-scale fire tests, all performed in fire compartments with a single door-type opening.Nine different correlations available in the open literature are evaluated in this work; the original correlations are cast in a common format, using identical symbols to facilitate comparison. Aiming to evaluate the various correlations’ performance, measurements of air mass flow rate through the opening, obtained in eight different large-scale fire tests, are used; the employed fire tests span a broad range of operational conditions, such as fire heat release rate (32 - 320 kW), burner position (back, centre, front), burner elevation (0 - 0.3 m) and ventilation factor (1.83 - 2.17 m5/2).In general, the investigated correlations are found to exhibit disparate levels of prediction accuracy. The majority of correlations over-predict the measured air inflow rate values, therefore yielding “conservative” results, appropriate for fire engineering design purposes. The correlation of Tewarson (1984) is found to provide the most accurate results (average error: 12.9%). Predictions are further analysed to determine the main physical parameters affecting the prediction accuracy; changing the burner location and elevation is found to significantly affect the quality of the obtained results.