adverse wind and conditions [Marchant, 1984 & Morgan et. al., 1981], and even if functioning satisfactorily, would generally require vents located on different external walls. Thirdly, the performance of natural vents is influenced also by the building stack effect, which may be particularly significant on the upper or lower most floors for taller buildings where there may be a strong tendency for either inflow or outflow from all natural vents on that floor.
Other experiments [e.g. Morgan et. al., 1987] have shown that smoke clearance in corridors by natural dilution/dispersal is not a reliable in terms of maintaining conditions suitable for means of escape, and the performance would be subject strongly to the external environmental conditions. Here smoke was allowed to enter the experimental corridor through door cracks, and a ‘visibility distance’ of 10 m defined conditions acceptable for means of escape. It should be noted that acceptable 'visibility distances' can only be defined in an approximate sense. However, the choice of 10 m can be judged to be a reasonable one for a conservative (safe) analysis, and is recommended by CIBSE [CIBSE, 2003], and is being proposed for inclusion in BSI PD 7974-Part 6 (Evacuation). A good summary on the published work on acceptable visibility distances is provided in [Jin, 2002], where the range of reported distances proposed by various authors is from under 2 m to 20 m. However, the shorter distances correspond generally to conditions where the occupants are familiar with their surroundings or where evacuation is in the context of a 'dash to safety' [e.g. Rasbash, 1967].
A scheme to achieve dilution through the use mechanical extraction, which would provide a designed flow rate and be less susceptible to adverse wind and stack effects, has been proposed [Marshall, 1985], but does not appear to have found favour in practice. A relationship between required mechanically induced flow rate and size of door crack (from the fire compartment to the corridor) to achieve a required smoke density was derived.
Efficient dilution will generally be expected on external corridors and lobbies. While still referenced in various codes and standards, this approach is nowadays not commonplace, in particular where the climatic conditions may be undesirable. External corridors and lobbies are not currently being considered for modelling in the current project.
Smoke dilution using a building HVAC system is unlikely to be effective since the ventilation flow rates will be too low to provide much benefit in conditions [Klote, 1988]. The HVAC system may still have a role to play in zoned smoke control, see below, where its function is to create pressure differentials to prevent smoke spreading away from the storey of fire origin.
A lobby or vestibule is used in some instances, and this acts as an extra line of defence for a protected stairwell or may provide an area of refuge. Vestibules may have no ventilation, or in some instance may be pressurised (on their own or together with the stairwell). An alternative is to vent the vestibule with fresh air, and exhaust at an equal rate [NFPA, 1992], providing in effect a mechanical dilution scheme within the vestibule.
© Building Research Establishment Ltd 2005