International Journal on Engineering Performance-Based Fire Codes, Volume 10, Number 2, p.36-39, 2011 36 OBSERVATION OF SMOKE MOVEMENT PATTERN IN A SCALE TUNNEL MODEL WITH LONGITUDINAL VENTILATION C.Y. Tso and W.K. Chow Research Centre for Fire Engineering, Department of Building Services Engineering The Hong Kong Polytechnic University, Hong Kong, China (Received 7 February 2011; Accepted 9 August 2011) ABSTRACT Smoke control by operating the longitudinal ventilation systems was studied by scale models. The tunnel model was of length 2 m and of semi-circular cross-sectional area 0.0665 m 2 . This model is 1/25 the size of a real tunnel. It was made of acrylic plastics so that smoke movement can be observed from outside. A small propanol pool fire of 0.097 kW was placed inside as the heat source with burning pellets to generate smoke. Longitudinal ventilation design was studied by driving air by a fan placed at one end. Different ventilation rates were set by using a transformer to control the power input to the fan motor. Smoke movement pattern in a tunnel fire was recorded. 1. INTRODUCTION In the recent past long railways and vehicular tunnels have been constructed in Mainland China, Hong Kong and Taiwan, as a crucial part in the development of highway systems in hilly areas. Similar projects and developments are still continuing today and are likely to occur in the future. [1,2]. In such transport systems, the passenger density is generally observed to be high and fire safety strategy should be worked out with care and foresight. In particular, attention should be paid to heavy goods vehicles (HGV) in vehicular tunnels as they may cause many problems. The heat release rates in burning HGV have been measured to be 200 MW at maximum [3,4]. Therefore, the fire temperature estimated in some tunnel fires burning HGV would be up to 1365 o C, far exceeding the heat-withstanding limit of 250 o C used in standard tunnel design [1,5,6]. Longitudinal ventilation systems are now a commonly specified installation requirement [7,8] in many developing cities with crowded buildings and dense population. In a fire occurrence, big jet fans installed on side walls would drive air towards the fire source and propel the smoke to one end. The other opposite end of the tunnel would be kept free from smoke [9,10]. The people inside would then be able to proceed in upwind directions by walking against the air movement. The system is particularly suitable for tunnels with a small cross-sectional area where transverse ventilation is difficult to install, and has been demonstrated to be effective on many occasions including the 2007 tunnel fire incident in Hong Kong [11]. Before determining the key design parameters, smoke movement patterns in tunnels with longitudinal ventilation systems should be assessed and studied carefully, so as to obtain data in order to work out the appropriate air speed, the flow rates, fan pressure, fan power and operating scheme. As discussed, there are significant challenges on the application of fire models, particularly on Computational Fluid Dynamics, while applying performance-based design for fire safety provisions [12]. There have been more requests by the Authority to conduct full-scale burning tests in order to provide justification of design details. It is obvious that physical experiments would be exceedingly expensive, and repetitions in an effort to reproduce results could often not be attempted. Consequently, scale modeling studies are more widely used [13,14]. Scale modeling technique would be demonstrated as appropriate and effective [8,15-17] in this paper. A scale tunnel model of length 2 m and semi-circular cross-sectional area 0.0665 m 2 has been used in the experiments. It was created from acrylic plastics such that smoke movement may be observed from its exterior. A small propanol pool fire of 0.097 kW was placed inside as the heat source with burning pellets as smoke generators. The effect of operating longitudinal ventilation on smoke movement was the study objective. Air flow was provided by a fan placed at one end of the tunnel. Varying ventilation rates were achieved by using a transformer to control the input to the fan motor.