An experimental study on the burning rates of interacting fires in tunnels Sina Shafee, Ahmet Yozgatligil * Mechanical Engineering Department, Middle East Technical University, Universiteler Mahallesi, Dumlupinar Bulvari No:1, Ankara 06800, Turkey ARTICLE INFO Keywords: Burning rate Heat release rate Interacting fires Pool fires Reduced scale tunnel ABSTRACT Multiple fires may occur in close proximity in process industries, power generation and fuel storage facilities and confinement conditions such as tunnels, which can lead to a considerable alteration in fire characteristics and safety design. The topic is of significant importance to the fire safety research because there is little work in the literature that investigates the case of interacting fires, which have a destructive potential. In this work, we study the effects of an adjacent fire source on the burning rate and heat release rate characteristics of tunnel fires. Square ethanol pools of 10 and 15 cm in size and 0.22–1 cm in depth were used as fire sources in a reduced scale tunnel model. Ventilation to the tunnel was varied between 0 and 1.5 m/s. Pool fires were configured in single and dual pool orientations. Variations in the pool fire burning rates were discussed as being functions of pool size and depth, and a result of the interaction with the secondary fire. The maximum burning rate enhancement factor, defined as the ratio of the parameter for interacting fires to non-interacting ones, was shown to be 2.3. This was due to the enhancing effect of the secondary fire on the heat feedback to the fuel, and the increased combustion mass transfer. Tests with relatively larger pool sizes burned faster, with an advanced onset of the transition to a bulk boiling phase, which was attributed to the controlling heat feedback mechanism associated with the pool size. 1. Introduction Research on tunnel fire safety has gained more importance owing to the rise in serious fire accidents, possibly resulting from the increased construction and utilization of road and railroad tunnels, which can now be many kilometers in length. The literature indicates that the source of tunnel fires is generally burning carriers, heavy good vehicles (HGVs) and pool/spill fires following the leakage of combustible materials from tankers [1–3]. Consequently, pool fires are of special interest to the fire research community in general and tunnel fire safety in particular. Pool fires are also recognized as a source of industrial fires [4,5]. There is an abundance of experimental and numerical research literature on pool and tunnel fires. The critical ventilation velocity (defined as the minimum ventilation velocity required for the prevention of smoke movement in an upstream direction), smoke flow backlayering, tunnel temperature dis- tribution, fire Heat Release Rate (HRR) and burning rate have been studied using real scale or reduced scale tunnel models. These works have contributed to the current knowledge on tunnel fire dynamics and the development of fire safety standards [6–20]. Among the above fac- tors, the burning rate and HRR of a fire are considered to be the most prominent factors in considering pertinent fire hazards [21,22]. Research on pool fires was pioneered by Blinov and Khudiakov [23] and elsewhere by Rasbash [24]. In more recent works, pool fire com- bustion has been characterized according to fuel type, pool size, the dominant heat transfer regime and flame attributes [25–28]. An infor- mative summary of relevant studies on the matter was given by Ditch [29] and elsewhere by Hu [30]. Chen et al. investigated the burning rates and temperature variations of 0.2 m circular n-heptane pool fires under quiescent conditions [31]. The results indicated that there are two-stages in the increase of the burning rates of a fire, in which the second peak corresponds to fuel bulk boiling. The effect of vessel materials and free- board heights on the burning rates of small ethanol pool fires was investigated by Dlugogorcki and Wilson [32]. Glass, copper, and steel were used as vessel materials. They concluded that the effect of the lip height could be a controversial aspect of the study of pool fires. Shafee et al. investigated variations in the Mass Loss Rate (MLR) of n- heptane pool fires in a 1/13 scale model of an underground tunnel [33]. Square and rectangular pans were used for the pool fire. The critical ventilation velocity was shown to be achieved at around 1 m/s in the model, which corresponded to 3.6 m/s in the real scale tunnel. H.Y. Wang simulated octane pool fires in a ventilated real scale tunnel using Fire Dynamics Simulator (FDS) numerical code [34]. Large Eddy Simulation (LES) was used to model turbulence in this work. FDS incorporates a finite difference solver, which is commonly used in fire simulations by * Corresponding author. E-mail addresses: sina@metu.edu.tr (S. Shafee), ahmety@metu.edu.tr (A. Yozgatligil). Contents lists available at ScienceDirect Fire Safety Journal journal homepage: www.elsevier.com/locate/firesaf https://doi.org/10.1016/j.firesaf.2018.01.004 Received 30 July 2017; Received in revised form 21 December 2017; Accepted 9 January 2018 0379-7112/© 2018 Elsevier Ltd. All rights reserved. Fire Safety Journal 96 (2018) 115–123