(624k) Simulation Based Optimization of Ventilation Strategies for Subway Tunnel Fires

As the population density in large cities of the world is getting higher and the traffic volume is  increasing continuously, the roads cannot accommodate such traffic increase any more so that construction of new subways or extension work of existing subways are constantly going on. With this higher subway traffic, the subway stations are needed to be better prepared for fire. But the preparation for fire emergency is lacking in subway tunnels at the present. In Korea, the risk of subway fire has been widely recognized since the arson fire took place at the Daegu Jungangro subway station in 2003, resulting in 340 casualties. As a result of this accident, people recognized problems related to fire equipment, the emergency management agency and ventilation operation. Since most subway stations are easily exposed against the possibility of danger such as, it is necessary to set up substantial prevention measures against any possible fire or terror attack.

Many studies on the prediction of fire damage have been performed at subway platform. However, the research of the simulation on subway car fire have lacked in tunnels. Unlike other building fire on the ground, the combustion is unstable because fire on underground tunnel may be limited to get O₂ from the outside, therefore mass toxic gas would be released. Considering most death from suffocation by the fumes, passengers in the tunnel are more exposed to the hazard from toxic gas. The better adaptive ventilation strategies have been needed to help people survive in dangerous situations.

This study is focused on the heat and toxic gas characteristics in an underground subway tunnel where burning train is stopping in, by using Fire Dynamics Simulator (FDS). The tunnel, which is a part of Seoul Metro Line 1, has 6 sessions for ventilation. Due to structural limitation, only four fans were installed for forced ventilation and other two fans operate by natural ventilation. Usually when a fire occurs in a tunnel, they control forced ventilation facilities to set up the fire escape route for people: fans at the side of escape route are operated on the supply mode. Fans at the other side are operated on the exhaust mode for the removal of smoke. But this tunnel cannot use ordinary ventilation strategy because of one-sided ventilation facilities. Therefore, this study carries out a variety of simulations to find out the best ventilation control strategy according to the fire situations.   

To simulate the combustion of real-size subway tunnel model in detail for the prediction of combustion gas behavior, we should calculate large amount of numerical analysis due to model of the complex structure and variety of oxidation, heat transfer as turbulent flow. For this reason, we used parallelized FDS, which based on Message Passing Interface (MPI), running on a high performance computer cluster. The objective of simulation is comparing each simulation result about air flow, temperature, visibility, toxic gas according to the selected method of fan operations. Then the best ventilation method for each fire situation is to be decided by the results. The result of this study will contribute to the improved saving of passengers in case of fire accidents in subway stations with optimized ventilation system design and operations.