Dynamic Risk Evaluation For Passenger-carried Fire Load In Metro Stations:From Theoretical Modeling To Application

Mass transit rail/metro is becoming one of the most important transportation modes in our daily lives as the rapid urbanization in many large cities.It has greatly improved the efficiency of our daily travel.Metro stations,particularly those in large cities like Beijing,Hong Kong,or Shanghai,are equipped with massive facilities such as shops,banks,fast food kiosks,and exhibitions,as well as vast population activities.Although most of these facilities are non-combustible and with fire extinguisher equipment nearby,existing statistics/reports show that fire accidents are always un-avoidable,which remains one of the serious threats that metro faced with.A very important yet underlying cause for these fire accidents is the fact that the baggage/luggage associated with the passengers may be combustible and ignited during the crowd transportation(termed as Passenger-carried Fire Load).However,fire risk associated with the passenger-carried fire load has rarely been studied because one of the difficulties underneath this topic is the dynamic nature of passenger-carried fire load.To study such dynamic passenger-carried fire load and evaluate the potential fire risk in metro stations,we need to study in detail the passenger movement pattern in stations.In this dissertation,we aim to establish an effective simulation model to study fire risk of the passenger-carried fire load for metro stations.The simulation output of this model could represent not only the detailed features of pedestrian traffic but also the dynamic variation of fire risks in the simulation scenarios.Firstly,a microscopic simulation model named dynamic risk evaluation(DRE)model is proposed through the combination of pedestrian movement module and fire risk calculation module.Pedestrian movement module mainly deals with the movement of passengers,while fire risk calculation module determines the variation of fire risks induced by passenger-carried fire load.Specifically,two types of pedestrians,passengers with large luggage and those with small/no luggage,are defined and the detailed features and algorithms of them are described(where large luggage occupies a specific moving space while the small luggage does not).The calculation of spatiotemporal fire risk distribution is executed by these two modules with mutual communication.Once pedestrian properties are provided,pedestrian movement module will initialize the pedestrians along with their belongings(passenger-carried fire load)in the simulated area.At the same time,fire risk calculation module will rank the fire risk levels of passenger-carried fire load and generate their influencing area.After that,a modified multi-velocity floor field cellular automaton(FFCA)model in pedestrian movement module would govern passengers’ actual movement and then calculate the moving direction for the next time step.Meanwhile,fire risk calculation module analyzes the interaction of fire risks caused by passenger-carried fire load when these pedestrians gather together or get separated.Besides,empirical data in published literature and field observation are used to validate the model,and parameters in DRE model are analyzed to identify the influences.As passenger-carried fire load is the property of passengers,and the dynamic nature of it is associated with the walking pattern of passengers.Therefore,the reliability of pedestrian movement module determines the credibility of DRE model,and the model validation is focused on the validation of pedestrian movement module.Flow rate at the bottlenecks/exits under different bottleneck/exit widths and fundamental diagram(the flow rate-density relationship)are considered in the quantitative test.At the same time,pedestrian movement trajectories and characteristics are adopted in the qualitative test.Simulation results obtained from DRE model agree well with the empirical data.Influencing area of passenger-carried fire load,time interval of fire risk map,and desired velocity of passengers are found to have significant effects on the variation of fire risk.A larger influencing area results in a smoother fire risk surface and a wider range of the risk regions,while a shorter time interval may provide a more accurate distribution of fire risk but have a larger variance.Faster desired velocity of passengers makes a larger fluctuation of fire risk distribution as passengers move quicker.With these insights into the dynamic passenger-carried fire load,we introduce the fire risk of passenger-carried fire load in crowd management studies.DRE model is used to assess the fire risk of the passenger-carried fire load to analyze spatial setting and crowd movement patterns by focusing on the influence of the initial distribution,the ratio of passengers with high-risk fire load,as well as system scale on the fire risk pattern.Simulation results demonstrate that farther distance to the exits would result in a relatively higher fire risk for a relatively long period.Both the ratio of passengers with high-risk fire load and entrance flow rate show exponential relationships with the proportion of high-risk regions.Larger system sizes would result in more stable outcomes and are better to take proper steps.To keep a safe metro environment,the critical point is to reduce the proportion of high-risk regions in the station,measures like fasten the speed of passengers with high-risk fire load,and separating them with other people can be taken.The findings of this work could serve as a useful tool to find out high-risk regions and provide scientific support for initial spatial design and crowd management of transportation hubs like metro stations as well as the management of facilities.