Optimisation of evacuation process in tall buildings

Abstract

The growing demand for residential and commercial space in urban areas has resulted in increased number of constructed tall buildings around the world. Regarding the safety, tall buildings were subject of many disaster events in the last two decades, where the disaster on 9/11 highlighted the importance of providing robust evacuation. The main reason for that is the fact that a single fire may cause considerable number of casualties in such buildings. Therefore, relocation strategy represents a major role in the safety design of any tall building. The first step to attain a safe and quick egress is understanding people behaviour in different evacuation scenarios, while the next step is the integration of it with building egress design and finding optimal evacuation strategy for all people. In this way, the optimal evacuation strategy would result in all people being evacuated from a tall building in the shortest possible time considering their actions.

In spite of the increase in the availability of advance evacuation models, the main drawback to model evacuation process close to a real one is not the simulation of people’s movements, but the current understanding of decision-making steps that people take in specific evacuation situations. The same applies to tall buildings being unable to model route choice of people between stairs and lifts with low uncertainty in order to achieve an effective evacuation. Therefore, the route choice was investigated in the thesis in an effort to understand people behavioural actions for different evacuation scenarios, and considering this afterwards for finding the optimal evacuation strategy for people. In this regard, it was comprehended that directing people to use a particular egress component (e.g., evacuation lifts) leads to most predictable evacuation procedure.

After grasping the knowledge related to the route choice of people, modelling of other people behavioural actions and their movements was undertaken by establishing an evacuation model which calculates total evacuation time for tall buildings. Thus, the mathematical model successfully integrated people performance, the stochastic nature of people behaviour at different egress components and capacities of egress components to represent a realistic evacuation process in tall buildings. In order to find the optimal evacuation strategy for all people the model is solved using branch and bound algorithm. The relative comparison of the results in several aspects showed that the research problem should be considered as multi-objective minimising the total evacuation time, the number of people waiting to be released on their floors and the number of people waiting at the refuge floors. The results indicated that 5400 people would be evacuated in approximately 1 hour and 15 minutes; less time compared to another model.

In the end, the conclusions drawn from research conducted in the thesis along with the implications and future research were presented. Thus, regarding the decision making of people, the importance of several parameters such as density of people on stairs, vertical position of refuge floors, number of people in front of lifts, navigation strategy and people’s familiarity with tall buildings is concluded. In terms of the optimal evacuation strategy, it was found that people should take at the beginning and the end the evacuation lifts or switch between stairs to achieve a balanced evacuation strategy, which would lead to faster evacuation of all people from a tall building.