Building evacuation: principles for the analysis of basic structures through dynamic flow networks

Purpose: The main purpose of this paper is to perform an analysis of the factors determining the architectural configuration of buildings for the mobility of people, using dynamic flow networks and considering group formation in the evacuation process. Design/methodology/approach: For a long time it has been considered that once an evacuation begins, movement on the evacuation route mainly obeys mechanical factors; people occupy the free spaces which lead to evacuation more or less automatically. However, recent research has emphasized the need to consider people’s behavior; one of the aspects considered in this work is group formation, with its significant influence on the evacuation process. In positions of convergence and their branches things become considerably more complicated; as well as occupants' behavioral aspects other relevant factors such as the geometry of the premises are critical in this process. Authors propose models, in which nodes are strategically placed, besides taking into consideration aspects of behavior. Several cases are analyzed. Findings: The solution proposed in this paper is to analyze the problem through dynamic flow networks, using a macroscopic model in a deterministic environment in which the evolution of the quantities characterizing the problem at regular intervals is represented, obtaining a reasonably accurate and reliable understanding of the development of the evacuation. Originality/value: A precise model of evacuation routes, convergence points and branches which includes a consideration of occupants' behavior is obtained using stochastic models with microscopic analysis, in which people's behavior is considered individually, this solution is complex, difficult to apply with many occupants and in large enclosures, and also, this way does not lead to optimal solutions.