Variability of Behavioural Parameters in Egress Simulations of Stadiums

Danielle Aucoin – York University

Abstract

Events hosted at stadiums attract large crowds, which requires an in-depth consideration of human behaviour to safely design and manage these structures. There is a growing demand to build new and revitalize current stadiums to a larger capacity to maximize revenue potential and minimize construction costs. However, there is little public data available to develop contemporary stadium guidance in order to optimize stadium performance and maintain an appropriate level of safety. This emphasizes the need for egress studies to provide validation for simulation models, since the reliability of these egress models in performance-based stadium design is dependent on the confidence of the input data.

The research presented in this study describes novel data collection and subsequent egress modeling of an anonymous contemporary Canadian stadium to provide further insights to practitioners regarding the potential variability of behavioural parameters between experimental and simulated trials. In 2018, the case study stadium has had a documented fire while at low capacity. Interrogating safety and the stadium’s egress performance at high capacity, two studies were carried out in which all events experienced crowds of over 20,000 people. High-resolution cameras were strategically placed to capture crowds in the stadium stands and all egress routes. Various cases were observed: one event with all exits available and another with one primary exit closed, presenting a worst-case scenario with converging crowds. Pedestrian facial cues were observed to monitor a benchmark applied to stadium design that suggests people show signs of stress when total egress exceeds eight minutes. Additional data collected includes percentage distributions of people egressed over time and total post-game egress times which all exceeded 17 minutes. A scale stadium model was built to simulate the various cases using the MassMotion pedestrian movement modeling software and compared the pedestrian behaviour in the model to that of the experimental trials. Running various simulations allowed assessing the sensitivity of the design according to parameters such as exit closures and varying demographics, all of which may impact human decision-making and walking speeds during an evacuation. Future work will include the addition of another Canadian stadium to the study, where a custom agent profile will be created in addition with further focus on qualitative behavioural aspects.

Presentation

Resources

Paper Presentation
PDF PDF

Return to the Fire and Evacuation Modeling Technical Conference (FEMTC) 2018 Home Page