PSAM 16 - Session W23 Overview

PSAM 16 Conference Session W23 Overview

Session Chair: Richard John (richardj@usc.edu)

Paper 1 MA211

Lead Author: Marilia Ramos Co-author(s): Mohammad Pishahang , mh.pishahang@risksciences.ucla.edu Enrique Andrés Lopez Droguett, eald@ucla.edu Ali Mosleh, mosleh@ucla.edu Andres Alfredo Ruiz-Tagle Palazuelos , aruiztag@umd.edu

Integrating Human Behavior Modeling into a Probabilistic Wildfire Egress Planning Framework
Wildland Urban Interface (WUI) can be defined as “the zone of transition between unoccupied land and human development.” The communities in these areas are particularly vulnerable to wildfires that start and propagate in wildlands. According to the U.S. Fire Administration, the U.S. has more than 70 thousand communities at risk for WUI fires, and the WUI area grows by approximately 2 million acres per year. Numerous efforts have been undertaken to address the dangers of wildfires, including building more resilient infrastructures, advancing techniques for extinguishing fires and exploring the possibilities of controlled fires. Associated with these efforts is the pressing need to ensure the safe evacuation of communities in WUI once they are threatened by wildfires. Evacuation modeling and planning is a challenging and complex problem. It involves human decisions and actions concerning if, when, and how to evacuate; directly impacting the traffic flow during the evacuation. Furthermore, the available time for a community to evacuate is a dynamic element: it changes according to the fire progression, which, in turn, depends on vegetation, weather, among other factors. The models for traffic and fire progression have advanced considerably in the past years. Human behavior modeling during wildfire evacuations has also received significant attention, leveraging existing studies for other natural hazards, such as hurricanes. However, these are mainly developed as standalone, qualitative approaches rather than integrated into a complete egress framework that accounts for traffic modeling and wildfire progression. This paper discusses the existing methods for modeling human behavior during wildfire evacuation and the requirements for such a model to be integrated into a quantitative, probabilistic evacuation planning tool. It further presents the method adopted for integrating human behavior model into the evacuation planning tool WISE (Wildfire Safe Evacuation). WISE calculates the probability of successful evacuations through a framework that incorporates human behavior, traffic, and fire progression models using Bayesian Networks, Agent-Based models, real-world socio-demographic data, and Geographic Information System. Finally, the paper showcases the impact of the socio-demographic profile of different communities on a safe evacuation probability.

Integrating Human Behavior Modeling into a Probabilistic Wildfire Egress Planning Framework

Wildland Urban Interface (WUI) can be defined as “the zone of transition between unoccupied land and human development.” The communities in these areas are particularly vulnerable to wildfires that start and propagate in wildlands. According to the U.S. Fire Administration, the U.S. has more than 70 thousand communities at risk for WUI fires, and the WUI area grows by approximately 2 million acres per year. Numerous efforts have been undertaken to address the dangers of wildfires, including building more resilient infrastructures, advancing techniques for extinguishing fires and exploring the possibilities of controlled fires. Associated with these efforts is the pressing need to ensure the safe evacuation of communities in WUI once they are threatened by wildfires. Evacuation modeling and planning is a challenging and complex problem. It involves human decisions and actions concerning if, when, and how to evacuate; directly impacting the traffic flow during the evacuation. Furthermore, the available time for a community to evacuate is a dynamic element: it changes according to the fire progression, which, in turn, depends on vegetation, weather, among other factors. The models for traffic and fire progression have advanced considerably in the past years. Human behavior modeling during wildfire evacuations has also received significant attention, leveraging existing studies for other natural hazards, such as hurricanes. However, these are mainly developed as standalone, qualitative approaches rather than integrated into a complete egress framework that accounts for traffic modeling and wildfire progression. This paper discusses the existing methods for modeling human behavior during wildfire evacuation and the requirements for such a model to be integrated into a quantitative, probabilistic evacuation planning tool. It further presents the method adopted for integrating human behavior model into the evacuation planning tool WISE (Wildfire Safe Evacuation). WISE calculates the probability of successful evacuations through a framework that incorporates human behavior, traffic, and fire progression models using Bayesian Networks, Agent-Based models, real-world socio-demographic data, and Geographic Information System. Finally, the paper showcases the impact of the socio-demographic profile of different communities on a safe evacuation probability.

Paper MA211 | Download the paper file. |

Name: Marilia Ramos (marilia.ramos@ucla.edu)

Bio:

Country: USA
Company: University of California Los Angeles
Job Title: Research Scientist

Paper 2 MH163

Lead Author: Mohammad Pishahang Co-author(s): Andres Ruiz-Tagle, aruiztag@umd.edu Enrique Lopez Droguett, eald@g.ucla.edu Marilia Ramos, mariliar@g.ucla.edu Ali Mosleh, mosleh@g.ucla.edu

Presenter of this paper: Marilia Ramos (marilia.ramos@ucla.edu)

WISE: a probabilistic wildfire egress planning framework
Wildfire is a significant threat to many communities in Wildland Urban Interface (WUI) areas, and ensuring an efficient evacuation of these communities in case of wildfire is a pressing challenge. Wildfire evacuation modeling consists of three main layers: fire model, human decision-making, and traffic models. An efficient evacuation planning needs thus a comprehensive understanding of each of these layers and their mutual interactions. Numerous methods have been proposed for wildfire risk assessment, focusing on each of these components, but few address the issue considering all these layers. This paper presents a framework for probabilistic evacuation planning in the case of wildfires. The Wildfire Safe Egress (WISE) framework integrates a human decision model, a traffic model, and wildfire dynamics modeling for estimating the probability that a community safely evacuates when in danger by a wildfire. The evacuation success is calculated through a comparison between two competing variables. The Available Safe Egress Time (ASET) determines the total amount of time before the fire reaches a community's borders. This variable depends on the wildfire dynamics. The Required Safe Egress Time (RSET) determines the amount of time a community needs to evacuate safely. The RSET considers population distribution, demographic characteristics, warning system timing and its reliability, available roads network, and the traffic travel times. These variables are modeled in a Bayesian Belief Network (BBN). Next, a Monte Carlo simulation of a Poisson process defined by the community's socio-demographic profile generates the evacuation demand curve. Finally, the traffic model is developed through agent-based modeling of evacuees mobilization on the roads network. The final node of the BBN estimates the probability of a successful evacuation. Having a realistic estimation of this probability helps decision-makers and stakeholders to plan evacuation time, routes, and strategies to mitigate the consequences of a wildfire on a community considering different scenarios. This framework is implemented as a web platform, allowing users to have a practical egress assessment in a visual GIS-based environment.

WISE: a probabilistic wildfire egress planning framework

Wildfire is a significant threat to many communities in Wildland Urban Interface (WUI) areas, and ensuring an efficient evacuation of these communities in case of wildfire is a pressing challenge. Wildfire evacuation modeling consists of three main layers: fire model, human decision-making, and traffic models. An efficient evacuation planning needs thus a comprehensive understanding of each of these layers and their mutual interactions. Numerous methods have been proposed for wildfire risk assessment, focusing on each of these components, but few address the issue considering all these layers. This paper presents a framework for probabilistic evacuation planning in the case of wildfires. The Wildfire Safe Egress (WISE) framework integrates a human decision model, a traffic model, and wildfire dynamics modeling for estimating the probability that a community safely evacuates when in danger by a wildfire. The evacuation success is calculated through a comparison between two competing variables. The Available Safe Egress Time (ASET) determines the total amount of time before the fire reaches a community's borders. This variable depends on the wildfire dynamics. The Required Safe Egress Time (RSET) determines the amount of time a community needs to evacuate safely. The RSET considers population distribution, demographic characteristics, warning system timing and its reliability, available roads network, and the traffic travel times. These variables are modeled in a Bayesian Belief Network (BBN). Next, a Monte Carlo simulation of a Poisson process defined by the community's socio-demographic profile generates the evacuation demand curve. Finally, the traffic model is developed through agent-based modeling of evacuees mobilization on the roads network. The final node of the BBN estimates the probability of a successful evacuation. Having a realistic estimation of this probability helps decision-makers and stakeholders to plan evacuation time, routes, and strategies to mitigate the consequences of a wildfire on a community considering different scenarios. This framework is implemented as a web platform, allowing users to have a practical egress assessment in a visual GIS-based environment.

Paper MH163 | Download the paper file. |

A PSAM Profile is not yet available for this author.
Presenter Name: Marilia Ramos (marilia.ramos@ucla.edu)

Bio:

Country: United States of America
Company: University of California Los Angeles
Job Title: Research Scientist

Paper 3 MB291

Lead Author: Matthew Bucknor

Sodium Fire Protection Systems, Mitigation Strategies, and Risk Analysis
Liquid sodium coolant at nominal sodium-cooled fast reactor (SFR) operating temperatures, approximately 350°C – 550°C, will readily ignite in air environments. The severity of the resulting fire scenario depends on several factors including sodium temperature, amount of available oxygen, geometric factors such as the size and configuration of the leaked sodium volume, and the type of fire (pool fire versus spray fires). To minimize the consequences associated with these fires, sodium fire protection systems (SFPSs) and mitigation strategies are utilized to provide a means of detecting, locating, containing, and suppressing the fires. The SFPSs and mitigation strategies proposed for future U.S. SFRs are based on experience and data from historical SFR facilities and testing programs which no longer exist. Software tools utilized to simulate sodium fires are also based on historical efforts and testing data and their development has been fragmented throughout the last few decades due to interest and sporadic funding. This presentation provides a brief overview of sodium fires, SFPSs and mitigation strategies, and a discussion on the status of sodium fire risk analysis in the U.S.

Sodium Fire Protection Systems, Mitigation Strategies, and Risk Analysis

Liquid sodium coolant at nominal sodium-cooled fast reactor (SFR) operating temperatures, approximately 350°C – 550°C, will readily ignite in air environments. The severity of the resulting fire scenario depends on several factors including sodium temperature, amount of available oxygen, geometric factors such as the size and configuration of the leaked sodium volume, and the type of fire (pool fire versus spray fires). To minimize the consequences associated with these fires, sodium fire protection systems (SFPSs) and mitigation strategies are utilized to provide a means of detecting, locating, containing, and suppressing the fires. The SFPSs and mitigation strategies proposed for future U.S. SFRs are based on experience and data from historical SFR facilities and testing programs which no longer exist. Software tools utilized to simulate sodium fires are also based on historical efforts and testing data and their development has been fragmented throughout the last few decades due to interest and sporadic funding. This presentation provides a brief overview of sodium fires, SFPSs and mitigation strategies, and a discussion on the status of sodium fire risk analysis in the U.S.

Paper MB291 | |

Name: Matthew Bucknor (mbucknor@anl.gov)

Bio: Matthew Bucknor is a Principal Nuclear Safety/Risk Analyst with the Nuclear Science and Engineering Division at Argonne National Laboratory. He has expertise in leading and performing advanced reactor safety and risk assessments for the Department of Energy and in collaboration with industry partners. His areas of technical expertise include: advanced reactor safety analysis, probabilistic risk assessments, component reliability evaluations, mechanistic source term evaluations, and advanced reactor modeling and simulation. Matthew Bucknor holds a Ph.D. in Nuclear Engineering, an M.S. in Nuclear Engineering, and a B.S. in Electrical and Computer Engineering from the Ohio State University.

Country: USA
Company: Argonne National Laboratory
Job Title: Principal Nuclear Safety/Risk Analyst