Project Summary: The accelerated growth of modern cities is creating unique challenges to the interdependent urban infrastructures that support contemporary technical, social, and economic vitality.  Infrastructures such as power grids, water distribution networks, and telecommunication systems are growing in size and complexity at a pace that prevents understanding the effects of this rapid evolution on their functionality and resilience.  If natural hazards or human-induced disruptions strike, urban infrastructures become the object of the perturbation and the means for prompt recovery.  This dual role demands understanding of the mechanisms that enable failure propagation within and across urban systems, and the strategies that can be adopted to improve their residual functionality and recovery time in the event of unforeseen disruptions.  Hence, this project aims at developing high fidelity models for interdependent urban infrastructures that can capture cascading failures and interdependencies.  The adopted approach will simultaneously model network topology, internal flow patterns, and the pathways that support interaction across different systems.  These models will be able to quantify global interdependent effects, and help prioritizing intervention actions to improve redundancy and accelerate urban recovery if disruptions occur.  Interdependent optimization, which requires coordination of multiple objectives and constraints, will yield the most desirable socio-technical strategies.

 

This project also supports the concept of system-level thinking for the new generation of civil engineers.  A new course on “Complex Urban Systems” at Rice University, and guided-visits to utilities of the city of Houston, Texas, will provide students the opportunity to engage in modern problem-based learning strategies. This project will also create an enriched summer experience for undergraduate students interested in urban infrastructures and working on network data mining and interdependent failure analysis problems.  Since urban systems reach multiple demographics, a culturally and ethnic diverse pool of interns is expected to bring in unique perspectives and insights for their analysis.  This project will also leverage the successes of the NSF-funded Mid-America Earthquake Center by providing new tools for the network analysis modules of MAEviz, the Center’s seismic impact assessment software package used today by urban risk analysts, city planners, and emergency managers.  Also, interdependent infrastructure analysis frameworks will find applications in cybersecurity assessment of technologically advanced interdependent lifeline systems.

Project Summary: The Houston metropolitan area is home to over 5.5 million inhabitants; it is the sixth-largest in the United States. When Hurricane Rita bore down on Houston, 2.5 million Houston area residents were evacuated, making it the largest evacuation in the history of the United States. Critics of the evacuation process believe that authorities waited far too long to permit outward- bound cars to use both sides of the Interstate highways. It is estimated that residents who decided to leave Houston, and that did not need to evacuate, had a significant impact in the mobility of populations at true risk.  Indeed, if a mere 15% of Harris County residents inside the loop had sheltered in place in their homes, the transportation system could have absorbed evacuees from the more vulnerable Galveston area. Instead, thousands of residents clogged major freeways and witnessed gasoline shortages, and difficult recovery.  The need for good planning and information integration tools which can help regional authorities develop sound evacuation and sheltering policies has never been more critical.

 

The goal of this research is to develop tools at the household level for risk assessment and evaluation of evacuation policies under hurricane hazards for the City of Houston and its surrounding communities. The tools integrate engineering analyses of existing infrastructure at the level of individual households, and empirically gathered sociological data on human behaviors in response to threats and advisories, to drive detailed computational simulations of sheltering and evacuation plans at the level of individual streets. In contrast to coarse-grained models and tools used by FEMA and the FHWA  (HAZUS, OREMS, ETIS) into which it is difficult to fold in detailed structural and behavioral data, our tools will provide emergency planning and management authorities with information to study the impact of decisions at a much finer granularity. Examples are the evacuation of specific streets or neighborhoods, choice of evacuation or sheltering destinations at the level of zip codes, and the opening of contra-flow lanes at the level of specific exits on specific streets. Since the information and simulations to be generated will be directly comprehensible to the general public, emergency planners can use this project’s tools and simulations to easily communicate the impact/value of their policies to the public. It is envisioned that weather stations can be used to relay regional advisories to residents on individual streets showing them the status of the relevant roadways and providing a set of dynamically updated alternate routes with predictions on clearance times on each of them.  Risk assessment data on households can also be made available to the public through HCAD to guide decision making at the level of individual households. By empowering people with reliable information on the hurricane resistance of their properties, there is a potential to significantly reduce pressure on evacuation routes.