Freight Distribution During Disasters: Measuring and Improving Operational Performance of Critical Systems

Abstract

The frequency and intensity of weather-related natural disasters have increased in the last őve decades. Moreover, the US faces more than a third of the disaster-related economic losses globally, majority of which are from storms. As the demand for distribution of essential freight increases during disasters, the physical and operational constraints decrease the capacity of the freight distribution systems. Accordingly, public and private-sector stakeholders seek disaster preparedness and response interventions to ensure timely and economic distribution of vital freight to the population in need. The goal of this thesis is to facilitate better strategic and tactical planning that results in higher operational performance of essential freight distribution systems during disasters. We study two critical freight distribution systems, namely, downstream fuel distribution and full truckload transportation of general freight. Truckload transportation plays a vital role in distributing relief supplies during emergencies, and fuel is required for humanitarian operations such as running generators, moving emergency response crews, and evacuation of the affected population. We collaborate with The US Federal Emergency Management Agency in response to multiple North-Atlantic storms and measure the operational performance of these systems under regular and disaster conditions, as well as identify public and private-sector interventions to make the performance better during future disasters. Our research contributes to the bodies of disaster modeling and management, fuel distribution, service procurement, and truckload procurement literature by, i) creating system level understanding of multi-server tandem cyclic queues with time-limited customers, ii) studying process improvement interventions for disasters, iii) quantifying the magnitude, geographical extent, timing, and duration of the causal effects of disaster conditions and consequent disaster relief activities on transportation procurement prices, iv) using datadriven analysis to design ŕexible truckload contracts that consider uncertainty in demand, and v) modeling dynamic-pricing where the buyer offers the price to service providers. In this thesis, we provide several actionable insights for public and private-sector stakeholders to manage freight distribution during future disasters. We identify which process improvement interventions are best suited for which type of downstream fuel distribution system, and which storage terminals should be prioritized under limited budget. We also measure how private-sector shippers should account for changes in truckload spot procurement prices during disaster episodes to manage their budgets and operational decisions. Moreover, we offer an alternate dynamic-priced truckload contract solution for public-sector shippers that deal with uncertain episodic demand in response to disasters. We demonstrate the impact of our research by implementing it to multiple real-life case studies in the US. Furthermore, our methodologies and results are generalizable to other geographical regions as well as other disaster conditions. Thus, we hope that they are used by public and private-sector actors to better manage essential freight distribution moving forward.