Off-campus UMass Amherst users: To download campus access dissertations, please use the following link to log into our proxy server with your UMass Amherst user name and password.
Non-UMass Amherst users: Please talk to your librarian about requesting this dissertation through interlibrary loan.
Dissertations that have an embargo placed on them will not be available to anyone until the embargo expires.
Open Access Dissertation
Doctor of Philosophy (PhD)
Year Degree Awarded
Agha Iqbal Ali
Management Sciences and Quantitative Methods | Operational Research | Operations and Supply Chain Management
This dissertation examines two salient issues that arise in the strategic planning of disaster management operations for providing relief to populations that are impacted by a disaster, such as an earthquake. The first issue is the alleviation of destitution faced by affected populations in the immediate aftermath of a disaster. The second is the establishment of an infrastructure for provision of relief, for a much longer period of time, until normalcy is restored. Central to the alleviation of destitution is the avoidance of critical shortages in meeting the demand for relief supplies. The literature on pro-active and strategic planning of relief operations primarily focuses on the minimization of unmet demand to address shortfalls in meeting required levels of relief. However, compromised handling capacity, which can be attributed to insufficient manpower, can deteriorate provision of supplies. The same is true for transport capacity of which there can be limited capability in the immediate aftermath of a disaster. Better planning and management of resource and supply allocations is made possible by quantifying the destitution faced by affected populations and revealing its relationship to delays in provision of relief. The second issue is that of on-going provision of relief, whether supplies, resources, or information. Key to such relief provision is the establishment of relief centers that are easily accessible by affected populations using the available transport network. Both issues are addressed using mathematical programming in the context of strategic planning of humanitarian logistics for a catastrophic earthquake in Istanbul, of which there is a high probability of occurrence over the first thirty years of the 21st Century.
Bringing visibility to the impact of critical shortages on affected and vulnerable population segments caused by the lack of any commodity has not been addressed in the literature on relief operations. Extant research is on post-disaster reactive operations. While disasters cannot be forecasted with pinpoint accuracy, it is possible to devise pro-active contingency plans for regions of the world that are prone to natural disasters. Proactive contingency plans, proposed in this dissertation, focus on the strategic planning of the geographical and temporal staging of relief supplies with a view to minimize the impact of critical shortages. The manner in which destitution can be alleviated by revealing the impact of delays in the provision of supplies, the availability of transport units, and the deployment of manpower has not been explicitly addressed in the literature. Delays can lead to a critical shortage of one or more of the supplied commodities causing destitution which is quantified by the number of periods a population segment is without provision of supplies. The destitution levels of population segments as they are replenished with supplies that become available over time are tracked using a complex mixed-integer goal programming model which is developed. The model is used to study the impacts of delays in providing relief on destitution and criticality among affected populations for a highly probable (62±15%), catastrophic earthquake in the greater Istanbul area. Making use of the estimates for seismic hazard and damage for the districts of Istanbul that are provided in the report published by Japan International Cooperation Agency (JICA) in collaboration with the Istanbul Metropolitan Municipality, an empirical study reveals the impacts of the three sources of delay and their significance for different types of supplies and need in different segments of impacted populations.
Centers for relief operations meet the continued need for resources, supplies, and information among affected populations until the restoration of normalcy. The literature on location-allocation of centers for humanitarian logistics have employed location methodology which comprises a suite of models such as the p-median or maximal cover models. These models have not accounted for traffic networks, the travel times on links of the network, and the potential delays that can occur due to congestion on the links. The centrality of the existing traffic network and travel times on links of the network as the flow of traffic increases in locating relief centers is not accounted for in such models. In this dissertation two significant aspects of locating centers are addressed in a new mathematical programming model: First, the number of, and locations for, the centers to be established to provide a given level of access to populations in various neighborhoods of the affected region. Second, the implementation plan for the centers, detailing the identification of the specific centers that are made available over time. Further, the model also addresses the two key issues of the implied capacity of each center and of the assumed patterns of access to, and demand at, each center. These two issues are intertwined in that varying frequencies of access among different population segments dictate that populations be provided with equally, with respect to travel time, accessible alternative relief centers. The inherent stochasticity of frequency of access needs necessarily to be accounted for when determining the locations of centers. The optimization model that is developed to determine locations of supply sites and locations of centers is a two-stage stochastic mixed integer non-linear programming model over a network in which supplies move from selected supply sites to selected relief centers and subsequently acquired by affected populations accessing the relief centers over the traffic network. The travel time on each link grows exponentially with the traffic, or flow, on the link. The nonlinearity reflects the behavior of populations headed from any neighborhood, i.e. population center, to any one of centers that can be made available to them. The model identifies relief centers to locate from among a set of potential sites for centers such that the total travel time over the network is optimized. The model assumes different levels of access for populations in different neighborhoods that are defined by pre-specified distance thresholds for access to a center. The solution of the model is addressed via a piece-wise linear approximation of the objective function which is separable, convex, and monotone increasing. The model is employed in a computational study for the identification of supply sites and relief centers for stochastically varying frequencies of access by populations in the one hundred twenty-three neighborhood of Greater Istanbul. The stochastic variations examined range from fixed daily, bi-weekly, and weekly access frequencies to totally randomized access frequencies during an hour. Further, a computational study reveals an implementation plan for establishing relief centers to ensure that easy access with minimal degradation of travel times is enabled for all populations in the neighborhoods of Greater Istanbul.
Ince, Guven, "Resource and Supply Allocation and Relief Center Location for Humanitarian Logistics" (2015). Doctoral Dissertations. 305.