Date of Graduation

12-2022

Document Type

Thesis

Degree Name

Master of Science in Industrial Engineering (MSIE)

Degree Level

Graduate

Department

Industrial Engineering

Advisor/Mentor

Zhang, Shengfan

Committee Member

Parnell, Gregory S.

Second Committee Member

Barry, Michelle L.

Keywords

Arkansas; Levee system; Maintenance; MCDM; Ranking

Abstract

There are 208,009 properties in Arkansas that have more than a 26% chance of being severely affected by flooding over the next 30 years, which represents 13% of all properties in the state. A levee system is designed to reduce the flooding risk for urban and rural communities; however, most of the state's levees have been significantly outdated or built with engineering standards less rigorous than current best practices. The Levee Safety Action Classification (LSAC), as recorded in the National Levee Database (NLD), communicates the risk associated with living behind a particular levee and assists local, state, and federal stakeholders in identifying and prioritizing funding needs. It is expected that LSAC will decrease as flood risk decreases. However, in some cases, the LSAC for a particular levee may stay High even if it is in perfect condition when the area behind it is densely populated or significantly developed. We develop a multi-criterion ranking framework, integrating Principal Component Analysis (PCA) and Multi-Criteria Decision-Making (MCDM) methods (i.e., a CRITIC-TOPSIS approach), for prioritizing maintenance of the Arkansas levee systems in Arkansas using the NLD data. The results show the rankings from each method are not significantly different from each other. When compared to the LSAC, it is important to note that the top-ranked levee systems obtained using the PCA or CRITIC-TOPSIS ranking method often have a low to moderate LSAC. Therefore, they are mostly in low priority of maintenance by according to the US Army Corps of Engineers (USACE), despite having a higher probability of being in poor conditions or having design and performance issues. Moreover, we perform a cost-benefit analysis, comparing the operating and maintaining costs with the associated benefits to determine maintenance prioritization. Additionally, we propose other modeling frameworks, such as multi-objective decision analysis (MODA) and sequential decision-making methods, that are suitable for this problem but cannot be implemented in this research due to limited data and stakeholder involvement.

Download

Share

COinS