Margin of Safety Based Flood Reliability Evaluation of Wastewater Treatment Plants: Part 2- Quantification of Reliability Attributes

Abstract

In a companion paper (Part 1), the basic concepts are discussed in details. In this paper, the application of the concepts and the proposed methodology in that paper are utilized to set the strategies in order to quantify reliability attributes. The case study is Hunts Point wastewater treatment plant and its sewershed in Bronx, New York City, so that it could function during flood. The strategies are selected and reliability of the Hunts Point plant is estimated before and after using best management practices (BMPs). Therefore, the copula based non-stationary 100–year flood frequency analysis of rainfall and storm surge analyzed in the companion paper (Part 1), are used to obtain the design values of surge and rainfall. A differential evaluation Markov Chain with Bayesian interface is used in that paper for parameter estimation. Because the co-occurrence of surge and rainfall is more critical in coastal areas, the design value resulted from bivariate analysis has been considered in this study. In this Part 2, a multi-criteria decision making (MCDM) approach, which characterized the uncertainty of sub-criteria in load-resistance concept, is applied. The Margin of Safety (MOS) concept is extended to numerically assess a rather non-probabilistic MCDM evaluation of load and resistance to estimate the reliability of the plant. Finally, Effective strategies are selected for flood hazard mitigation to improve the reliability and performance of the plant. The results present an increase of about 46% in reliability after using BMPs. The framework presented in this paper is applicable to other coastal watersheds and could be a platform for addressing some pressing issues in coastal preparedness with impacts on design criteria of coastal infrastructure.