Best Practices for Assessing Culvert Health and DeterminingAppropriate Rehabilitation Methods

Several culvert structures in the United States are in a deteriorated state needing immediate attention to prioritize critical culverts and rehabilitate them in a timely and economical manner. The overarching goal of this study is to provide technical guidance to SCDOT in effectively managing their culvert infrastructure. This study identified best practices for effectively inspecting deteriorating culvert infrastructures and choosing appropriate rehabilitation methods. The synthesis of literature on best practices for culvert rehabilitation is formulated into a simple decision-making architecture using the principles of analytical hierarchy procedure (AHP). Specifically, a Microsoft Excel-based Culvert Renewal Selection Tool (CREST) is developed to assist SCDOT in shortlisting suitable construction methods for the renewal of failing culverts. Furthermore, a multinomial logistic regression (MLR) model as well as an artificial neural network (ANN) model is developed to predict the condition scores of a culvert based on historic inspection data recorded into SCDOT’s culvert inspection database. A simple analytical hierarchical procedure (AHP) is used to subsequently prioritize critical culvert structures based on their condition scores evaluated on various defect categories. The prioritization model has been demonstrated using the inspection data available in the SCDOT’s culvert inventory database.