Defect-Based Condition Assessment Model for Sewer Pipelines Using Fuzzy Hierarchical Evidential Reasoning

Abstract

Assessing the condition of sewer pipelines is essential for addressing associated problems as a result of deterioration to maintain functionality and avoid additional social costs. This paper presents the development of a defect-based condition assessment model for sewer pipelines. The model aims to cover the structural, operational, and rehabilitation defects that are associated with the pipeline, joints, and maintenance holes on pipe segments. The analytic network process (ANP), which considers the interdependencies between pipeline components and defect types, was used to deduce the relative importance weights through questionnaires and interviews with infrastructure experts and professionals. The model employs the defects’ severities to develop fuzzy membership functions based on a predefined linguistic condition grading scale that would precisely indicate the degree of distress. Hierarchical evidential reasoning (HER) and the Dempster-Shafer (D-S) theory were used to integrate the defects’ conditions and to evaluate the overall condition of sewer pipelines, which is represented as a crisp value calculated by the weighted average defuzzification method. A validation dataset was used in the model validation and protocol verification. This model is expected to minimize the inaccuracy of sewer condition assessment through the application of severity, uncertainty mitigation, and robust aggregation. It could also benefit asset managers by providing a precise condition overview for maintenance, rehabilitation, and budget allocation purposes.