Clustered Models for the Integration of Sewer Condition Classification Protocols

Abstract

Adoption of a suitable sewer pipeline condition classification protocol is recognized as an indispensable first step in world wide sewer rehabilitation industry. Various condition classification systems for sewers have been developed in this regard. These systems differ according to local requirements in which there is no integrated, unified sewer condition assessment protocol available. Therefore, in order to standardized sewer condition assessment procedures, there is an urgent need of developing such a system. This paper reviews the historical development of different sewer condition classification protocols. The protocols developed by the Water Research Centre (WRc), UK and by the Centre for Expertise and Research on Infrastructures in Urban Areas (CERIU), Canada, have been discussed in detail. In order to integrate both protocols, an unsupervised neural network model has been developed. The integrated protocols are verified by municipal practitioners and experts of the CERIU sub-committee for developing a unified sewer condition assessment system. The integrated models will assist municipal engineers in developing a unified sewer condition assessment system.