A multi-objective optimisation model for sewer rehabilitation considering critical risk of failure

Abstract

A unique methodology for the optimal specification of sewer rehabilitation investment is presented in this paper. By accounting for the critical risk of asset failure, this methodology builds on previously successful work which explored the application of multi-objective optimisation tools to assist engineers with the specification of optimal rehabilitation strategies. The conventional sewerage rehabilitation specification process relies on the expertise of professional engineers to manually evaluate CCTV inspection information when determining the nature and extent of the rehabilitation solution. This process is not only tedious and subjective but it has no quantifiable means of identifying optimal solutions, or, possible combinations of optimal solutions in the delivery of catchment wide rehabilitation programmes. Therefore, the purely manual process of sewer rehabilitation design leaves a number of unanswered questions, such as: (1) Does the solution offer the greatest structural benefit to the network? (2) Is the solution the most cost effective solutions available? And, (3) Does the solution most greatly reduce the risk of critical asset failure? The application of a multi-objective genetic algorithm optimisation model, coupled with an enhanced critical risk methodology, has successfully answered these questions when applied to a case study data set provided by South West Water (UK).

An optimised total expenditure approach to sewerage management

“…In light of the above, it is widely foreseen that the UK water industry will begin to evaluate investment on a total expenditure (totex) basis. To assist in this transition, a previously successful sewer rehabilitation optimisation model (Ward and Savić, 2012) has been modified to adopt a totex approach to the problem of strategic sewer rehabilitation. This approach allows the user to consider the trade-offs that exist between the capital and operational benefits of different intervention strategies.…”

“…Unlike in previous work by the authors (Ward and Savić, 2012), the model described in this paper presents the problem in a much more computationally efficient manner. Rather than assigning decision variables to each 1-m section of the sewer to indicate whether the section is being repaired (1) or not (0), the decision variables are now only applied to defective sections of the sewer.…”

“…It uses conventional CCTV data to identify An optimised total expenditure approach to sewerage management Ward, Kawalec and Savić which sewers to rehabilitate and the extent and nature of the rehabilitation required, while also considering the serviceability benefits that different rehabilitation schemes bring to the customer. The study was commissioned by the UK water utility South West Water with the aim of embedding sewer network performance, as a decision-making criterion, into the previously successful sewer rehabilitation model developed by Ward and Savić (2012). The 2012 study documents the benefits associated with using a multi-objective optimisation model to evaluate the trade-offs between structural condition improvement (objective function 1) and cost (objective function 2).…”

Identificación de factores de tipo categórico relacionados con la condición estructural de tuberías de alcantarillado de Bogotá a partir de conceptos de entropía de la información

“…Según reportes provenientes de países desarrollados y en vías de desarrollo, esta infraestructura presenta tasas alarmantes de envejecimiento y deterioro [2]. Es así como la mayoría de redes de alcantarillado son cada vez más propensas a fallar [3], además de traer como consecuencia su deterioro estructural, hidráulico e impactando el nivel de servicio y, por consiguiente, la calidad de vida de las comunidades [2,4,5].…”

Prediction of CSO Chamber Level Using Evolutionary Artificial Neural Networks

“…Moradi & Dariane [11] proposed an EANN to derive operating policies for complex reservoir operations. A number of EAs have been developed for use in sewer networks, largely to help identify optimum design and rehabilitation strategies of sewer systems to reduce spill events [12,13] and for operational control of networks to minimise overflows [14,15]. EANNs have not yet, however, been applied to CSO level forecasting.…”