The challenges of predicting pipe failures in clean water networks: a view from current practice

Abstract

Pipe failure models can aid proactive management decisions and help target pipes in need of preventative repair or replacement. Yet, there are several uncertainties and challenges that arise when developing models, resulting in discord between failure predictions and those observed in the field. This paper aims to raise awareness of the main challenges, uncertainties, and potential advances discussed in key themes, supported by a series of semi-structured interviews undertaken with water professionals. The main discussion topics include data management, data limitations, pre-processing difficulties, model scalability and future opportunities and challenges. Improving data quality and quantity is key in improving pipe failure models. Technological advances in the collection of continuous real-time data from ubiquitous smart networks offer opportunities to improve data collection, whilst machine learning and data analytics methods offer a chance to improve future predictions. In some instances, technological approaches may provide better solutions to tackling short term proactive management. Yet, there remains an opportunity for pipe failure models to provide valuable insights for long-term rehabilitation and replacement planning.

An evolution of statistical pipe failure models for drinking water networks: a targeted review

“…For structured data, such as that seen in pipe failure data, supervised grey-box models are more appropriate (see Figure 3). Supervised models can be tuned for improved model accuracy and are interpretable through variable importance measures and partial plots (Wols et al 2019), which is more appealing than black-box approaches to industry professionals (Barton et al 2021). There are various studies in the literature that have applied machine learning and explored its effectiveness.…”

Asset management analytics for urban water mains: a literature review

“…Consequently, removing missing values from a dataset can result in negative effects on data-driven models, unreliable parameter predictions, loss of valuable information, bias, and poor models (Tang et al 2019). Therefore, it is necessary to keep as much information as possible (Barton et al 2022).…”

“…Left truncation occurs when the records of pipe failures before collecting data are missing. Like censoring, this is also always the case in a water main dataset and it is acknowledged widely (Barton et al 2022). The effects of left truncation have been overlooked in many studies (Snider and McBean 2020b) even though this issue causes a systematic bias, especially for survival analysis models (Scheidegger et al 2015;Xu and Sinha 2019).…”

“…The existing review articles have mostly focused on water main prediction models (St. Clair and Sinha 2012;Dawood et al 2020b), the effect of availability and quantity of the database on WM failure models (Snider and McBean 2020a;Chen et al 2022), the effect of the limited, uncertain dataset on WM failure models (Jenkins et al 2014), and the effect of combined datasets from different utilities on the performance of machine learning models for predicting future breaks (Chen et al 2022). Other reviews have discussed different approaches to optimizing rehabilitation and maintenance strategies for WMs and integrated infrastructures (Abusamra 2018;Ghobadi et al 2021;Ramos-Salgado et al 2022;Shahata et al 2022; Barton et al 2022). This review aims to address issues of the failure prediction models developed recently and optimal maintenance strategies conjointly.…”

“…What are the pros and cons? • Failure models of various types have been used to analyze water main datasets (Economou et al 2012; American Water Works Association 2019; Snider and McBean 2020a; Snider 2021; Barton et al 2022).…”