TY - JOUR
T1 - Development of sustainable water infrastructure
T2 - A proper understanding of water pipe failure
AU - Taiwo, Ridwan
AU - Shaban, Ibrahim Abdelfadeel
AU - Zayed, Tarek
N1 - Funding Information:
In addition, the consequences of water pipe failure are numerous. Some of them are property loss, repair and remediation costs, deterioration of human health, damage to the environment, and customer dissatisfaction (Dawood et al., 2021; Fares and Zayed, 2010; Katarzyna Pietrucha-Urbanik and Tchórzewska-Cieślak, 2020). Moreover, the failure of the WDN and its associated system affects global economic conditions. For instance, in 2006, although the original need to maintain and rehabilitate the WDNs in the US was $6 billion, only about $1.2 billion was spent because of the unavailability of funds (Paradkar, 2012). Therefore, a WDN should be developed in such a way that several technical requirements, i.e., functionality, serviceability, and durability (Farshad, 2006), are achieved. To function optimally and sustainably, the material properties of components of a WDN in relation to mechanical, thermal, and durability, which determine the serviceability state of the network, are highly important. The system should be designed to provide adequate resistance against chemical, biological, and other aggressive elements from the environment (American Works Water Association, 2002). Importantly, to foster sustainability, the economic aspect of the system should not be overlooked in the design stage.Soil type: Soil that acts as a foundation for the pipe and soil used as backfilling play an important role in the number of loads applied to the pipe. For instance, external loading can get increased on the water pipe due to soil compaction, frost penetration, and shrinking and swelling. These phenomena are common in clayed soils (Mackey et al., 2014). Additional stresses are induced on the pipes when soil shrinks due to a decrease in soil volume. This means that the pipe will not be supported uniformly (bedding), which makes it act like a beam, thus, creating high bending moments. Apart from the produced bending moments, shear forces are also developed as a result of differential soil movement (Hu and Hubble, 2007). Furthermore, soft soil, such as organic soil, is not suitable for bedding support due to its high retention capacity of moisture and low bearing capacity of the soil (Pritchard et al., 2013). On the other hand, sandy soil contributes less to pipe failure due to its higher drainage capacity and resistivity (Doyle et al., 2003). Singh (2011) attributed the lowest number of failures to sandy soil (722 breaks), while materials such as mud and coral had a higher number of failures (1080 and 1423 breaks, respectively). However, due to their larger particle size, care must be taken when using sandy soil for bedding support in locations prone to excessive erosion (Pritchard et al., 2013).The probability of failure model predicts the likelihood of failure of a given water pipe. The prediction outcome from this model is expressed as a value between 0 and 1, with 0 indicating that the pipe is functioning properly, and 1 indicating complete failure of the pipe with no residual reliability. The prediction outcome provides valuable information to water utilities in assessing the condition of the pipes in their network and in making decisions related to maintenance, repair, and replacement. Probability of failure models can be broadly grouped into two: physical and statistical-based models. For the physical models, the probability of failure is determined by making a comparison between the stress exerted on a pipe and the pipe resistance (Amaya-Gómez et al., 2019; Li et al., 2021). While experimental data are used to develop physical models, historical data of WDNs are often used for statistical-based models. The statistical-based models include classical mathematical models such as logistic regression (Fan et al., 2022), Poisson distribution (Singh and Adachi, 2012), Weibull distribution (Ward et al., 2017), proportional hazard (Debón et al., 2010), hierarchical beta process (Luo et al., 2017) and machine learning algorithms such as artificial neural network (ANN) (Fan et al., 2022), support vector machine (SVM) (Robles-velasco et al., 2020), random forest (Raspati et al., 2022), and light-gradient-boosting machine (Fan et al., 2022), amongst others.
Funding Information:
The authors gratefully acknowledge the support from the Innovation and Technology Fund (Innovation and Technology Support Programme (ITSP)) under grant number ITS/033/20FP, and the Water Supplies Department of Hong Kong.
Publisher Copyright:
© 2023 Elsevier Ltd
PY - 2023/4/20
Y1 - 2023/4/20
N2 - The need for sustainable water infrastructure systems continues to grow as clean water is essential for daily life. Despite efforts to sustain water distribution networks (WDNs), they often experience frequent failures, leading to several environmental, social, and economic consequences. Previous studies have investigated the causes of water pipe failure in different contexts. However, a comprehensive and holistic understanding of these causes is lacking in the literature. Therefore, this study contributes to the existing knowledge by presenting 1) a scientometric analysis of the previous literature, 2) a systematic discussion of the causes, 3) an Analytical Hierarchy Process model and fault tree logic to prioritize and map the causes, respectively, and 4) an overview of techniques used in developing failure prediction models. The scientometric analysis reveals that little attention has been paid generally to the operational causes of water pipe failure. The same trend was supported by the systematic review, which divides a total of 33 causes into three main categories: pipe-related, environment-related, and operation-related causes. This study gives insights to academics and practitioners working in this domain on the contributions of various factors to the failure of water pipes, which would be useful in designing a sustainable and resilient WDN.
AB - The need for sustainable water infrastructure systems continues to grow as clean water is essential for daily life. Despite efforts to sustain water distribution networks (WDNs), they often experience frequent failures, leading to several environmental, social, and economic consequences. Previous studies have investigated the causes of water pipe failure in different contexts. However, a comprehensive and holistic understanding of these causes is lacking in the literature. Therefore, this study contributes to the existing knowledge by presenting 1) a scientometric analysis of the previous literature, 2) a systematic discussion of the causes, 3) an Analytical Hierarchy Process model and fault tree logic to prioritize and map the causes, respectively, and 4) an overview of techniques used in developing failure prediction models. The scientometric analysis reveals that little attention has been paid generally to the operational causes of water pipe failure. The same trend was supported by the systematic review, which divides a total of 33 causes into three main categories: pipe-related, environment-related, and operation-related causes. This study gives insights to academics and practitioners working in this domain on the contributions of various factors to the failure of water pipes, which would be useful in designing a sustainable and resilient WDN.
KW - Environment-related causes
KW - Failure modes
KW - Operation-related causes
KW - Pipe-related causes
KW - Sustainable water infrastructure
KW - Systematic review
KW - Water pipe failure
UR - http://www.scopus.com/inward/record.url?scp=85149440761&partnerID=8YFLogxK
U2 - 10.1016/j.jclepro.2023.136653
DO - 10.1016/j.jclepro.2023.136653
M3 - Review article
AN - SCOPUS:85149440761
SN - 0959-6526
VL - 398
JO - Journal of Cleaner Production
JF - Journal of Cleaner Production
M1 - 136653
ER -