A review of secondary contamination of drinking water quality in distribution systems: sources, mechanisms, and prospects

The Sustainable Development Goals, as established by the United Nations, emphasize the critical importance of securing access to sustainable, adequate, and clean water (Howard 2021). While the primary concern regarding drinking water quality is the pollution of source water (Vyas & Bhatnagar 2023; Kennedy et al. 2024), secondary contamination of treated water during distribution constitutes an equally significant challenge (Han et al. 2018; Gleeson et al. 2023). Such contamination is attributed to the deterioration of water quality resulting from prolonged transportation and storage within drinking water distribution systems (DWDSs) (Nieuwenhuis et al. 2021).

As the majority of pipes in DWDSs are composed of metals such as iron and copper (Li et al. 2016), metal release is the primary cause of secondary contamination in drinking water (Li et al. 2019). Corrosion scales that develop on the inner surfaces of metallic pipes can act as reservoirs for heavy metal contaminants (Salehi 2022), which may be mobilized under varying water chemistry and hydraulic conditions (Li et al. 2020a). The presence of elevated levels of heavy metals in drinking water poses serious health risks (Arunakumari et al. 2023), including liver damage associated with Cu leaching and developmental disorders in children due to Pb exposure (Guo et al. 2022; Chu et al. 2024).

Microbial regrowth is another factor contributing to the secondary contamination of drinking water, particularly evident with the consumption of chlorine residuals in DWDSs (Hu et al. 2021; Xiong & Wang 2022). Microorganisms possess the ability to adhere to pipe surfaces, resulting in the formation of biofilms (Muhammad et al. 2020), which, in turn, promote microbial regrowth by ingesting nutrients from bulk water (Yan et al. 2022). Moreover, these biofilms can harbor opportunistic pathogens, thereby posing potential health risks to consumers (Thom et al. 2022).

In addition, disinfection by-products (DBPs), generated by the reaction of precursors and disinfectants in DWDSs, are also a significant factor involved in the secondary contamination of drinking water (Lin et al. 2022; Villanueva et al. 2023). DBPs have been associated with a range of adverse health effects, including skin allergies, immune system deficiencies, reproductive system abnormalities, and endocrine disorders (Li & Mitch 2018). So far, over 800 types of DBPs have been identified, such as trihalomethanes (THMs), haloacetic acids (HAAs), and bromate (BrO₃⁻). The formation of DBPs in drinking water is primarily influenced by the composition of the source water and the disinfection methods employed (Xiao et al. 2023).

Accordingly, secondary contamination of water quality within DWDSs has received considerable attention, leading to a growing body of research in this domain. This study aims to deliver a comprehensive review of the existing literature concerning secondary contamination in DWDSs, with a focused examination of the causes, impacts, and mitigation strategies associated with heavy metal release, microbial regrowth, and DBP formation. Emphasis is also placed upon elucidating the interrelations among these contamination types. In addition, this study endeavors to propose directions for future research based on a thorough analysis of the scholarly literature and pertinent policies. The insights garnered from this investigation are expected to significantly enhance the academic community's comprehension concerning secondary contamination issues and contribute substantively to the advancement of strategies aimed at safeguarding drinking water quality.

The release of heavy metals is fundamentally associated with electrochemical corrosion processes, alongside the dynamics of metal oxide dissolution and precipitation (Li & Mitch 2018; Jia et al. 2022). This section reviews the prevailing knowledge regarding the metal release, its associated health risks, and the implementation of preventive strategies against hygienic emergencies in DWDSs.

The release of metals within DWDSs is predominantly attributed to the metallic materials constituting the infrastructural pipelines (Tian et al. 2022a). The exposure of these metallic components to aqueous environments initiates a series of chemical and electrochemical oxidation processes, leading to the release of metal ions (Qin et al. 2021). Concurrently, corrosion scales that develop on the internal surfaces of the pipes can detach under varying hydraulic conditions or dissolve due to alterations in water chemistry (Faustino Magalhães et al. 2022). Iron pipes, due to their extensive application and prevalence, are the principal contributors to the dissemination of Fe, which is the most frequently reported metal in DWDSs (Li et al. 2016). Likewise, empirical studies have demonstrated that systems composed of copper and lead pipes exhibit measurable concentrations of Cu and Pb, respectively (Li et al. 2019; Fu & Xi 2020).

Although metallic pipelines are recognized as the primary sources of metal release into DWDSs, other components of the system may also contribute to this phenomenon. Specifically, the deployment of galvanized iron pipes in certain infrastructures can result in the leaching of Zn due to the degradation of coatings over time (Li et al. 2016, 2020b). Additionally, the galvanized layer might contain trace levels of other metals, such as 0.001–2% Pb, thus posing a risk of Pb contamination in water supplies (Li et al. 2020b). Moreover, ancillary system components, including pipe fixtures, water meters, and welded joints, have also been identified as sources of Cu and Pb release (Ghoochani et al. 2022). Notably, the metal release issue is not confined to metallic pipelines; non-metallic plastic pipes have been shown to include trace-level metals, such as Pb, which are used as stabilizers during the manufacturing process (Endale et al. 2022). For instance, analysis has revealed that plastic pipes can contain a significant proportion of Pb (1.4–2.8%), originating from flame retardants, antioxidants, and coloring agents (Zhang & Lin 2015; Diera et al. 2023). As a result, levels of Pd released after a 4-h stagnation period exceeded the 15 μg/L action level established by the US Environmental Protection Agency (US EPA) (Zhang & Lin 2015). Besides, in regions utilizing aged asbestos-cement pipes for water transport, there is a documented propensity for leaching trace metals such as Al, Mn, and Zn (Zavasnik et al. 2022).

Furthermore, corrosion scales that develop on metallic pipe surfaces over extended periods of operation can serve as reservoirs for retaining trace-level metallic substances from the water (Anderson et al. 2022). These scales can enrich metallic concentrations to significant levels. A prominent study conducted by Li et al. identified substantial Pb enrichment (1–3 g/kg) within corrosion scales formed on low-lead (<10 mg/kg) galvanized steel coupons exposed to treated water (Li et al. 2020b). The accumulation of these metals raises concerns about their potential release back into the water, posing a considerable risk of severe contamination (Li et al. 2020a).

As aforementioned, the dynamics of the metal release in drinking water are intrinsically linked to the materials constituting the pipe systems. Beyond that, however, both water quality and hydraulic conditions in DWDSs significantly affect the rate and extent of the metal release.

Changes in pH and dissolved oxygen levels play a critical role in modulating the oxidation processes of metal bases, thereby affecting the stability of oxide scales. This stabilization or destabilization critically impacts the rate and likelihood of metal release within the system (Tian et al. 2022b). Research indicates that elevated concentrations of sulfate and chloride can enhance metal release from corrosion scales (Peng et al. 2013; Fabbricino & Korshin 2014; Rabeiy 2018). In contrast, increased alkalinity produces adverse effects on various metals; specifically, while the release of Fe may be inhibited, the release of Al and Cu can be enhanced under stagnant conditions (Li et al. 2020a). Furthermore, the strategic application of specific corrosion inhibitors, such as orthophosphate, has demonstrated efficacy in mitigating the release of Fe, Mn, Zn, and Pb (Liggett et al. 2024) by promoting the formation of stable carbonate scales (Zhang et al. 2021). These scales serve to protect the underlying metal base from further corrosion. Besides, water temperature also significantly influences pipe corrosion and metal release, as it affects reaction rates, chemical equilibrium, and microbial processes (Li et al. 2020b; Tian et al. 2022b). Notably, the practice of switching water sources, frequently implemented in water management systems, induces changes in water quality that disrupt the original equilibrium in DWDSs, leading to substantial metal release (Li et al. 2020a; Rockey et al. 2021). The cumulative impact of these multifarious changes, caused by the collective alteration of various water quality parameters, is contingent upon the physiochemical characteristics of the involved water sources and the operational capacity of the water treatment infrastructures in place (Lin et al. 2021; Kyritsakas et al. 2023).

The hydraulic conditions within DWDSs encompass both the dynamic flow patterns and the residence time of water throughout the pipeline network (Fabbricino & Korshin 2014). An elevation in flow velocity has been found to increase the shear stress exerted on pipe surfaces, which can result in the detachment of heavy metals from corrosion scales. This phenomenon leads to the formation of a compact scale layer, thereby mitigating the potential for metal release (Weston et al. 2019; Sunny et al. 2020; Boxall et al. 2023). For instance, previous studies have indicated that an increase in shear stress from 0.5 to 3 Pa correlates with an enhancement in Fe release from 30 to 75 μg/L (Sharpe et al. 2019). However, prolonged water retention fosters an anoxic environment that promotes the reduction and dissolution of corrosion products, which could consequently facilitate metal release (Li et al. 2020a; Preciado et al. 2021; Weston et al. 2021). Empirical research has shown that under stagnant conditions, the release of metals such as Fe, Mn, and Ni is significantly enhanced, with concentrations being 2, 90, and 4 times greater, respectively, compared to those under flowing conditions (Li et al. 2020a). Furthermore, alterations in hydraulic retention time induce shifts in microbial communities and activities, which subsequently influence the redox dynamics of heavy metals through various microbial metabolic processes (Proctor et al. 2020; Tian et al. 2021; Li et al. 2024a).

As excessive intakes of heavy metals pose serious health risks to humans, numerous countries and organizations have established guidelines governing the permissible levels of metallic components in drinking water. A comparative analysis of quality standards prescribed by various entities (Table 1) reveals a general consistency in these guidelines across different geopolitical regions.

The World Health Organization (WHO) guidelines do not stipulate any specific restrictions on Fe concentration, despite the acknowledgment that the excessive release of Fe can give rise to aesthetic concerns, resulting in consumer complaints commonly referred to as the ‘red water’ issue (Wang et al. 2021a). Likewise, surpassing permissible limits of Cu is implicated in the ‘blue water’ issue, which is associated with an enhanced risk of gastrointestinal, liver, neurological, and immune system complications (Li et al. 2020a; Meng et al. 2024). An upper limit of Al concentration is also stipulated in several standards, with the US guidelines recommending a range of 0.05–0.2 mg/L. Concerns regarding overexposure to Al are significant, as it is linked to an elevated risk of renal and bone pathologies and may exacerbate the pathogenesis of Alzheimer's disease (Rondeau et al. 2009).

The presence of Pb, As, and Cr in DWDSs raises significant health concerns due to their inherent toxicity and potential for bioaccumulation (Zielina et al. 2022). Prolonged exposure to Pb is particularly detrimental, as it has been associated with severe adverse effects on renal and hematopoietic systems, especially in pediatric populations (Chowdhury et al. 2016; Sun et al. 2021). A prominent case illustrating this issue is the Flint water crisis in Michigan, USA, which stemmed from infrastructural deficiencies and regulatory oversights, resulting in elevated Pb concentrations in the municipal water supply (Roy et al. 2021). This crisis has had profound neurological and developmental implications for children and pregnant women, with recovery efforts to date amounting to nearly half a billion dollars (Edwards & Pruden 2016; Rockey et al. 2021). Likewise, As, even at diminutive levels in drinking water, can result in dermal disorders such as melanosis and is associated with an increased risk of skin cancer, as well as a 6-fold surge in stillbirth rates (Chakraborti et al. 2010; Zhou et al. 2023a). Consequently, regulatory frameworks have established a maximum allowable concentration of As at 0.01 mg/L in potable water (Table 1).

The toxicological profile of Cr is contingent upon its valence state (Moreira et al. 2018; Singh et al. 2022). Long-term exposure to elevated levels of Cr(VI) has been shown to impair kidney function (Liu & Yu 2020). In contrast, Cr(III) is recognized as an essential nutrient, crucial for the regulation of glucose, lipid, and protein metabolism, and it plays a pivotal role in antioxidant defense mechanisms within the human organism (Zielina et al. 2022). Unfortunately, commonly used disinfectants in DWDSs, including chlorine, ozone, and permanganate, can oxidize Cr(III) to its more toxic form, Cr(VI) (Chebeir et al. 2016), thereby exacerbating the potential risks associated with this metal in drinking water.

Given the detrimental impacts of heavy metals on public health, mitigation measures including the implementation of pipe lining, adjustment of water quality, and regulation of flow conditions are frequently employed to reduce metal release in DWDSs. As previously indicated, specific water quality parameters, such as pH, alkalinity, sulfate, and chloride levels, exert a significant influence on the metal release potential in DWDSs. Consequently, moderately adjusting these parameters within the allowable ranges is regarded as a practical approach to the management of water quality stability (Kumar et al. 2016; Li et al. 2020a; Zhang et al. 2022). Likewise, reducing water retention time serves as a vital strategy for mitigating metal release potential. Technologies such as routine flushing and the optimization of pipe network configurations to eliminate stagnant zones are widely implemented within DWDSs (Sunny et al. 2020; Weston et al. 2021). The application of cementitious coatings in metal piping systems is another prevalent mitigation strategy, as this technique facilitates the development of a stable hydroxide layer at the pipe-water interface (Zielina et al. 2022). Moreover, substituting metallic piping with non-corrosive alternatives, such as high-density polyethylene, emerges as a viable solution to the issue of metal release (Okibe et al. 2016; Wei et al. 2020). Nonetheless, it is imperative to consider the potential inclusion of trace metal components within these non-metal plastic piping systems, as discussed in Section 2.1.

Although the factors influencing metal release, the associated health risks, and the mitigation measures in DWDSs have been the focus of academic and industrial attention for decades, several challenges remain to be addressed to enhance water quality management. It is essential to establish a more effective methodology for accurately identifying the sources of heavy metal secondary contamination within DWDSs (Chowdhury et al. 2016; Li et al. 2019), with the ultimate objective of formulating efficacious control strategies to mitigate this contamination. The development of source-tracking methodologies and water quality prediction models constitutes a particularly significant avenue in the pursuit of effective control strategies (Chowdhury et al. 2021).

Microbial presence in DWDSs manifests in two primary forms: suspended in the bulk water or adhered to pipe surfaces, with the latter predominantly forming biofilms (Douterelo et al. 2013). It has been established that over 90% of microbial biomass in DWDSs is encapsulated within biofilms (Fish et al. 2016), attributed to the protective and nutritive environments they provide, which enhance microbial resilience against external perturbations (Wang et al. 2012). Biofilms are complex aggregates composed of diverse microorganisms and their extracellular polymeric substances (EPS), which primarily comprise proteins and carbohydrates (Fish et al. 2017). The developmental processes of biofilms on pipe surfaces are multifaceted, encompassing several stages including initial attachment, colony formation characterized by EPS production, biofilm maturation, detachment, and subsequent recolonization (Liu et al. 2016a; Berne et al. 2018; Sun et al. 2022). In the early stage of biofilm development within DWDSs, notable shifts in microbial community composition and succession of dominant taxa are frequently observed (Douterelo et al. 2014a; Waak et al. 2019). Additionally, a decrease in microbial diversity is particularly pronounced in systems subjected to disinfection procedures (Douterelo et al. 2013).

The microbiome of DWDSs is significantly influenced by a variety of factors, including the quality of the supply water, the materials comprising the pipe system, and the types of disinfectants employed (Learbuch et al. 2022; Oliveira et al. 2024; Li et al. 2024a). A comprehensive investigation, which involved the collection of water samples from 142 full-scale DWDSs across seven countries, revealed a predominance of bacteria in these microbial communities (Dai et al. 2020). The phylum Proteobacteria emerge as the dominant taxon irrespective of geographical location or the specific disinfectant used, aligning with existing research on both regional and laboratory-scale DWDSs (Khu et al. 2023). Regarding pathogenic microorganisms, Legionella spp. showed increased abundance in DWDSs devoid of disinfectants, while Mycobacterium and Pseudomonas were more prevalent in systems that retained disinfectant residuals (Falkinham et al. 2015; Dai et al. 2020; Oliveira et al. 2024). Notably, free-living amoebae were identified as ubiquitous in DWDSs, with their prevalence correlated to various water quality parameters such as organic matter levels, conductivity, pH, and temperature (Delafont et al. 2016).

Microbial activities may result in the secretion of acidic metabolites, which facilitate the corrosion of pipe substances and the subsequent release of contaminants (Douterelo et al. 2018). Additionally, adhered microorganisms or biofilms are susceptible to detachment due to external interventions, such as an elevation in flow shear stress (Douterelo et al. 2019). This process is linked to an increase in water turbidity and alterations in the sensory attributes of water, including odor and taste (Douterelo et al. 2013, 2016; Adams et al. 2022). More importantly, however, microbial regrowth and biofilm development pose significant hygienic concerns, primarily due to their role in the propagation of pathogenic microorganisms (Falkinham et al. 2015; Gomez et al. 2015).

Coliform bacteria, inclusive of species such as Escherichia spp., Enterobacter spp., and Citrobacter spp., are recognized as pathogenic entities capable of eliciting acute gastroenteritis (Mirsepasi et al. 2019; Gilliland et al. 2024). Consequently, their detection in drinking water contravenes established regulatory standards (Table 1). An emblematic incident occurred in northern Finland in 1998, wherein routine maintenance activities resulted in coliform contamination within DWDSs, leading to over 200 reported cases of acute gastroenteritis (Kuusi et al. 2005; Thom et al. 2022). Moreover, the protozoan Cryptosporidium is recognized as a significant etiological agent of diarrhea (Kotloff et al. 2013), particularly in developing nations, where it contributes substantially to diarrheal morbidity among children under 5 years of age (Troeger et al. 2018). A review of global outbreak data during the 2011–2016 period revealed that Cryptosporidium spp. was the most prevalent causative agent, accounting for 63% of reported cases, followed by Giardia spp. (Efstratiou et al. 2017). Therefore, the occurrence of these two species in drinking water is strictly proscribed by regulatory agencies in countries such as China, the United States, and South Africa (Table 1). Research conducted in the United States indicated that waterborne infections are implicated in over 40,000 hospitalizations annually, resulting in a significant economic burden, estimated at $970 million per year (Collier et al. 2012).

For decades, culture-based methodologies have been the primary approach for assessing microbiological water quality (Farhat et al. 2020; Feng et al. 2021). This traditional technique involves the inoculation of water specimens onto agar plates, followed by an incubation period that allows for microbial growth, culminating in the identification of species present (Chowdhury 2012). Microbes-related water quality indicators, such as total bacterial count, total coliforms, and heterotrophic plate count (Table 1), are determined utilizing this culture-dependent technique (Palma et al. 2024). Nonetheless, despite its practicality, the culture process is time-consuming and may not accurately represent the microbial diversity present in aquatic environments (Douterelo et al. 2020).

To address these limitations, molecular biology techniques, particularly those centered on PCR, have markedly revolutionized the detection of microorganisms in drinking water (Nnadozie & Odume 2019). PCR facilitates the rapid amplification and identification of microbial DNA or RNA, thereby enabling the precise characterization of specific microorganisms in water samples (Douterelo et al. 2014b). Among the various PCR-based methodologies, multiplex-PCR and quantitative real-time PCR (qPCR) are recognized for their applicability in drinking water analysis (Kubista et al. 2006). Multiplex-PCR employs multiple oligonucleotide probes, allowing for simultaneous identification of a diverse range of microorganisms. Conversely, qPCR quantifies the accumulation of PCR products during the exponential phase of the reaction through the use of fluorescent reporters. This capability enables qPCR to not only detect the presence of microorganisms but also quantify their abundance in environmental samples, thus providing a more comprehensive assessment of water quality (Perkel 2013; Zhang et al. 2024a).

Recent advancements in next-generation sequencing technologies, particularly 16S rRNA gene sequencing and metagenomic sequencing, have substantially enhanced the detection and characterization of microbial communities in drinking water (Feng et al. 2021). These technologies facilitate high-throughput sequencing of microbial DNA, thereby enabling a comprehensive analysis of microbial community composition present in water samples (Caporaso et al. 2012). Moreover, the incorporation of bioinformatics tools has proven essential for the precise identification of microbial species and the elucidation of community structures in drinking water (Feng et al. 2021).

The application of these sophisticated detection methods, coupled with advances in socioeconomic factors and an elevated concern for public health, has precipitated a marked increase in research focused on microbiological water quality (Figure 1(c)). Specifically, to understand how microbial communities respond to chlorine treatment in DWDSs, qPCR analysis has been employed to target gene markers associated with nitrifying populations, revealing a predominance of ammonia-oxidizing bacteria over ammonia-oxidizing archaea (Wang et al. 2014). Moreover, research by Hull et al. has indicated the presence and activity of Actinobacteria spp., Mycobacterium spp., and Legionella spp. in DWDSs through the application of 16S RNA analysis, highlighting their implications for water safety (Hull et al. 2017). Additionally, the synergistic use of 16S rRNA gene sequencing and qPCR has also been instrumental in demonstrating shifts in the microbiome and the potential for pathogen regrowth within secondary water supply systems (Li et al. 2024a).

Primary disinfection serves as an essential strategy for mitigating microbial regrowth and pathogen propagation within DWDSs. The predominant disinfectants employed in water treatment processes include chlorine, chloramine, and chlorine dioxide, which are typically administered in excess to ensure the maintenance of effective residual concentrations throughout the distribution network (Lu et al. 2023; Oliveira et al. 2024). These disinfectants function by inhibiting microbial proliferation through various mechanisms, including oxidative stress and the destruction of protein structures (Dai et al. 2020). However, the challenge of maintaining an optimal disinfectant concentration arises due to the potential depletion of disinfectants at the termini of expansive networks, while elevated residual levels in other areas can precipitate the formation of DBPs (Liu et al. 2016a). To enhance microbial control, the implementation of secondary disinfection methods, such as UV disinfection, is employed as an additional safeguard (Bairoliya et al. 2022; Li et al. 2024a). The UV disinfection technique is recognized for its effectiveness in damaging the DNA and RNA structures of microorganisms, thereby preventing their replication and transcription (Wang et al. 2023).

The conditions that promote microbial regrowth in DWDSs are significantly influenced by factors such as optimal water temperature, anoxic–oxic conditions, and the availability of organic matter originating from source water (Li et al. 2018). Accordingly, strategies such as biofiltration and membrane filtration have been adopted to diminish the levels of assimilable organic carbon in drinking water prior to its introduction into DWDSs (Liu et al. 2016a). This approach effectively reduces nutrient levels, which, in turn, inhibits microbial regrowth. Furthermore, the inner surfaces of ageing pipes can serve as additional niches for microbial colonization (Ammar et al. 2015; Chinnaraj et al. 2021), underscoring the necessity for judicious selection of pipeline materials and proactive measures to mitigate pipe corrosion. Additionally, the meticulous regulation of water retention time within DWDSs is critical, as extended periods of stagnation are conducive to microbial regrowth and pose an increased risk of pathogen proliferation (Li & Mitch 2018; Li et al. 2024a).

Public concerns regarding pathogens and the advancement of detection methodologies have catalyzed heightened research interest in microorganisms within DWDSs. Recently developed adenosine triphosphate luminescence-based assays represent a novel approach for the rapid assessment of biostability in drinking water (Zhang et al. 2019a). Continued innovation in detection techniques and their applications is anticipated to garner significant interest from both industrial and academic sectors. While changes in microbial community composition have been documented within DWDSs (Li et al. 2024a), further investigative efforts are necessary to elucidate how alterations in these communities throughout the water distribution process may influence public health outcomes.

NOM serves as the principal precursor in the formation of DBPs, comprising a complex amalgamation of organic molecules that occur in particulate or dissolved forms (He et al. 2020; Wu et al. 2020). Water sources significantly influence the composition of NOM, resulting in its various molecular constitutions, structural characteristics, and functional groups. NOM is typically categorized into hydrophilic components, such as carbohydrates, and hydrophobic components, like humic acids. Evidence suggests that the potential for DBP formation exhibits regional variability, closely correlated with the specific characteristics of NOM in the source water (Li et al. 2014). Notably, the hydrophobic portion of NOM, particularly those of higher molecular weight, has been identified as a critical factor in the formation of THMs and HAAs (Hua & Reckhow 2007a), whereas NOM of lower molecular weight emerges as vital precursors for the genesis of HANs (Wang et al. 2021b).

Halogen ions such as Br⁻ and I⁻ are prevalent in natural water systems, especially within coastal and desalinated waters. Their presence critically influences the generation of brominated and iodinated DBPs (Yang et al. 2014; Ankoliya et al. 2023; Khan et al. 2024). Research indicates that a higher initial concentration of bromide ions in source water correlates with an increased formation of bromate following ozonation (Aranda-Rodriguez et al. 2017). The metric of total organic halogen (TOX) serves as a comprehensive indicator for assessing the integration of halogen atoms into organic compounds, which can be further delineated into specific categories, including total organic chlorine, total organic bromine, and total organic iodine (Yang et al. 2014). The relative concentrations of Br⁻ and I⁻ have been demonstrated to be associated not only with the overall formation of TOX but also with the distribution of various TOX species (Kristiana et al. 2009). Besides, further investigations have established a correlation between increased bromide concentrations and elevated molar yields of THMs and HAAs (Liu et al. 2018). The formation of iodo-trihalomethanes, especially iodoform, can be substantially mitigated through enhanced exposure to free chlorine and an elevated Br⁻/I⁻ ratio (Criquet et al. 2012).

A diverse range of disinfection methodologies is employed in DWDSs, including chlorine, chloramine, chlorine dioxide, hydrogen peroxide, ozone, UV irradiation, as well as alternative agents such as silver ions and nanoparticles (Mian et al. 2018; Radwan et al. 2021; Feng et al. 2022). Among these, chlorine, chloramine, and chlorine dioxide are the most extensively utilized, which also serve as principal agents for secondary disinfection (Mian et al. 2018; Knocke et al. 2022; Oliveira et al. 2024). The selection of a particular disinfectant can result in significant variances in both the profile and concentration of DBPs generated. Free chlorine, for instance, is implicated in the formation of elevated levels of THMs, HAAs, and TOX compared to its counterparts, chloramines, and chlorine dioxide (Hua & Reckhow 2007a, 2007b). However, while the application of monochloramine disinfection substantially decreases the yields of THMs and HAAs, it conspicuously enhances the production of N-nitrosodimethylamine (NDMA) (Ma et al. 2016). Additionally, the use of chlorine dioxide generates inorganic by-products, such as chlorite, during the disinfection process, posing potential health risks (Zhou et al. 2016). Moreover, it is critical to realize that the secondary disinfection strategy implemented in DWDSs, which aims to mitigate microbial regrowth, as mentioned in Section 3, can lead not only to the generation of new DBPs but also to the transformation of existing DBPs (Huang et al. 2017a).

In addition to the precursors and disinfectants directly involved in reactions, factors such as pipe materials and physico-chemical properties of water also significantly affect the formation of DBPs (Ye et al. 2009; Hu et al. 2024a). Pipe materials, for instance, alter the attenuation process of disinfectants, thereby impacting the genesis of DBPs. Specifically, polyethylene pipes have been identified as the largest contributors to THMs and HAAs (Kali et al. 2021). Furthermore, there is a discernible escalation in the formation of THMs and HAAs as temperature increases from 15 to 30 °C. Conversely, the concentrations of these by-products decrease upon heating or boiling of tap water (Wu et al. 2001; Ye et al. 2009). Besides, elevated pH levels in water correlate with increased THM production, whereas an inverse relationship is observed for HAAs (Siddique et al. 2023).

In 1974, Rook et al. first identified trichloromethane (TCM) and other THMs in chlorinated drinking water. Subsequently, the US National Cancer Institute published a study regarding the carcinogenicity of TCM (Rook 1972; Bellar et al. 1974). Since then, significant attention has been placed on DBPs and their associated health risks. Over the past few decades, researchers worldwide have documented various toxicological effects of DBPs, including cytotoxicity, genotoxicity, mutagenicity, teratogenicity, carcinogenicity, and endocrine disruption (Wang et al. 2018; Mazhar et al. 2020).

Epidemiological studies have demonstrated the health risks associated with consuming chlorinated tap water containing increased concentrations of THMs, notably an elevated potential in bladder cancer (Shi et al. 2024). For instance, an aggregated analysis of European populations revealed a significantly higher risk of bladder cancer in men with greater THMs exposure (>50 μg/L) compared to those exposed to lower or negligible levels (≤5 μg/L) (Costet et al. 2011). In 1979, the US EPA established regulations to limit THMs in finished drinking water to a maximum of 100 μg/L (US EPA 1979). Subsequently, another group of DBPs known as HAAs were discovered in drinking water at levels comparable to THMs, posing additional reproductive and developmental health risks (Zhang & Minear 2002). A retrospective cohort study found that women exposed to mean dibromoacetic acid (DBAA) concentrations of ≥5 μg/L during late pregnancy had an increased risk of delivering low-birth-weight infants at term, compared to women exposed to concentrations of <4 μg/L (Hinckley et al. 2005). In 1998, the US EPA implemented the Stage 1 Disinfectant/DBP (D/DBP) Rule, which reduced the allowable concentration of THMs to 80 μg/L and regulated five HAAs (namely, mono-, di-, trichloroacetic acid and mono-, dibromoacetic acid), bromates, and chlorites for the first time (US EPA 1998).

Currently, approximately 20 DBPs, including THMs and HAAs, have been subjected to regulation by global authorities to safeguard the safety of drinking water and reclaimed water, as shown in Table 1. The United States and the European Union have formulated guidelines governing the total concentration of THMs in drinking water (US EPA 2015; European Union, The European Parliament and the Council of the European Union 2020), while the WHO, alongside nations such as China, Japan, and South Africa, has specified limits for individual THM species (SABS T.S.A.B.o.S. 2015; WHO Team 2017; SAC T.S.A.o.t.P.s.R.o.C 2022). Notably, the regulation of nitrogenous DBPs, specifically HANs and NDMA, is exclusively encompassed in the WHO guidelines (WHO Team 2017). By contrast, the European Union and South Africa impose comparatively less stringent regulatory measures regarding DBP formation, with South Africa only regulating THMs and the European Union solely regulating total THMs and bromate (SABS T.S.A.B.o.S. 2015;  European Union, The European Parliament and the Council of the European Union 2020).

Nevertheless, it has been estimated that these regulated THMs and HAAs may account for less than 5% of the total cytotoxicity of organic extracts derived from tap water (Plewa et al. 2017). In vitro evaluations of genotoxicity and cytotoxicity have revealed that nitrogenous DBPs, such as HANs and HNMs, exhibit significantly greater toxicity compared to the regulated THMs and HAAs (Shah & Mitch 2012). Moreover, NDMA, a prominent nitrogenous DBP, is reported to be implicated in a lifetime cancer risk of 10⁻⁻⁶ at a drinking water concentration of merely 0.7 ng/L (Grasso 1979; IARC 1979; National Academies of Sciences Division on Earth Life Studies Board on Environmental Studies & Committee to Review Advances Made to the IRIS Process 2018). This substantial toxicity associated with unregulated DBPs raises significant concerns for global regulatory agencies, despite their typically low concentrations within DWDSs (Wang et al. 2018).

To minimize the detrimental impacts on human health, it is imperative to implement practical interventions at various stages of water distribution (Pandian et al. 2022; Peng et al. 2023; Qiu et al. 2023). These interventions, aimed at regulating the concentration of DBPs, can be categorized into three primary strategies: source control, process control, and end control.

First of all, precursors could be appropriately removed during water treatment processes to reduce the potential formation of DBPs, referred to as the source control. Among these precursors, NOM and halogen ions are considered the primary targets for reduction. The effective removal of NOM from source waters can be accomplished through several treatment processes, including coagulation, adsorption, membrane filtration, photocatalysis, and pre-oxidation (Liang & Singer 2003; Boyer & Singer 2005; Selcuk et al. 2007). Besides, the concentration of halogen ions in source water can be significantly diminished through preventative measures aimed at minimizing seawater intrusion and regulating the discharge of industrial wastewater into water sources (Watson et al. 2012; Islam et al. 2023; Trejo-Albuerne & Canul-Macario 2024). Additionally, advanced treatment techniques, such as ion exchange, have been demonstrated to be effective in removing inorganic ions and certain charged organic compounds (Winter et al. 2018; Caltran et al. 2020; Yang et al. 2023), thereby contributing to the reduction of DBP formation potential.

The process control involves implementing strategies to minimize the formation of DBPs and alter the types of DBPs generated. This is achieved by optimizing disinfection parameters, which encompass the selection of disinfectants, determination of appropriate dosages, and adjustment of reaction times (Kali et al. 2021; Siddique et al. 2023). As elucidated in Section 4.1, chloramine exhibits lower reactivity compared to chlorine, resulting in a reduced potential for the overall formation of DBPs (Mian et al. 2018). However, the application of chloramine is associated with increased production of nitrogenous DBPs (Huang et al. 2017b; Oliveira et al. 2024) and augments the formation of iodinated DBPs in the presence of iodide (Hua & Reckhow 2007b). This underscores the necessity of selecting disinfection methods based on the inherent characteristics of diverse water sources. Furthermore, a decrease in disinfectant concentration can substantially reduce DBP formation (Ma et al. 2016), on the premise of ensuring the effective inhibition of microbial regrowth (Ye et al. 2009). Moreover, in circumstances where secondary disinfection is adopted, reducing the ratio of primary to secondary chlorination dosage, as well as postponing the timing of secondary chlorine addition, can effectively reduce DBP generation while maintaining adequate chlorine residual, provided that the total chlorine dosage remains consistent (Ding et al. 2019; Zhu et al. 2022).

Following the generation of DBPs in DWDSs, various measures can also be implemented for the partial removal of these contaminants to mitigate their adverse health impacts on consumers, referred to as the end control. For example, the application of a nanocomposite comprising zero-valent iron and activated carbon has demonstrated efficacy in the removal of THMs (Xiao et al. 2015). Additionally, advanced membrane filtration techniques, such as reverse osmosis, ultrafiltration, and nanofiltration, can be utilized to intercept a significant proportion of DBPs (Li et al. 2024b). Research conducted by Liu et al. showed that nanofiltration achieved an average removal efficiency of 78.2% for THMs, HAAs, and various typical nitrogenous DBPs (Liu et al. 2020). Boiling tap water is also an effective household method for reducing DBPs and lowering human exposure (Ye et al. 2009; Li et al. 2023). Previous studies have demonstrated that routine home cooking can remove approximately 90% of THMs, 80% of HANs, and 15% of HAAs (Ma et al. 2024). Furthermore, Li et al. found that boiling exhibits high removal efficiencies for volatile DBPs such as THMs, halo-acetaldehydes, HANs, and haloketones, with corresponding removal rates of 90.5, 100, 100, and 100%, respectively (Li et al. 2024b). However, these post-formation measures are either energy-intensive or inconvenient for consumers. It is preferable to implement appropriate management strategies aimed at controlling the formation of DBPs prior to and during water distribution.

Despite the implementation of various control strategies, additional considerations are necessary to effectively address the issue of DBPs in DWDSs. Source water quality demonstrates considerable variability, influenced by both natural events and anthropogenic activities. This variability significantly affects the composition of precursors that contribute to the formation of DBPs (Xiao et al. 2024). The accurate prediction of the spatial and temporal variations in source water quality is essential for developing targeted preventive measures aimed at mitigating DBP formation. Furthermore, the continuous identification of emerging DBPs poses challenges to conventional water treatment methodologies (Ding et al. 2019). Therefore, sustained innovation in advanced treatment techniques, alongside the implementation of composite treatment systems and the effective control of treatment costs, is likely to garner interest from both academic researchers and industrial practitioners.

The presence of heavy metals in DWDSs significantly influences the structure, composition, and functional dynamics of microbial communities (Jia et al. 2022). Previous research has demonstrated that elevated levels of metal stress exerts a selective effect on metal-sensitive microorganisms, leading to a decrease in microbial diversity while concurrently increasing the relative abundance of metal-resistant organisms (Wang et al. 2021c; Li et al. 2024a). Nonetheless, specific heavy metals, such as Cr, Cu, and Mn, which serve as essential micronutrients, play crucial roles within microbial flora, yet their elevated levels disrupt the homeostasis of microbial communities by selectively inhibiting the growth of particular microorganisms (Wang et al. 2021c; Stefan et al. 2023). From another perspective, the presence of excess heavy metals in water supply systems can alter the dynamics of pathogenic bacteria. For example, Pd has been shown to affect the growth and viability of Escherichia coli, while Cd may inhibit the proliferation of Salmonella by compromising the integrity and functionality of its cell membrane, thereby impacting its pathogenic potential (Liu et al. 2016b; Rong et al. 2021).

Besides, the release of heavy metals, such as Pd, Fe, and Mn, can react with disinfectants, diminishing their efficacy and DBP generation while facilitating the regrowth of bacteria and other pathogens (Wang & Zhu 2010). These heavy metals can also form complexes with NOM, thereby influencing the production of DBPs (Hu et al. 2024b). Furthermore, the presence of heavy metals alters water chemistry, including parameters such as pH and redox potential. Such changes can significantly affect the reactivity of disinfectants and the stability of precursor compounds, ultimately influencing the production of DBPs. For instance, the catalytic oxidation produced by Fe and Mn can facilitate the transformation of bromide to bromate (Yan et al. 2020).

The metabolic activities of microorganisms, especially sulfate-reducing bacteria (SRB) and iron-oxidizing bacteria (IOB), play a significant role in exacerbating metal corrosion within DWDSs (Sun et al. 2017). These microorganisms engage in electron exchange with pipeline materials through extracellular electron transfer mechanisms, thereby expediting the corrosion process. Specifically, SRB derive electrons by oxidizing organic matter in anaerobic environments and subsequently transferring them to sulfates, leading to the reduction of sulfates to hydrogen sulfide (H₂S). This metabolic process not only provides essential energy for microbial growth but also contributes to localized metal corrosion (Kurajica et al. 2022).

Furthermore, the development of microbial resistance to disinfectants can lead to an augmented dosage of these agents, which, in turn, can impact the formation of DBPs (Zhang et al. 2019c; Zhao et al. 2023a). The careful selection and quantification of disinfectants are crucial considerations in this regard (Zhang & Lu 2021). Microbial metabolism, in conjunction with the presence of organic matter, is also linked to the production of DBPs (Zeng et al. 2021). Additionally, protozoan organisms such as ciliates can impede the growth of other corrosive microorganisms through predation or the synthesis of antibacterial substances, thereby influencing both the overall corrosion dynamics within the pipeline network and the generation of DBPs (Rakic et al. 2012; Rilstone et al. 2021; Jia et al. 2022; Shi et al. 2022; Dong et al. 2023).

The formation of DBPs in DWDSs has been found to have significant implications for the structure and diversity of microbial communities. Primarily, DBPs cause changes in environmental parameters such as dissolved oxygen and organic matter concentrations, thereby influencing microbial growth and metabolic activities (Zhou et al. 2023b). Additionally, the generation of DBPs can affect the diversity of microbial communities in DWDSs (Zhang et al. 2024b), leading to a decrease in specific functional microbial populations such as SRB, IOB, and nitrifying bacteria (Andra et al. 2014; Qi et al. 2022).

Moreover, DBPs can alter the activity of microorganisms involved in the transformation of heavy metals in natural environments, consequently affecting the release of these metals. For instance, chloroform has been identified as a substance that negatively affects the metabolic activities of microorganisms, inhibiting their ability to oxidize manganese and iron. This inhibition may lead to a reduction in the transformation of insoluble heavy metal oxides into soluble forms, potentially decreasing the release and mobility of heavy metals in aquatic environments (Wu et al. 2017; Han et al. 2020).

This study presents a comprehensive analysis of the three primary substances contributing to secondary contamination in DWDSs, namely metal release, microbial regrowth, and DBP formation. It thoroughly elucidates their sources, underlying mechanisms, health risks, and potential mitigation strategies. Additionally, the article delves into the interrelationships among these contaminants.

The release of heavy metals is intricately related to the corrosion processes of metallic components within distribution pipelines. Key factors such as pipe materials, water quality, and hydraulic conditions in DWDSs significantly affect the rate and extent of metal release. Consequently, mitigation measures often involve the moderate adjustment of these parameters to minimize metal leaching. The proliferation of pathogenic microorganisms could occur with the depletion of disinfectant residuals in DWDSs, potentially causing hygienic issues. Improving disinfection strategies can effectively mitigate the hazards associated with microbial regrowth. Additionally, the adoption of disinfection strategies plays a critical role in the formation of DBPs, which arise from reactions between residual disinfectants and precursors. The generation of DBPs can be effectively reduced through strategies including source control, process control, and end control. Moreover, a comparative analysis of drinking water quality standards reveals a general consistency in guidelines across various entities; albeit, specific organizations may emphasize distinct parameters.

The construction of a multi-scale water supply network represents a significant hydraulic engineering endeavor in China. Therefore, emphasizing secondary contamination control within DWDSs has emerged as a critical concern to be addressed (Guo et al. 2023; Zhao et al. 2023b). Further exploration is warranted regarding the implications of utilizing multiple water sources on the induction of metal release, pathogenic bacteria propagation, and DBP formation within DWDSs (Daly & Harris 2022; Lancioni et al. 2023). Moreover, artificial intelligence (AI) technologies, particularly those based on machine learning (ML), have been increasingly applied to enhance water quality control within DWDSs. For instance, ML algorithms have been employed to assess the concentration and distribution of heavy metals in drinking water, thereby improving the accuracy and effectiveness of contaminant removal predictions (Oh et al. 2021). Additionally, AI facilitates the monitoring of contamination levels of Giardia spp. and the evaluation of biofilm development within DWDSs (Ligda et al. 2020). It can also predict DBP formation based on operational conditions while elucidating the underlying mechanisms (Li et al. 2021). However, the application of AI technology in water quality management remains largely limited to specific cases and encounters challenges, such as in the availability of sufficient data. Continuous advancements in AI technology raise the anticipation for a more comprehensive application in the management of secondary contamination within DWDSs.

The work is supported by the Shenzhen Key Laboratory of Advanced Technology for Marine Ecology (ZDSYS20230626091459009) and the Open Research Fund Program of the State Key Laboratory of Hydroscience and Engineering (sklhse-2023-A-03).

All relevant data are included in the paper or its Supplementary Information.

The authors declare there is no conflict.