Biological stability in drinking water distribution systems: A novel approach for systematic microbial water quality monitoring

Abstract

Challenges to achieve biological stability in drinking water distribution systems Drinking water is distributed from the treatment facility to consumers through extended man-made piping systems. The World Health Organization drinking water guidelines (2006) stated that “Water entering the distribution system must be microbiologically safe and ideally should also be biologically stable”. The biological stability criterion refers to maintaining the microbial drinking water quality in time and distance from the point of drinking water production up to the point of consumption. However, uncontrolled growth of indigenous bacteria during water transport can result in the deterioration of aesthetic aspects of water, such as taste, colour, and odour, in exceeding of guideline values, and/or in technical problems. Controlling bacterial growth in piping systems and premise plumbings is very challenging (Chapter 2), and changes in drinking water microbial characteristics are often measured in networks distributing water with or without residual disinfectant such as chlorine, monochloramine or chlorine dioxide. In the Netherlands, drinking water is distributed without detectable residual disinfectant. Quantitative and qualitative knowledge on the indigenous bacterial communities and microbiological processes taking place during drinking water distribution is limited and in-depth investigations are required. New opportunities with novel analytical methods One reason for the lack of knowledge on bacterial growth controlling factors in drinking water distribution systems is that methods for characterizing drinking water bacterial communities are still relying heavily on culture-based techniques such as plate counts, developed more than 130 years ago. The conventional cultivation-based methods have major limitations: only a minute fraction (<0.1 %) of drinking water bacteria is detected, which is not representative of the drinking water bacterial community, and results are obtained only after a minimum of two days. During the last decade, new cultivation-independent techniques have emerged for the characterization of water bacterial communities. Among them, flow cytometry (FCM) enables the rapid detection and counting of all bacterial cells in water (within 15 minutes), and provides information on bacterial cell properties such as viability. Besides, high-throughput sequencing methods (e.g. 454-pyrosequencing or Illumina) enable characterization of the total bacterial community composition and structure at various taxonomic levels. FCM and high-throughput sequencing methods offer new perspectives for better and faster water microbiology monitoring and for increased understanding of the complex bacterial dynamics occurring during drinking water distribution up to the point of consumption (Chapter 2). Method development The primary goal of this study was to develop a methodological approach, based on advanced analytical methods, for the assessment of biological stability in drinking water distribution systems. A standardized, rapid and simple FCM method was shown to be highly reproducible and sensitive for total and intact bacterial cell enumeration. Changes in bacterial community characteristics could be detected based on bacterial cell concentrations and FCM fluorescence fingerprints, which are characteristic of each water sample (Chapter 3). Changes in fluorescence fingerprints were proven to be a rapid indication for changes in bacterial community composition, by comparing FCM and 16S rRNA gene pyrosequencing data obtained from the same drinking water samples. Combining the two methods enabled both quantitative and qualitative characterization of water bacterial communities (Chapter 4). An integrated approach was proposed for the assessment of bacterial growth-controlling factors in drinking water and for the evaluation of the impact of full-scale distribution conditions on bacterial growth extent. The approach combines (i) characterization of autochthonous bacterial communities in water samples collected at several locations in full-scale drinking water distribution systems, using FCM and high-throughput sequencing methods, (ii) comparison of changes in bacterial abundance recorded during water distribution and during controlled laboratory bacterial growth tests, and (iii) stepwise assessment of bacterial growth limitations in drinking water using straightforward bacterial growth potential tests (Chapter 5). Application of developed methodological approach to a full-scale drinking water system The developed methodological approach was applied to a Dutch full-scale drinking water treatment and distribution system operated without detectable disinfectant residual. Spatial and temporal variations were studied on short-term (hour, day, week) and long-term (seasonal) time-scales, and bacterial growth-limiting factors were investigated. Bacterial growth in the produced drinking water was limited both by organic carbon and inorganic nutrients (Chapter 5). Large seasonal variations in bacterial cell concentrations were recorded at the treatment effluent, which were congruent with water temperature fluctuations. Changes in bacterial community characteristics in the distribution system were minor compared to temporal variations in the treatment effluent (Chapter 6). However, all studies univocally showed that changes in bacterial community abundance, viability and/or community composition occurred during water distribution in the well-maintained network (Chapters 4, 5, 6 and 7). Changes were not detected with conventional bacterial detection methods. In-depth analysis of bacterial community composition in water samples, using pyrosequencing, showed that the core bacterial community did not change during water distribution, whereas high dynamicity was found in rare taxa (Chapter 7). Different bacterial cell concentrations were measured in the full-scale system and after incubation of the same water under controlled conditions, highlighting the effect of distribution conditions (e.g. temperature, pipe material, residence time) on drinking water microbial quality (Chapter 5). The results suggest that the extent of bacterial growth at one specifically studied location in the distribution system was not determined by the concentration of assimilable organic carbon in the treatment effluent. Likely not only one single parameter can be considered as controlling factor of microbial growth in drinking water distribution systems (Chapter 6). Recommendations From these observations, it is recommended to study microbial dynamics in drinking water distribution systems using a combination of controlled laboratory growth potential tests and in-situ characterization of the drinking water bacterial communities in the distribution network, which includes both spatial and temporal investigations. Applying such an approach to individual systems would provide better understanding of microbial dynamics during drinking water production and distribution, enabling (i) rapid and sensitive drinking water monitoring, (ii) effective corrective and maintenance actions and (iii) funded decisions for the optimization of water treatment production and/or distribution conditions to control bacterial growth in drinking water distribution systems. In this regard, the recent emergence of on-line flow cytometers will promote flow cytometry as an ideal monitoring method, for the rapid detection of system failure and targeted maintenance management.