A systematic approach is presented for the assessment of (i) bacterial growth-controlling factors in drinking water and (ii) the impact of distribution conditions on the extent of bacterial growth in full-scale distribution systems. The approach combines (i) quantification of changes in autochthonous bacterial cell concentrations in full-scale distribution systems with (ii) laboratory-scale batch bacterial growth potential tests of drinking water samples under defined conditions. The growth potential tests were done by direct incubation of water samples, without modification of the original bacterial flora, and with flow cytometric quantification of bacterial growth. This method was shown to be reproducible (ca. 4% relative standard deviation) and sensitive (detection of bacterial growth down to 5 µg L−1 of added assimilable organic carbon). The principle of step-wise assessment of bacterial growth-controlling factors was demonstrated on bottled water, shown to be primarily carbon limited at 133 (±18) × 103 cells mL−1 and secondarily limited by inorganic nutrients at 5,500 (±1,700) × 103 cells mL−1. Analysis of the effluent of a Dutch full-scale drinking water treatment plant showed (1) bacterial growth inhibition as a result of end-point chlorination, (2) organic carbon limitation at 192 (±72) × 103 cells mL−1 and (3) inorganic nutrient limitation at 375 (±31) × 103 cells mL−1. Significantly lower net bacterial growth was measured in the corresponding full-scale distribution system (176 (±25) × 103 cells mL−1) than in the laboratory-scale growth potential test of the same water (294 (±35) × 103 cells mL−1), highlighting the influence of distribution on bacterial growth. The systematic approach described herein provides quantitative information on the effect of drinking water properties and distribution system conditions on biological stability, which can assist water utilities in decision-making on treatment or distribution system improvements to better control bacterial growth during water distribution.

ABBREVIATIONS

  • AOC

    assimilable organic carbon

  • BDOC

    biodegradable dissolved organic carbon

  • DWDS

    drinking water distribution system

  • FCM

    flow cytometry /flow cytometer

  • OTU

    operational taxonomic unit

  • TCC

    total cell concentration

  • WQ

    water quality

  • WTP

    water treatment plant

INTRODUCTION

Water utilities aim to distribute biologically stable drinking water, i.e. to operate drinking water distribution systems (DWDS) in which no adverse bacterial growth occurs. Uncontrolled growth of bacteria in distribution pipelines can lead to aesthetic, operational and even hygiene-related problems (van der Kooij 2003). Assessment of the bacterial growth potential of drinking water and identification of microbial growth-controlling factors are therefore required to enable better control of bacterial growth in DWDS.

The bacterial growth potential of drinking water is an experimental quantification of the extent of bacterial growth that water can promote under defined laboratory conditions (e.g. temperature), and depends on (i) all the growth-promoting/limiting compounds (e.g. organic and inorganic nutrient composition), (ii) the presence of growth-inhibiting substances (e.g. residual disinfectant) and (iii) the autochthonous bacterial community. Organic carbon is often the main growth-controlling compound in drinking water (Joret et al. 1991; Escobar et al. 2001). Hence, conventional bacterial growth assays, such as the assimilable organic carbon (AOC) and biodegradable dissolved organic carbon (BDOC) methods either assume organic carbon limitation (van der Kooij et al. 1982; Servais et al. 1989) or incorporate excess inorganic nutrients to ensure organic carbon limitation (Miettinen et al. 1999). However, previous studies have revealed phosphate to be the growth-limiting compound in several drinking waters (Miettinen et al. 1997; Sathasivan et al. 1997; Juhna & Rubulis 2004).

The current use of conventional bacterial growth assays has three shortcomings: (1) although growth potential methods are available, systematic investigation of specific bacterial growth limitations in drinking water are almost never performed, which provides limited understanding of water characteristics; (2) conventional growth potential methods are essentially used to estimate the concentration of the growth limiting compounds (e.g. AOC) rather than to estimate the bacterial growth; and (3) water transport in full-scale DWDS can alter the water characteristics and subsequently affect bacterial growth. For example, DWDS conditions can promote disinfectant residual decay (LeChevallier et al. 1996; Nescerecka et al. 2014), and the release of organic or inorganic compounds into the water from pipe walls and sediments (Niquette et al. 2000; Zacheus et al. 2001; Lehtola et al. 2004; Bucheli-Witschel et al. 2012) or from the mixing of waters differing in composition (Niquette et al. 2001). Bacteria present in large numbers in pipe and reservoir biofilms and sediments can also compete with bulk water bacteria for available biodegradable nutrients and be released into the drinking water. In addition, varying water residence times and temperatures (Kerneis et al. 1995; Uhl & Schaule 2004; Hammes et al. 2010) are likely to modify bacterial growth kinetics in comparison with controlled and standardized batch conditions applied for laboratory growth tests. Consequently, controlled laboratory-scale growth potential tests are not sufficient to predict bacterial growth actually occurring during water distribution. In situ water sampling is needed to provide a picture of the extent of bacterial growth occurring in full-scale DWDS (Niquette et al. 2001; Lautenschlager et al. 2013; Nescerecka et al. 2014).

To address these issues we propose to integrate full-scale DWDS investigations and step-wise laboratory-scale growth potential tests into one systematic approach. For the laboratory tests we adapted the approach used by Gillespie et al. (2014), which consists of the direct incubation of drinking water samples without modifying the autochthonous community, and combined it with a step-wise assessment of bacterial growth-controlling factors. In addition, the bacterial growth behaviour of the same drinking water in well-defined laboratory conditions and the full-scale distribution network was directly compared. The objective of this study was to evaluate the integrated approach for (i) quantification of bacterial growth in DWDS, (ii) identification of bacterial growth-controlling factors in drinking water and (iii) evaluation of the effect of full-scale drinking water distribution on the extent of bacterial growth.

Description of the systematic approach

Figure 1 provides an overview of the proposed three-step approach for determining bacterial growth-controlling factors in drinking water and evaluating the effect of full-scale distribution.
Figure 1

Schematic overview of the systematic approach for the evaluation of bacterial growth-controlling factors during drinking water distribution. WQ-1: water quality at the treatment outlet; WQ-2: water quality in the distribution system; WQ-2b,c: water quality after bacterial growth potential tests under controlled laboratory conditions. WQx are compared.

Figure 1

Schematic overview of the systematic approach for the evaluation of bacterial growth-controlling factors during drinking water distribution. WQ-1: water quality at the treatment outlet; WQ-2: water quality in the distribution system; WQ-2b,c: water quality after bacterial growth potential tests under controlled laboratory conditions. WQx are compared.

Quantification of changes in bacterial cell concentrations during full-scale distribution

As a first step, water samples are collected at the drinking water treatment plant effluent, and at full-scale distribution system (DWDS) sampling points with various estimated water residence times, for the quantification of suspended bacterial cells in the DWDS. The total and intact planktonic bacterial cell concentrations are accurately quantified using flow cytometry (FCM) (Lautenschlager et al. 2013; Prest et al. 2013). We argue herein that the difference in bacterial cell concentrations between the treatment plant and the network locations reflects the extent of bacterial growth, and thus biological stability, in the DWDS (Niquette et al. 2001; Lautenschlager et al. 2013; Prest et al. 2014).

Comparison of bacterial growth in laboratory- and full-scale conditions

In a second step, laboratory-scale tests are used to quantify bacterial growth in the unmodified treatment plant effluent, which is then compared with the changes detected during full-scale water distribution. The treatment plant effluent sample is incubated as such in AOC-free glassware at the same temperature as the distributed water and bacterial cell concentrations are measured at time intervals representative of the DWDS samples. Differences in the extent of bacterial growth in the laboratory batch test and in the DWDS at comparable residence times and temperature are reflective of the effect of the prevailing distribution conditions on the extent of bacterial growth in the DWDS.

Determination of bacterial growth-controlling factors in the treatment effluent water

The third step is an assessment of bacterial growth-controlling factors in the treatment plant effluent. This is performed with a combination of growth potential tests with varying incubation conditions (Table 1). The tests step-wise assess (1) the presence of residual disinfectant, (2) absence of viable bacterial cells, (3) organic carbon limitation, (4) inorganic nutrient limitations, and/or combinations of these. To enable comparison, all growth tests are performed at a fixed sample incubation temperature and incubation time. Figure 2 proposes a supportive scheme to (i) select appropriate tests for the studied drinking water system and/or (ii) interpret data obtained from combined tests for the evaluation of risks associated with bacterial growth limitations in a water sample.
Table 1

Scheme for comprehensive investigation of bacterial growth-limiting factors in drinking water

Test Quench* Inoculum** Nutrients*** Acetate Compare with Investigation 
T1 − − − − − What is the water growth potential? 
T2 − − − T1 Is there oxidative toxicity (disinfectant)? 
T3 − − T2 Are there live cells? 
T4 − T3 Is a non-carbon compound limiting growth? 
T5 – T3 Is the water carbon-limited? 
T6 T4 and T5 Are there multiple limitations? 
Test Quench* Inoculum** Nutrients*** Acetate Compare with Investigation 
T1 − − − − − What is the water growth potential? 
T2 − − − T1 Is there oxidative toxicity (disinfectant)? 
T3 − − T2 Are there live cells? 
T4 − T3 Is a non-carbon compound limiting growth? 
T5 – T3 Is the water carbon-limited? 
T6 T4 and T5 Are there multiple limitations? 

*Add only in case of treatments including chemical disinfection.

**Add only in case of treatments using chemical disinfection or physical barrier (membrane filtration, e.g. UF, NF, RO).

***Contains all nutrients needed for bacterial growth except carbon. A similar approach can be applied if a non-carbon limitation is detected, by adding individually different compounds (e.g. phosphorus, iron).

Figure 2

Decision-making tree for the assessment of growth-limiting factors in a water sample and associated risks for uncontrolled bacterial growth in a full-scale distribution system. Tx refer to the test numbers as indicated in Table 1. T1 = T2: same extent of growth in both tests; T2 > T1: extent of growth higher in T2 than in T1.

Figure 2

Decision-making tree for the assessment of growth-limiting factors in a water sample and associated risks for uncontrolled bacterial growth in a full-scale distribution system. Tx refer to the test numbers as indicated in Table 1. T1 = T2: same extent of growth in both tests; T2 > T1: extent of growth higher in T2 than in T1.

Novelty and value

While individual components of this approach have been applied in previous studies, the value of the present work is in the systematic and combined investigations at laboratory and full scales. The combination of laboratory tests and on-site investigations aims at the prediction of scenarios that can affect the water microbial growth properties, causing uncontrolled bacterial growth. This provides a basis for further detailed investigation and/or decision-making on treatment or distribution condition improvements, to better control bacterial growth in DWDS.

MATERIALS AND METHODS

Water samples

Experiments were carried out on bottled water (Evian, France) and on drinking water produced at a treatment facility in the Netherlands (Kralingen). At this location, surface water is treated by coagulation, flocculation and sedimentation, followed by ozonation (1.0–1.3 mg L−1, 10 min), dual medium filtration, and granular active carbon (GAC) filtration (approximately 15 min bed contact time). Chlorine dioxide (0.1 mg L−1) is added at the end of the treatment and the water is collected in a reservoir before distribution, resulting in a ClO2 concentration in the water after storage of 0.006 mg L−1 on average (Table 2). Chlorine dioxide concentrations within the subsequent distribution system are below the detection limit (<0.001 mg L−1) in most cases and were not reported by the water utility. An overview of the drinking water treatment effluent properties is given in Table 2.

Table 2

Overview of Kralingen water treatment plant effluent properties. Minimum, maximum and average values of each parameter measured in (N) number of samples over a 1-year period

Parameter Unit Min Max Average N 
Temperature °C 2.3 22 12.5 416 
Dissolved oxygen mg/L O2 10.3 19.4 13.9 52 
Turbidity FTU <0.05 0.36 <0.05 466 
pH  7.79 8.32 8.08 468 
Conductivity (20 °C) mS/m 40.5 49.6 44.8 417 
UV-extinction, 254 nm 1/m 1.1 2.6 2.1 27 
DOC mg/L C 1.07 1.97 1.61 28 
AOC µg/L ac-C 5.3 43 28 20 
Chlorine dioxide mg/L ClO2 0.034 0.006 93 
Carbon dioxide mg/L CO2 <1 2.8 1.9 13 
Hydrogen carbonate mg/L HCO3 106 134 121 13 
Chloride mg/L Cl 50.7 60.2 53.5 13 
Sulphate mg/L SO4 43 66 52 13 
Sodium mg/L Na 34 42 37 13 
Potassium mg/L K 6.5 5.6 13 
Calcium mg/L Ca 43.4 50 46.9 13 
Magnesium mg/L Mg 6.7 7.8 7.2 13 
Hardness mmol/L 1.363 1.538 1.463 13 
Ammonium mg/L NH4 <0.03 0.05 <0.03 52 
Nitrite mg/L NO2 <0.01 <0.01 <0.01 13 
Nitrate mg/L NO3 11.2 14.6 12.8 13 
Silicon mg/L Si 1.25 2.1 13 
Iron µg/L Fe <5 <5 36 
Manganese µg/L Mn <5 <5 <5 
Aluminium µg/L Al <10 <10 <10 13 
Antimony µg/L Sb <1 <1 <1 
Arsenic µg/L As <1 <1 <1 
Barium µg/L Ba 14 22 18 
Boron mg/L B 0.042 0.049 0.046 
Cadmium µg/L Cd <0.05 <0.05 <0.05 
Chromium µg/L Cr <1 <1 <1 
Mercury µg/L Hg <0.03 <0.03 <0.03 
Nickel µg/L Ni 
Selenium µg/L Se <1 <1 <1 
Strontium µg/L Sr 140 160 150 
Bromide mg/L Br 0.07 0.111 0.087 13 
Fluoride mg/L F 0.19 0.27 0.22 13 
Cyanide µg/L CN <0.5 <0.5 <0.5 13 
Bromate µg/L BrO3 2.3 6.8 3.6 26 
Chlorate µg/L ClO3 <40 57 <40 13 
Parameter Unit Min Max Average N 
Temperature °C 2.3 22 12.5 416 
Dissolved oxygen mg/L O2 10.3 19.4 13.9 52 
Turbidity FTU <0.05 0.36 <0.05 466 
pH  7.79 8.32 8.08 468 
Conductivity (20 °C) mS/m 40.5 49.6 44.8 417 
UV-extinction, 254 nm 1/m 1.1 2.6 2.1 27 
DOC mg/L C 1.07 1.97 1.61 28 
AOC µg/L ac-C 5.3 43 28 20 
Chlorine dioxide mg/L ClO2 0.034 0.006 93 
Carbon dioxide mg/L CO2 <1 2.8 1.9 13 
Hydrogen carbonate mg/L HCO3 106 134 121 13 
Chloride mg/L Cl 50.7 60.2 53.5 13 
Sulphate mg/L SO4 43 66 52 13 
Sodium mg/L Na 34 42 37 13 
Potassium mg/L K 6.5 5.6 13 
Calcium mg/L Ca 43.4 50 46.9 13 
Magnesium mg/L Mg 6.7 7.8 7.2 13 
Hardness mmol/L 1.363 1.538 1.463 13 
Ammonium mg/L NH4 <0.03 0.05 <0.03 52 
Nitrite mg/L NO2 <0.01 <0.01 <0.01 13 
Nitrate mg/L NO3 11.2 14.6 12.8 13 
Silicon mg/L Si 1.25 2.1 13 
Iron µg/L Fe <5 <5 36 
Manganese µg/L Mn <5 <5 <5 
Aluminium µg/L Al <10 <10 <10 13 
Antimony µg/L Sb <1 <1 <1 
Arsenic µg/L As <1 <1 <1 
Barium µg/L Ba 14 22 18 
Boron mg/L B 0.042 0.049 0.046 
Cadmium µg/L Cd <0.05 <0.05 <0.05 
Chromium µg/L Cr <1 <1 <1 
Mercury µg/L Hg <0.03 <0.03 <0.03 
Nickel µg/L Ni 
Selenium µg/L Se <1 <1 <1 
Strontium µg/L Sr 140 160 150 
Bromide mg/L Br 0.07 0.111 0.087 13 
Fluoride mg/L F 0.19 0.27 0.22 13 
Cyanide µg/L CN <0.5 <0.5 <0.5 13 
Bromate µg/L BrO3 2.3 6.8 3.6 26 
Chlorate µg/L ClO3 <40 57 <40 13 

Flow cytometric (FCM) measurements

FCM total cell concentration (TCC) was measured as proposed in the standardized Swiss guideline for drinking water analysis (SLMB 2012; Prest et al. 2013). In short, samples (500 µL) were pre-heated to 35 °C (5 minutes) and then stained with 10 µL mL−1 SYBR® Green I (1:100 dilution in DMSO; Molecular Probes), and incubated in the dark for 10 minutes at 35 °C before FCM measurement. For the assessment of intact bacterial cell concentrations (ICC), a working solution containing SYBR® Green I (1:100 dilution in DMSO; Molecular Probes) and propidium iodide (0.3 mM) was used for bacterial staining following the same protocol as described above. FCM measurements were performed using a BD Accuri C6® instrument (BD Accuri cytometers, Belgium) equipped with a 50 mW laser emitting at a fixed wavelength of 488 nm. The FCM was equipped with volumetric counting hardware, calibrated to measure the number of particles in 50 µL of a 500 µL sample. Measurements were performed at a pre-set flow rate of 35 µL min−1. A threshold value of 700 a.u. was applied on the green fluorescence channel (FL1). Triplicate FCM measurements had a relative standard deviation below 4%.

AOC-free material and experiments

All glassware used for water sampling and growth potential tests was prepared as described previously to make it AOC-free (Greenberg et al. 1993; Hammes & Egli 2005). In short, the glassware was soaked in acid overnight (HCl, 0.2N) and subsequently rinsed with ultrapure water (mQ, Biocel), air-dried and heated at 550 °C for 6 hours. The caps to cover the incubation glass vials (40 mL) and sampling bottles were Teflon coated and were soaked in a 10% sodium persulfate solution at 60 °C for at least 1 hour, rinsed with ultrapure water and air-dried. All glassware cleaning procedures, preparation of nutrient solutions and growth potential tests were performed in a clean laboratory environment, with reduced input of volatile organic compounds in the air. During growth potential tests, pipette tips were rinsed 10 times with ultrapure water before pipetting water from vials for FCM measurements, to prevent AOC from the pipette tip plastic being released into the sample.

Preparation of stock solutions

All stock solutions of quenching agent (sodium nitrite), organic carbon (sodium acetate), and inorganic nutrients were prepared with ultrapure water in AOC-free glass bottles. Final concentrations of quenching agent and organic carbon in water samples are given in the relevant sections below. The acidified inorganic nutrient stock solution was adapted from Ihssen & Egli (2004) and contained 1.28 g L−1 Na2HPO4.2H2O, 0.3 g L−1 KH2PO4 and 1.77 g L−1 (NH4)2SO4, as well as the following trace elements: 80 mg L−1 CaCO3, 11 mg L−1 MnCl2.4H2O, 1.5 mg L−1 CuSO4.5H2O, 1.3 mg L−1 CoCl2.6H2O, 4 mg L−1 ZnO, 1.2 mg L−1 H3BO3, 134 mg L−1 MgCl2.6H2O, 10 mg L−1 NaMoO4.2H2O, and 30 mg L−1 EDTA Na4.2H2O. A solution containing 2.7 g L−1 FeCl3.6H2O was prepared separately. To ensure carbon limitation during a growth potential experiment, 680 µL of the inorganic nutrient stock solution and 10 µL of the iron stock solution were added to a 20 mL water sample.

Direct incubation method for the assessment of growth potential: proof of principle

Control experiments were done with bottled mineral water (Evian, France) for the evaluation of the direct water sample incubation method for bacterial growth potential tests. This bottled water was specifically selected, as it contains low organic carbon concentrations, a range of inorganic nutrients and an indigenous community of viable bacteria (Berney et al. 2008; Marcussen et al. 2013).

Evaluation of the direct incubation growth potential method

Bottled water was transferred into AOC-free glass vials. Sodium acetate was added to triplicate vials to obtain final carbon concentrations of 0, 5, 10, 15, 20 and 25 µg-C L−1. Inorganic nutrients were added as described above. All vials were incubated in the dark at 30 °C for 3 days, and the initial and final TCCs were measured with FCM. Measurements were not performed after longer incubation times, as previous studies have shown that a stationary phase is reached after 3 days for bacterial communities growing in bottled water at 30 °C (Hammes & Egli 2005). The extent of bacterial growth, or net bacterial growth, was calculated from the increase in TCC between the start and end of the water sample incubation.

Assessment of bacterial growth-controlling factor(s) in bottled water

Bottled water was transferred into AOC-free glass vials and treated with or without the addition of organic carbon, inorganic nutrients, and/or quenching agent (Table 1). Organic carbon (acetate) was added to triplicate vials to obtain a final concentration of 1 mg-C L−1 and ensure that organic carbon was not the limiting compound. Inorganic nutrients were added as described above. The bottled water did not contain any disinfectant residual, but in order to verify that the quenching agent solution did not promote additional bacterial growth (e.g. from contamination), the quenching agent was added in part of the glass vials at a final concentration of 5 mmol L−1 NaNO2. No inoculum was added in this experiment, as the bottled water already contained viable bacterial cells and no growth-inhibiting substance. The samples were thereafter incubated in the dark at 30 °C for 3 days, and the initial and final total bacterial cell concentrations were measured with FCM.

Application of the systematic approach to a full-scale drinking water distribution system

Quantification of bacterial growth within the distribution system

Quantification of actual bacterial growth in the full-scale distribution system was performed by collecting water samples at locations with varying residence times (0, 24, 48–72 and 120 h). However, as the available tools to estimate residence time provide only rough estimations, sampling of the ‘same water’ at both locations could not be accurately achieved, and all samples were taken on the same day. For each estimated residence time, three to five water samples were taken at different locations in the drinking water distribution network. The samples were collected in high-density polyethylene (HDPE) bottles containing 2 mL L−1 of a mixed solution of sodium thiosulfate solution (20 g L−1) and of nitrilotriacetic acid (25 g L−1), as routinely applied for microbial water analysis in the Netherlands. The water temperature in the network samples was 11.6 ± 1.8 °C. The samples were transported on melting ice in a cooling box and kept at 4 °C until analysis. FCM measurements were performed in <24 h after sampling for the determination of total and intact bacterial cell concentrations.

Evaluation of growth potential of unmodified treatment effluent samples

For the assessment of the growth potential of the treatment plant effluent, a water sample was taken in a 250 mL AOC-free glass bottle, after the reservoir at the end of the treatment train. The water was transferred into triplicate AOC-free glass vials within 2 h of sampling. The vials were thereafter incubated in the dark, without further treatment, for 5 days at 12 °C, and FCM measurements were performed at the start and after 1, 2, 3, and 5 days of incubation for a comparison with bacterial growth occurring in the full-scale distribution system.

Assessment of bacterial growth-controlling factors in the treatment plant effluent

Identification of bacterial growth-controlling factors was performed on water collected at the full-scale treatment plant after the chlorine dioxide addition point (before the reservoir), to ensure the presence of disinfectant and to evaluate the potential of the laboratory growth potential method to detect the impact of residual disinfectant on bacterial growth potential. A water sample was collected in a glass bottle without addition of a quenching agent to study the effect of chlorine dioxide on bacterial growth. Another sample was collected in a separate bottle and quenched with sodium nitrite (5 mmol L−1). An additional sample was taken before the addition of chlorine dioxide (i.e. GAC filtrate) to be used as inoculum. All samples were collected in AOC-free glass bottles. The water without the addition of the quenching agent was transferred into triplicate AOC-free glass vials, within 2 h of sampling, without further treatment (test 1 in Table 1). The water sample containing the quenching agent was also transferred into glass vials that were treated with the addition of bacterial inoculum, organic carbon and/or inorganic nutrients, according to Table 1 (tests 2–6). Inoculum (i.e. GAC filtrate containing 125 × 103 cells mL−1 as determined with FCM) was added in order to add 104 bacterial cells mL−1 to the individual vials. Acetate was added to triplicate vials to obtain a final concentration of 500 µg-C L−1. Inorganic nutrients were added as above. All treatments were done in triplicate vials, which were incubated in the dark at 30 °C for 10 days. The TCC in each vial was measured with FCM at the start and after 3, 7 and 10 days of incubation, to check for a stationary phase.

Table 3 provides an overview of the drinking water types and sampling locations of the drinking water samples collected for each experiment, as well as the incubation conditions for the bacterial growth potential tests.

Table 3

Overview of investigations with corresponding drinking water sampling locations and experiment descriptions

Investigation Water type and/or sampling location Experiment description Growth potential test incubation conditions Figure 
Direct incubation for the assessment of growth potential of water: proof of principle 
 Evaluation of the direct incubation growth potential method Bottled water Growth potential tests with varying acetate concentrations 30 °C, 3 days Addition of acetate at varying concentrations 3 
 Assessment of bacterial growth-limiting factor(s) in bottled water Bottled water Growth potential tests with additions of quenching agent, inorganic nutrients and carbon solutions 30 °C, 3 days Additions according to Table 1 4 
Application of the systematic approach to a full-scale distribution system 
 Quantification of changes in planktonic bacterial cell concentration during distribution Full-scale system, various locations Total and intact cell concentration measurements in water samples taken at various residence times No growth potential tests 5 
 Evaluation of growth potential of WTP effluent water WTP effluent water after reservoir Growth potential test without addition, incubated at full-scale system water temperature 12 °C, 5 days No addition 6 
 Assessment of bacterial growth-limiting factor(s) in water WTP effluent water before reservoir Growth potential tests with additions of quenching agent, inoculum, inorganic nutrients and carbon solutions 30 °C, 10 days Additions according to Table 1 7 
Investigation Water type and/or sampling location Experiment description Growth potential test incubation conditions Figure 
Direct incubation for the assessment of growth potential of water: proof of principle 
 Evaluation of the direct incubation growth potential method Bottled water Growth potential tests with varying acetate concentrations 30 °C, 3 days Addition of acetate at varying concentrations 3 
 Assessment of bacterial growth-limiting factor(s) in bottled water Bottled water Growth potential tests with additions of quenching agent, inorganic nutrients and carbon solutions 30 °C, 3 days Additions according to Table 1 4 
Application of the systematic approach to a full-scale distribution system 
 Quantification of changes in planktonic bacterial cell concentration during distribution Full-scale system, various locations Total and intact cell concentration measurements in water samples taken at various residence times No growth potential tests 5 
 Evaluation of growth potential of WTP effluent water WTP effluent water after reservoir Growth potential test without addition, incubated at full-scale system water temperature 12 °C, 5 days No addition 6 
 Assessment of bacterial growth-limiting factor(s) in water WTP effluent water before reservoir Growth potential tests with additions of quenching agent, inoculum, inorganic nutrients and carbon solutions 30 °C, 10 days Additions according to Table 1 7 

WTP: water treatment plant.

RESULTS

Direct incubation for the assessment of growth potential: proof-of-principle

Growth potential control experiments with bottled water showed that direct incubation of water samples without modification of the autochthonous bacterial community provides a sensitive and repeatable assessment of the extent of bacterial growth that a water sample can support (Figure 3). The extent of bacterial growth (net growth) was proportional to the carbon content in the water when no other nutrients were limiting and concentrations steps as low as 5 µg-CL−1 could be distinguished. A numerical yield of 1.4 (±0.2) × 107 cells (µg Ac-C)−1 was measured, which is in the upper range of previous observations (van der Kooij 2002; Ross et al. 2013). The results were obtained with good repeatability, with <4% error on average on triplicate samples.
Figure 3

Relationship between carbon concentration and bacterial growth in bottled water. Acetate was added to bottled water at varying concentrations, and each sample was incubated in triplicate AOC-free glass vials in the dark at 30 °C. The bacterial cell concentration in each vial was measured after 3 days using flow cytometry. The net growth is the increase in bacterial cell concentration between the start and end of the incubation period. The error bars indicate the error on triplicate samples.

Figure 3

Relationship between carbon concentration and bacterial growth in bottled water. Acetate was added to bottled water at varying concentrations, and each sample was incubated in triplicate AOC-free glass vials in the dark at 30 °C. The bacterial cell concentration in each vial was measured after 3 days using flow cytometry. The net growth is the increase in bacterial cell concentration between the start and end of the incubation period. The error bars indicate the error on triplicate samples.

The step-wise approach proposed in Table 1 was further used to identify bacterial growth-controlling compounds in the bottled water samples. Figure 4 shows that this particular bottled water without any additives supported the growth of 133 (±18) × 103 bacterial cells mL−1 (corresponding to approximately 9.8 µg Ac-C L−1 equivalent; based on the yield determined above). The addition of the quenching reagent did not significantly affect the extent of bacterial growth in the case of this water that did not contain residual disinfectant. The addition of excess inorganic nutrients had a negligible effect on growth (170 (±10) × 103 cells mL−1), compared to the addition of 1 mg L−1 carbon (5,500 (±1,700) × 103 cells mL−1), clearly indicating that the primary bacterial growth-limiting compound was organic carbon. However, the bacterial cell concentration obtained with the addition of acetate-carbon was lower than expected from the introduced carbon concentration (ca. 1.4 × 107 cells mL−1; Figure 3), suggesting that not all of the newly added carbon was consumed, due to a possible secondary limitation. This was confirmed by significantly higher cell concentrations (1.1 (±0.34) × 106 cells mL−1) in the samples where both acetate-carbon (1 mg L−1) and excess inorganic nutrients were added. In summary, these control experiments with the direct incubation method showed sensitive quantification (as low as 5 µg-C L−1) of proportional bacterial growth, and the identification of a primary organic carbon limitation of growth at 1.7 × 105 cells mL−1 and a secondary inorganic nutrient limitation of growth at 5.5 × 106 cells mL−1.
Figure 4

Investigation of bacterial growth-limiting factor(s) in bottled water. The water sample was treated with different solutions, as summarized in Table 1. The net growth is the increase in bacterial cell concentration between the start and end of the incubation period (3 days, 30 °C), as measured by flow cytometry. The error bars indicate the error on triplicate samples for each condition. −: No addition; +Q: addition of quenching agent; +NP: addition of inorganic nutrient solution; +C: addition of organic carbon.

Figure 4

Investigation of bacterial growth-limiting factor(s) in bottled water. The water sample was treated with different solutions, as summarized in Table 1. The net growth is the increase in bacterial cell concentration between the start and end of the incubation period (3 days, 30 °C), as measured by flow cytometry. The error bars indicate the error on triplicate samples for each condition. −: No addition; +Q: addition of quenching agent; +NP: addition of inorganic nutrient solution; +C: addition of organic carbon.

Application of the systematic approach to a full-scale distribution system

The combination of full-scale sample investigation and laboratory-scale growth potential tests (Figures 1 and 2, Table 1) was applied to samples from a Dutch full-scale DWDS.

Quantification of changes in planktonic bacterial cell concentration during water distribution

FCM analysis of samples taken at the treatment plant effluent and at different locations in the DWDS enabled the detection of changes in cell concentrations during water distribution (Figure 5). The water sampled at the treatment plant effluent contained 117 (±2) × 103 cells mL−1, with 38 (±2)% intact bacterial cells. This low percentage of intact cells is attributed to the addition of chlorine dioxide after the final treatment step, before storage in the clear water reservoir. Significantly higher TCCs (147 (±18) × 103 cells mL−1) were measured in DWDS samples with approximately 24 h residence time (t = 2.92, p = 0.04 based on independent sample t-tests; Table S1 in the Supplementary Information, available with the online version of this paper). TCC at longer residence times in the distribution network ranged from 165 × 103 to 176 × 103 cells mL−1 and were not significantly different from the 24 h samples (p > 0.1 in any case; Table S1 in the Supplementary Information). Notably, the TCC increase after 24 h residence time in the DWDS was mainly due to an increase in intact cell concentration (ICC) (from 45 (±1) × 103 cells mL−1 to 65 (±15) × 103 cells mL−1), with the percentage ICC consequently increasing up to 44 (±5)%.
Figure 5

Investigation of bacterial growth in a Dutch full-scale drinking water distribution system. Total, intact and damaged cell concentrations of water samples taken at various residence times from the network.

Figure 5

Investigation of bacterial growth in a Dutch full-scale drinking water distribution system. Total, intact and damaged cell concentrations of water samples taken at various residence times from the network.

Comparison of data from the full-scale system and laboratory tests

Bacterial growth in the treatment plant effluent incubated without any modifications under temperature and time conditions similar to the DWDS (12 °C; 24–120 h) showed discrepancies compared to the increase in bacterial cell concentrations in the full-scale DWDS (Figure 6). In the batch growth tests, a lag phase in TCC was observed up to 60 h, which was not observed in the full-scale samples. Comparable TCC values were measured in the batch tests and DWDS after approximately 80 h incubation (no significant differences; Table S2 in the Supplementary Information, available with the online version of this paper). However, at longer residence times (120 h), significantly higher TCC were recorded in the batch test (294 (±35) × 103 cells mL−1) than in the full-scale system (176 (±25) × 103 cells mL−1) (p = 0.004, t = 5.20; Table S2). The TCC increase in the batch test (179 (±39) × 103 cells mL−1) was necessarily due to bacterial growth and thus to an increase in ICC. In comparison, an increase in ICC of only 42 (±14) × 103 cells mL−1 was recorded in the full-scale system at long residence times. The differences in cell concentrations measured in the laboratory-scale batch growth tests and in the full-scale DWDS show that the distribution conditions affect bacterial growth in the pipelines.
Figure 6

Comparison of the total bacterial cell concentrations measured in water samples taken from a Dutch full-scale distribution system at different residence times (histograms) and the bacterial growth observed in the water treatment plant (WTP) effluent incubated in triplicate in the laboratory at network temperature (12 °C) (growth curve). The error bars on the growth curve indicate the standard deviation on triplicate samples. Three to five samples were taken in the full-scale system for each residence time range (0 hours = WTP effluent water, 24 hours, 48 to 72 hours, 120 hours). The histogram error bars indicate the standard deviation on the samples taken at each residence time range.

Figure 6

Comparison of the total bacterial cell concentrations measured in water samples taken from a Dutch full-scale distribution system at different residence times (histograms) and the bacterial growth observed in the water treatment plant (WTP) effluent incubated in triplicate in the laboratory at network temperature (12 °C) (growth curve). The error bars on the growth curve indicate the standard deviation on triplicate samples. Three to five samples were taken in the full-scale system for each residence time range (0 hours = WTP effluent water, 24 hours, 48 to 72 hours, 120 hours). The histogram error bars indicate the standard deviation on the samples taken at each residence time range.

Assessment of bacterial growth-controlling factors in the treatment plant effluent

Multiple growth limitations were identified and quantified in the treatment plant effluent samples assessed with a series of step-wise growth potential tests (Figure 7). Firstly, a decrease in TCC (−116 (±6) × 103cells mL−1) was observed in the absence of the addition of the quenching agent, suggesting cell death caused by residual disinfectant. However, when the disinfectant was quenched, bacterial growth occurred (net growth of 192 (±72) × 103cells mL−1). The extent of bacterial growth was not significantly affected by the addition of a viable bacterial inoculum or by the addition of inorganic nutrients. This suggests that viable bacteria were initially present in the water sample despite the disinfectant and that the primary bacterial growth-limiting compound was not an inorganic nutrient. The latter observation was supported by the addition of excess organic carbon (0.5 mg-C L−1) that resulted in slight but significantly higher net growth (375 (±31) × 103 cells mL−1; t = 4.08 and p = 0.008), showing that the water sample was primarily limited in organic carbon. However, the obtained cell concentrations were considerably lower than expected from the addition of 0.5 mg-C L−1, suggesting that not all of the newly added carbon was consumed, due to a possible secondary limitation. This was confirmed by the addition of both organic carbon (0.5 mg-C L−1) and excess inorganic nutrients, which resulted in significantly higher cell concentrations (net growth of 3.1 (±0.5) × 106 cells mL−1). Overall, the set of bacterial growth tests revealed: (1) bacterial growth inhibition by residual chlorine dioxide; (2) primary growth limitation by organic carbon at ca. 1.9 × 105 cells mL−1; and (3) secondary growth limitation by inorganic nutrients at ca. 3.8 × 105 cells mL−1 in the case of organic carbon contamination.
Figure 7

Investigation of the growth-limiting factor(s) in the treatment effluent water of a Dutch full-scale drinking water treatment plant, sampled after the chlorine dioxide addition point. The water sample was treated with different solutions, as summarized in Table 1. The net growth is the increase in bacterial cell concentration between the start and end of the incubation period (10 days, 30 °C), as measured by flow cytometry. The error bars indicate the error on triplicate samples for each condition. −: No addition; +Q: addition of quenching agent; +in: addition of inoculum; +NP: addition of inorganic nutrient solution; +C: addition of organic carbon. Note the y-axis break.

Figure 7

Investigation of the growth-limiting factor(s) in the treatment effluent water of a Dutch full-scale drinking water treatment plant, sampled after the chlorine dioxide addition point. The water sample was treated with different solutions, as summarized in Table 1. The net growth is the increase in bacterial cell concentration between the start and end of the incubation period (10 days, 30 °C), as measured by flow cytometry. The error bars indicate the error on triplicate samples for each condition. −: No addition; +Q: addition of quenching agent; +in: addition of inoculum; +NP: addition of inorganic nutrient solution; +C: addition of organic carbon. Note the y-axis break.

DISCUSSION

We propose and demonstrate here a systematic approach to (i) identify and quantify bacterial growth-controlling factors in drinking water and (ii) evaluate the impact of distribution conditions on the extent of bacterial growth in full-scale DWDS (Figures 1 and 2). Laboratory-scale growth potential tests were performed by direct incubation of water samples, thus measuring the response of the autochthonous community to varying conditions (incubation temperature and time, chemical additions; Figure 1, Table 1). This direct incubation growth potential method was shown to be accurate and sensitive (Figure 3), and suitable for the identification of the growth-controlling factors in water (Figures 4 and 7). When applied to a full-scale DWDS, the comprehensive systematic approach showed that changes in bacterial cell concentrations during distribution (Figure 5) were always lower but not strictly in accordance with the predictions of growth potential tests (Figure 6), highlighting the influence of the distribution conditions on the bacterial growth in the DWDS. Interestingly, the growth potential tests revealed multiple limitations (disinfectant residuals, organic carbon and inorganic nutrients) in these specific samples (Figure 7).

Direct incubation of water samples for the evaluation of bacterial growth potential

Advantages of the direct incubation method

The direct incubation growth potential method is a straightforward, simplified version of existing growth potential tests using indigenous bacterial communities (Werner & Hambsch 1986; Hammes & Egli 2005) and an improvement of the method recently proposed by Gillespie et al. (2014). It is based on the incubation of water samples, without modification of the autochthonous community with steps such as filtration (Hammes & Egli 2005) or pasteurization (van der Kooij et al. 1982). The reduction of these sample handling steps is a major advantage compared to existing growth potential tests such as AOC or BDOC assays, limiting both labour and risk of sample contamination with external nutrients (Ross et al. 2013). Our method differs from the approach suggested by Gillespie et al. (2014) in that we used specifically prepared AOC-free glassware, instead of plastic containers for water incubation, to avoid potential contamination (Niquette et al. 2000; van der Kooij & Veenendaal 2001; Bucheli-Witschel et al. 2012). In the direct growth potential method, the growth of the autochthonous bacteria is studied rather than pure culture(s). The bacterial community is grown in its own environment and is therefore likely to be the most adapted community that will consume the maximum amount of available nutrients, therefore revealing the water's maximum growth potential. Drinking water bacterial communities are typically very diverse and contain up to 30 different phyla, 1,500 operational taxonomic units (Pinto et al. 2012; Lin et al. 2014; Liu et al. 2014; Prest et al. 2014; Wang et al. 2014; El-Chakhtoura et al. 2015), and likely organisms already adapted to the specific water environment. The method was shown to be reproducible and sensitive, with the detection of bacterial growth from carbon concentration down to 5 µg-C L−1 (Figure 3), which is comparable with other methods (between 1 µg-C L−1, van der Kooij et al. 1982, and 10 µg-C L−1, Hammes & Egli 2005). The easy handling of the direct incubation growth potential method offers the possibility of it being applied on a regular basis (e.g. bi-weekly) by water utilities at lower costs for water characterization.

Assessment of growth potential vs nutrient content

In the proposed growth potential method, we evaluate the maximum extent of growth that the autochthonous bacterial community can reach in the water, rather than estimating the concentration of specific compounds (e.g. AOC) used for growth. Bacterial growth has been shown to be proportional to the concentration of easily AOC compounds (e.g. acetate), both in the cases of pure bacterial strains (van der Kooij et al. 1982; van der Kooij & Hijnen 1985) and in the case of mixed bacterial communities (Werner & Hambsch 1986; Hammes & Egli 2005; Figure 3 in this study), as long as all other elements required for growth are available in excess. Thus, net growth can in theory be converted to organic carbon (equivalent) concentrations using measured or estimated growth yields (van der Kooij & Hijnen 1985; Hammes & Egli 2005; Ross et al. 2013). However, the conversion can only be accurate when single bacterial species and a single substrate are used, as yields vary between substrates and organisms (Kaplan et al. 1993; van der Kooij 2002). Moreover, the linearity between the extent of bacterial growth and organic carbon concentration would not apply if the AOC content cannot be used entirely by bacteria due to limitation in another element such as phosphate (this study; Miettinen et al. 1997). Conversion to carbon concentrations would then be erroneous if no other element is added in the growth test (Miettinen et al. 1999; Ihssen & Egli 2004; Chow et al. 2009), and would lead to a misinterpretation of the water growth properties. Hence we chose to express the growth potential results as the actually measured bacterial cell concentrations (cells mL−1), thus avoiding potentially erroneous conversions of bacterial cell concentrations to, for example, carbon concentrations (van der Kooij et al. 1982; Hammes & Egli 2005). The absence of bacterial growth in such tests is a straightforward indicator of biologically stable water with its given characteristics at the sampling location and time, and under the incubation conditions. However, changing conditions during water distribution might induce instability in the same water (discussed below).

Identification of bacterial growth-controlling factors in water

Adapting the growth potential method with straightforward chemical additions is a simple procedure for the identification of bacterial growth-controlling factors in water. Until now, most drinking water studies that include growth potential assays focused only on organic carbon limitation (van der Kooij 1992; Escobar et al. 2001; Niquette et al. 2001). Here we argue for a comprehensive and systematic assessment of growth-controlling compounds in water and their effect on the growth potential by identifying (a) the limiting factors, (b) the order in which they are limiting, and (c) the extent of microbial growth at which they become limiting. The addition of a quenching agent, bacterial inoculum, organic carbon and/or inorganic nutrients needed for bacterial growth revealed multiple growth-limiting factors (Figures 4 and 7). This information is particularly relevant for practice by shedding light on scenarios that can convert apparently stable water into unstable water with uncontrolled bacterial growth, as described in Figure 2. For example, bacterial growth in the Dutch drinking water described in the present study was shown to be inhibited by the presence of disinfectant residuals, which would inevitably decay in the distribution system and thus would leave room for bacterial growth (results of the test T2 > T1 in Figure 2). Similarly, a phosphate-limited water could be subject to promoting unexpected bacterial growth if additional phosphate is brought to the water via the mixing of different water types, cell death or the resuspension of sediments (Niquette et al. 2001; Zacheus et al. 2001; Lehtola et al. 2004). Organic carbon leaching from pipe materials can also provide bacterial growth-promoting nutrients in carbon-limited waters (Niquette et al. 2000; van der Kooij & Veenendaal 2001; Bucheli-Witschel et al. 2012), but would not alter growth if, for example, phosphate is limited. In this regard the double nutrient limitation observed in the presently studied water (T4 > T3, T5 > T3 and T6 > T5 in Figure 2) is a particularly efficient barrier against uncontrolled and unpredicted bacterial growth in the distribution system.

Table 1 proposes basic tests for rapid evaluation of the bacterial growth-promoting properties of water. A primary limitation in organic carbon and a secondary limitation in inorganic nutrients have been detected in both bottled water and the full-scale water treatment effluent (Figures 4 and 7). The individual compound responsible for the secondary limitation was not identified in this study, but this can easily be done through addition of individual inorganic compounds instead of the combined inorganic nutrients solution. Based on the nutrient requirements of bacteria we propose the order of additions to be as follows: N, P, Fe, and trace elements. Consequently, improvements to drinking water treatment can be considered to further reduce concentrations of identified growth-limiting nutrient(s) in water and thus reduce the overall growth potential of the water. In addition, the water sample incubation temperature can be easily adapted to quantify the extent of bacterial growth that can result from temperature fluctuations in distribution systems (Niquette et al. 2001; Hammes et al. 2010), or from increased water temperatures at household levels (Lipphaus et al. 2014). The incubation time of the growth test is another straightforward variable to evaluate the risks for bacterial growth-related problems that can rise in distribution areas with high residence time, particularly dead-end zones or household taps, where residence times can be much longer than what is commonly found in DWDS. Overnight stagnation especially has been shown to lead to increased bacterial cell numbers (Lautenschlager et al. 2010).

Note that bacterial growth-controlling factors are susceptible to change over time, due to seasonal variations in the nutrient composition and concentrations in the treated water (e.g. Pinto et al. 2014). Such seasonal variations can be due to changes in raw water quality (particularly surface water), and/or treatment efficiency. Hence, the step-wise assessment of growth-limitations should be repeated on a regular basis to provide a representative picture of growth-limitations in the treatment effluent. In a first study, the tests could be applied at high frequency (e.g. every 2 weeks), but the frequency could be lowered (e.g. every 3 months) once the dynamics in the system are well understood. Additionally, growth-limitation investigations should be considered in case of changes in operating conditions or implementation of new treatment. Such systematic understanding of bacterial growth-controlling factors in drinking water can provide a solid basis for achieving biological stability in DWDS: water treatment processes can be improved by, for example, targeting the removal of identified growth-limiting compounds to lower the growth potential of the water, while distribution conditions can be improved by optimizing, for example, distribution materials to reduce carbon leaching and/or hydraulic conditions in the system to minimize residence times.

Combining laboratory- and full-scale investigations

Based on the present study, we argue that laboratory- and full-scale investigations are complementary and should always be conducted in parallel to link predictive growth potential measurements with actual changes in the DWDS. Laboratory-scale tests are applied to investigate the water bacterial growth potential under controlled conditions and are useful for the prediction of potentially problematic situations during full-scale water distribution (Figures 2, 6 and 7), and for the optimization of treatment and/or distribution conditions. However, growth potential assays are relatively straightforward planktonic batch growth experiments under defined conditions, and an increase in bacterial cell concentration is necessarily due to bacterial growth on readily available compounds in the water. Full-scale DWDS in turn are extremely complex with variable conditions, and an increase in bacterial cell concentration can be due to bacterial growth from compounds in the water, but also from an external contamination (e.g. substrate leakage from pipe material), and/or due to resuspension of sediments or biofilm detachment. The assessment of bacterial growth in such systems can only be achieved by direct investigation in the system (Niquette et al. 2001; Lautenschlager et al. 2013; Liu et al. 2014). This enables the detection of changes in microbial water characteristics (Prest et al. 2014) and the identification of so-called ‘hot-spot’ areas within the same network with bacterial growth concern (Lautenschlager et al. 2013; Nescerecka et al. 2014). Hence, the effect of drinking water distribution conditions on bacterial growth should be evaluated by comparing the full-scale actual measurements and laboratory-scale predictive data. The present study is the first to directly compare the extent of bacterial growth in the same water under laboratory- and full-scale conditions. We observed faster planktonic cell concentration increase in the DWDS compared to the laboratory-scale tests, but lower final cell concentrations at approximately 120 h (Figures 5 and 6). We argue here that the increase in DWDS planktonic cells is a direct consequence of growth, either directly in the planktonic phase or in the biofilms and/or sediments with subsequent detachment. The faster increase in DWDS cell concentrations is attributed to a rapid depletion of residual disinfectant in the network, as well as the presence of viable bacteria in distribution network biofilms. The lower final growth in the network at long residence times (120 h) (Figure 6) is attributed to various factors, such as predation by protozoa or viruses (Sibille 1998; van Lieverloo et al. 2002) and/or the competition between bulk water and biofilm bacteria for the available nutrients. Arguably, a significant proportion of the available nutrients would be consumed by sediment and biofilm bacteria, as >95% of the total biomass in distribution pipelines is recorded to be in the biofilm and sediment phases (Zacheus et al. 2001; Flemming et al. 2002; Liu et al. 2014). Investigation of the specific growth-controlling factors within the distribution system can be considered, based on the hypothesis formulated after comparison of results from growth tests and full-scale investigations.

CONCLUSIONS

A systematic approach combining full-scale investigations with laboratory-scale growth potential tests is proposed for (i) the assessment of bacterial growth-controlling factors in drinking water and (ii) the evaluation of drinking water distribution on the extent of bacterial growth in full-scale distribution systems. Controlled bacterial growth potential tests of bottled water showed that:

  • direct incubation of water samples with the unmodified autochtonous bacterial community provides a straightforward quantification of the water growth potential;

  • bacterial growth in bottled water was limited by carbon at 133 (±18) × 103 cells mL−1 and by inorganic nutrients at 5,500 (±1,700) × 103 cells mL−1.

Application of the systematic approach to a full-scale DWDS showed that:

  • bacterial growth in drinking water samples was (1) inhibited by the presence of disinfectant residual, (2) primarily limited by organic carbon up to 192 (±72) × 103 cells mL−1, and (3) secondarily limited by inorganic nutrients at 375 (±31) × 103 cells mL−1;

  • faster increase but lower final TCCs occurred in the full-scale DWDS than in laboratory bacterial growth potential tests.

The systematic approach provides the basis for understanding bacterial growth-controlling factors in DWDSs and optimization of water treatment and distribution conditions.

ACKNOWLEDGEMENTS

This publication is based upon work supported by Evides Waterbedrijf and the King Abdullah University of Science and Technology (KAUST) Office of Sponsored Research (OSR) under Award No. URF/1/1728-01-01.

REFERENCES

REFERENCES
Berney
M.
Vital
M.
Hulshoff
I.
Weilenmann
H. U.
Egli
T.
Hammes
F.
2008
Rapid, cultivation-independent assessment of microbial viability in drinking water
.
Water Research
42
(
14
),
4010
4018
.
Bucheli-Witschel
M.
Kötzsch
S.
Darr
S.
Widler
R.
Egli
T.
2012
A new method to assess the influence of migration from polymeric materials on the biostability of drinking water
.
Water Research
46
(
13
),
4246
4260
.
Chow
C. W. K.
Kuntke
P.
Fabris
R.
Drikas
M.
2009
Organic characterisation tools for distribution system management
.
Water Science and Technology: Water Supply
9
(
1
),
1
8
.
El-Chakhtoura
J.
Prest
E.
Saikaly
P.
van Loosdrecht
M.
Hammes
F.
Vrouwenvelder
H.
2015
Dynamics of bacterial communities before and after distribution in a full-scale drinking water network
.
Water Research
74
,
180
190
.
Flemming
H.-C.
Percival
S. L.
Walker
J. T.
2002
Contamination potential of biofilms in water distribution
.
Water Science and Technology: Water Supply
2
(
1
),
271
280
.
Gillespie
S.
Lipphaus
P.
Green
J.
Parsons
S.
Weir
P.
Juskowiak
K.
Jefferson
B.
Jarvis
P.
Nocker
A.
2014
Assessing microbiological water quality in drinking water distribution systems with disinfectant residual using flow cytometry
.
Water Research
65
,
224
234
.
Greenberg
A. E.
Clesceri
L. S.
Eaton
A. D.
(eds)
1993
Standard Methods for the Examination of Water and Wastewater
.
18th edn
,
American Public Health Association
,
Washington, DC.
Hammes
F.
Berger
C.
Köster
O.
Egli
T.
2010
Assessing biological stability of drinking water without disinfectant residuals in a full-scale water supply system
.
Journal of Water Supply: Research and Technology – AQUA
59
(
1
),
31
.
Joret
J. C.
Levi
Y.
Volk
C.
1991
Biodegradable dissolved organic carbon (BDOC) content of drinking water and potential regrowth of bacteria
.
Water Science and Technology
24
(
2
),
95
101
.
Juhna
T.
Rubulis
J.
2004
Problem of DOC removal during biological treatment of surface water with a high amount of humic substances
.
Water Science and Technology
4
(
4
),
183
187
.
Kaplan
L. A.
Bott
T. L.
Reasoner
D. J.
1993
Evaluation and simplification of the assimilable organic carbon nutrient bioassay for bacterial growth in drinking water
.
Applied and Environmental Microbiology
59
(
5
),
1532
1539
.
Kerneis
A.
Nakache
F.
Degui
A.
Feinberg
M.
1995
The effects of water residence time on the biological quality in a distribution network
.
Water Research
29
(
7
),
1719
1727
.
Lautenschlager
K.
Hwang
C.
Liu
W. T.
Boon
N.
Koster
O.
Vrouwenvelder
H.
Egli
T.
Hammes
F.
2013
A microbiology-based multi-parametric approach towards assessing biological stability in drinking water distribution networks
.
Water Research
47
(
9
),
3015
3025
.
LeChevallier
M. W.
Welch
N. J.
Smith
D. B.
1996
Full-scale studies of factors related to coliform regrowth in drinking water
.
Applied and Environmental Microbiology
62
(
7
),
2201
2211
.
Lehtola
M. J.
Nissinen
T. K.
Miettinen
I. T.
Martikainen
P. J.
Vartiainen
T.
2004
Removal of soft deposits from the distribution system improves the drinking water quality
.
Water Research
38
(
3
),
601
610
.
Lipphaus
P.
Hammes
F.
Kötzsch
S.
Green
J.
Gillespie
S.
Nocker
A.
2014
Microbiological tap water profile of a medium-sized building and effect of water stagnation
.
Environmental Technology
35
(
5
),
620
628
.
Liu
G.
Bakker
G. L.
Li
S.
Vreeburg
J. H.
Verberk
J. Q.
Medema
G. J.
Liu
W. T.
Van Dijk
J. C.
2014
Pyrosequencing reveals bacterial communities in unchlorinated drinking water distribution system: an integral study of bulk water, suspended solids, loose deposits, and pipe wall biofilm
.
Environmental Science & Technology
48
(
10
),
5467
5476
.
Marcussen
H.
Holm
P. E.
Hansen
H. C. B.
2013
Composition, flavor, chemical food safety, and consumer preferences of bottled water
.
Comprehensive Reviews in Food Science and Food Safety
12
(
4
),
333
352
.
Miettinen
I. T.
Vartiainen
T.
Martikainen
P. J.
1997
Phosphorous and bacterial growth in drinking water
.
Applied and Environmental Microbiology
63
(
8
),
3242
3245
.
Miettinen
I. T.
Vartiainen
T.
Martikainen
P. J.
1999
Determination of assimilable organic carbon in humus-rich drinking waters
.
Water Research
33
(
10
),
2277
2282
.
Nescerecka
A.
Rubulis
J.
Vital
M.
Juhna
T.
Hammes
F.
2014
Biological instability in a chlorinated drinking water distribution network
.
PLoS ONE
9
(
5
),
e96354
.
Niquette
P.
Servais
P.
Savoir
R.
2001
Bacterial dynamics in the drinking water distribution system of Brussels
.
Water Research
35
(
3
),
675
682
.
Pinto
A. J.
Xi
C.
Raskin
L.
2012
Bacterial community structure in the drinking water microbiome is governed by filtration processes
.
Environmental Science & Technology
46
(
16
),
8851
8859
.
Prest
E. I.
Hammes
F.
Kötzsch
S.
van Loosdrecht
M. C. M.
Vrouwenvelder
J. S.
2013
Monitoring microbiological changes in drinking water systems using a fast and reproducible flow cytometric method
.
Water Research
47
(
19
),
7131
7142
.
Prest
E. I.
El-Chakhtoura
J.
Hammes
F.
Saikaly
P. E.
van Loosdrecht
M. C. M.
Vrouwenvelder
J. S.
2014
Combining flow cytometry and 16S rRNA gene pyrosequencing: a promising approach for drinking water monitoring and characterization
.
Water Research
63
,
179
189
.
Ross
P. S.
Hammes
F.
Dignum
M.
Magic-Knezev
A.
Hambsch
B.
Rietveld
L. C.
2013
A comparative study of three different assimilable organic carbon (AOC) methods: results of a round-robin test
.
Water Science and Technology: Water Supply
13
(
4
),
1024
1033
.
Sathasivan
A.
Ohgaki
S.
Yamamoto
K.
Kamiko
N.
1997
Role of inorganic phosphorus in controlling regrowth in water distribution system
.
Water Science and Technology
35
(
8
),
37
44
.
Servais
P.
Anzil
A.
Ventresque
C.
1989
Simple method for determination of biodegradable dissolved organic carbon in water
.
Applied and Environmental Microbiology
55
(
10
),
2732
2734
.
Sibille
I.
1998
Stabilité biologique des réseaux de distribution d'eau potable
.
Année Biologique
78
,
117
161
.
SLMB
2012
Determining the Total Cell Count and Ratios of High and Low Nucleic Acid Content Cells in Freshwater Using Flow Cytometry. Analysis Method 333.1, Swiss Food Book (Schweizerische Lebensmittelbuch)
,
Federal Office of Public Health
,
Switzerland
.
Uhl
W.
Schaule
G.
2004
Establishment of HPC(R2A) for regrowth control in non-chlorinated distribution systems
.
International Journal of Food Microbiology
92
(
3
),
317
325
.
van der Kooij
D.
1992
Assimilable organic carbon as an indicator of bacterial regrowth
.
Journal of American Water Works Association
84
(
2
),
57
65
.
van der Kooij
D.
2002
Assimilable organic carbon (AOC) in treated water: determination and significance
. In:
Encyclopedia of Environmental Microbiology
(
Bitton,
G.
ed.),
John Wiley & Sons
,
New York
, pp.
312
327
.
van der Kooij
D.
2003
Managing regrowth in drinking-water distribution systems
. In:
Heterotrophic Plate Counts and Drinking Water Safety
(
Bartram
J.
Cotruvo
J.
Exner
M.
Fricker
C.
Glasmacher
A.
, eds),
IWA Publishing
,
London
,
UK
, pp.
199
232
.
van der Kooij
D.
Hijnen
W. A. M.
1985
Regrowth of bacteria on assimilable organic carbon in drinking water
.
Journal Français d'hydrologie
16
,
201
218
.
van der Kooij
D.
Veenendaal
H. R.
2001
Biomass production potential of materials in contact with drinking water: method and practical importance
.
Water Science and Technology
1
(
3
),
39
45
.
van der Kooij
D.
Visser
A.
Hijnen
W. A. M.
1982
Determining the concentration of assimilable organic carbon in drinking water
.
Journal of American Water Works Association
74
,
540
545
.
van Lieverloo
J. H. M.
van der Kooij
D.
Hoogenboezem
W.
2002
Invertebrates and protozoa (free-living) in drinking water distribution systems. In:
Encyclopedia of Environmental Microbiology
(G. Bitton, ed.)
,
John Wiley & Sons
,
New York
, pp.
1718
1733
.
Wang
H.
Masters
S.
Edwards
M. A.
Falkinham
J. O.
3rd
Pruden
A.
2014
Effect of disinfectant, water age, and pipe materials on bacterial and eukaryotic community structure in drinking water biofilm
.
Environmental Science & Technology
48
(
3
),
1426
1435
.
Werner
P.
Hambsch
B.
1986
Investigation of the growth of bacteria in drinking water
.
Water Supply
4
(
3
),
227
232
.
Zacheus
O. M.
Lehtola
M. J.
Korkhonen
L. K.
Martikainen
P. J.
2001
Soft deposits, the key site for microbial growth in drinking water distribution networks
.
Water Research
35
(
7
),
1757
1765
.

Supplementary data