Fig 1.
Schematic of the arrangement of service reservoirs (SR) within the four DWDS selected for further sampling (not to scale).
The SRs sampled for AOC are numbered from one to three in each distribution system, as indicated by the second number in each case with the first number denoting the DWDS to which the SR belongs. Unlabelled SRs are not sampled as part of this study and are only drawn to show the pathway of the water. SRs labelled with a * are subject to the pipe only effect, as water has not previously passed through a SR. Samples analysed included: WTW (raw): AOC, TCC & ICC; WTW (post-treatment): AOC, TCC, ICC & total chlorine; SR inlets: AOC; SR outlets: AOC, TCC, ICC & total chlorine.
Table 1.
Twenty WTW selected for AOC sampling within raw (pre-treatment) and final (post-treatment) water [10].
The final column represents the 4 DWDS selected for further AOC sampling at service reservoirs inlet and outlets.
Fig 2.
Comparison of cell counts enumerated using bacterial strains NOX (NOX HPC) and P-17 (P-17 HPC) with heterotrophic plate counts (HPC) or adenosine triphosphate (ATP), or using a natural microbial inoculum with flow cytometry (Nat Inoculum FC).
HPCs are recorded as CFU / mL and total cell counts (TCC) are recorded as cells / mL. ATP luminescence units converted to cells / mL using an ATP calibration curve. The average (n = 3) cell counts (± standard deviation) at each acetate carbon concentration are plotted. The flow cytometry count refers to the total cell count.
Fig 3.
Comparison of P-17 cell concentrations when enumerated using heterotrophic plate counts (HPC), flow cytometry (FC) or adenosine triphosphate (ATP) over a nine day period.
Cells are grown on in solutions containing 0 μg acetate carbon / L (control), and 100 μg acetate carbon / L. HPCs are recorded as CFU / mL and total cell counts (TCC) are recorded as cells / mL. ATP luminescence units converted to cells / mL using an ATP calibration curve. The average (n = 3) cell counts (± one standard deviation) are plotted.
Fig 4.
Comparison of P-17 cell growth over a 9 day period when incubated at (A) different temperatures and (B) with different inoculum densities. All samples were enumerated using flow cytometry and recorded as cells/mL. Cells were grown in solutions containing 100 μg acetate carbon/L. (A) Inoculum density was 500 cells/mL, temperature as indicated by the key. (B) Temperature was 15°C, inoculum density as indicated by the key. The average (n = 3) cell counts (± standard deviation) are plotted.
Fig 5.
The effect of sample location on natural inoculum growth rate.
The three inoculums were collected from three separate treated (post-treatment) water locations and prepared according to Hammes & Egli [21]. Inoculums were added to separate solutions containing 100 μg acetate carbon/L and incubated at 30°C for 9 days. Cell enumeration was via flow cytometry. The average (n = 3) cell counts (± standard deviation) are plotted.
Fig 6.
Yield factors produced when using NOX and P-17 grown in 0 – 1000 μg/L sodium acetate solutions.
Samples were inoculated with 10,000 cells ml-1, incubated at 15°C until maximum cell growth was achieved and enumerated using flow cytometry. Total cell counts are presented as averages ± standard deviation. The average (n = 3) cell counts (± standard deviation) are plotted.
Fig 7.
AOC concentration in raw (light bar) and treated (dark bar) water at 20 WTW (Table 1 for details of source water, treatment type and disinfectant) [26].
Data collected weekly over a two month period, average is presented (n = 24) ± standard deviation. Figure axis include the identification (ID) number of each WTW (1-20), the water source (GW = groundwater, Res = reservoir, Riv = river) and treatment type (RGF = rapid gravity filter, DAF = dissolved air flotation, GAC = granular activated carbon).
Fig 8.
Variation in the AOC concentration along 4 DWDS during A) summer and B) winter ( Table 1 for DWDS details). AOC plotted with respect to the time that water has spent in the network, i.e. hydraulic retention time (HRT). Hydraulic residence time is calculated from the pipe and/or SR volume divided by the flow rate (l/s). Locations of samples are indicated by the key and follow the sequence post-treatment (WTW outlet) and through 3 service reservoirs (SR), inlet and outlet. Data is the annual average ± standard deviation. A network schematic of the SRs within each system is provided in Fig 1. SR marked with a * are pipe only systems (water does not pass through an un-sampled SR).
Fig 9.
Seasonal water quality data for DWDS 2.
Variation in the mean a) AOC concentration b) total cell counts (TCC) and c) intact cell counts (ICC) in post-treated water and three service reservoirs (SR) within the DWDS. Nb different y-axis scale for each parameter, and different y axis scales used in Fig 9, Fig 10 and Fig 11. Panel B of Fig 9 has two y-axis so that both the TCC within post-treated water, and each of the SR, are visible. Data was split into spring, summer, autumn and winter seasons based on monthly averages in temperature.
Fig 10.
Seasonal water quality data for DWDS 3.
Variation in the mean a) AOC concentration b) total cell counts (TCC) and c) intact cell counts (ICC) in post-treated water and three service reservoirs within the DWDS. Nb different y-axis scale for each parameter, and different y axis scales used in Fig 9, Fig 10 and Fig 11. Data was split into spring, summer, autumn and winter seasons based on monthly averages in temperature.
Fig 11.
Seasonal water quality data for DWDS 4.
Variation in the mean a) AOC concentration b) total cell counts (TCC) and c) intact cell counts (ICC) in post-treated water and three service reservoirs within the DWDS. Nb different y-axis scale for each parameter, and different y axis scales used in Fig 9, Fig 10 and Fig 11. Data was split into spring, summer, autumn and winter seasons based on monthly averages in temperature.