Legionella growth potential of drinking water produced by a reverse osmosis pilot plant.

Treatment processes, such as membrane filtration with reverse osmosis (RO), are used to produce drinking water with a high degree of biostability. To our knowledge, the influence of RO water on biofilm formation and growth of L. pneumophila has not yet been investigated. Therefore, this study aimed (i) to determine the Legionella growth potential of (remineralised) RO-water produced by a pilot plant and to compare this to conventional treated groundwater, and (ii) to determine if different pipe materials, in contact with remineralised RO-water, can cause growth of L. pneumophila. The Legionella growth potential of water was determined with the boiler biofilm monitor (BBM) that mimics the flow of water in a premise plumbing system. The Legionella growth potential of materials in contact with remineralised RO-water was determined by using the biomass production potential (BPP)-test. ATP concentrations in the biofilm on the glass rings from the BBM fed with (remineralised) RO water fluctuated around 100 pg ATP cm-2. In contrast, BBMs fed with conventionally treated water resulted in ten-fold higher ATP concentrations in the biofilm. Moreover, conventionally treated water had a Legionella growth potential that was 1000-fold higher than that of (remineralised) RO-water. Furthermore, glass, copper and PVC-C had the lowest biofilm concentrations and Legionella growth potential in the BPP-test, followed by PE-Xb, PE-Xc and PE-100. The highest biofilm concentration and Legionella growth potential were with PVC-P. Hence, our study demonstrated that remineralised RO-water did not enhance growth of L. pneumophila in the BBM that mimics the premises plumbing system. However, when PE or PVC-P materials are used growth of L. pneumophila can still occur in the premises plumbing system despite the high quality of the supplied remineralised RO-water.

A comparison of additional treatment processes to limit particle accumulation and microbial growth during drinking water distribution.

Water quality changes, particle accumulation and microbial growth occurring in pilot-scale water distribution systems fed with normally treated and additional treated groundwater were monitored over a period of almost one year. The treatment processes were ranked in the following order: nanofiltration (NF) > (better than) ultrafiltration (UF) > ion exchange (IEX) for limiting particle accumulation. A different order was found for limiting overall microbial growth: NF > IEX > UF. There were strong correlations between particle load and particle accumulation, and between nutrient load and microbial growth. It was concluded that particle accumulation can be controlled by reducing the particle load in water treatment plants; and the microbial growth can be better controlled by limiting organic nutrients rather than removing biomass in water treatment plants. The major focus of this study was on microbial growth. The results demonstrated that growth occurred in all types of treated water, including the phases of bulk water, biofilm and loose deposits. Considering the growth in different phases, similar growth in bulk water was observed for all treatments; NF strongly reduced growth both in loose deposits and in biofilm; UF promoted growth in biofilm, while strongly limiting growth in loose deposits. IEX had good efficiency in between UF and NF, limiting both growths in loose deposits and in biofilm. Significant growth was found in loose deposits, suggesting that loose deposit biomass should be taken into account for growth evaluation and/or prediction. Strong correlations were found between microbial growth and pressure drop in a membrane fouling simulator which proved that a membrane fouling simulator can be a fast growth predictor (within a week). Different results obtained by adenosine triphosphate and flow cytometry cell counts revealed that ATP can accurately describe both suspended and particle-associated biomass, and flow cytometry files of TCC measurements needs to be further processed for particle loaded samples and/or a pretreatment protocol should be developed.