Biofilm Detachment during Intermittent and Continuous Water Supply and its Effect on Water Quality in a Controlled Pilot-Scale Water Distribution Study

Drinking water quality can deteriorate within piped distribution networks due to the detachment of biofilms. In continuous water supply (CWS) systems, pipes are continuously under fluid pressure and detachment is promoted by changes in hydraulic regime that increase shear stresses along the pipe walls. When services are intermittent, there are stagnant periods where water drains from the de-pressured pipes, and biofilm detachment occurs as the supply is reconnected. Here we investigate the microbial communities by comparing inlet and outlet bulk water conditions over three repeated sets of flushing experiments performed on 90 m long uPVC pipe sections in an above-ground test bed. The CWS section is primed under conditions of laminar flow, while the intermittent water supply (IWS) section is subjected to a 15 min daily supply window with 23.75 h of stagnation over a period of one month prior to each set of tests. Flow cytometry (FCM) reveals a substantial increase in total and live cell counts during initial flushing of the IWS section, which exhibits a higher biofilm detachment potential than CWS at similar flow rates. We find that the CWS section has much higher microbial alpha-diversity (2D Hill number) than the IWS section based on both 16S rRNA gene metabarcoding and fingerprinting of flow cytometric data. Through statistical analyses we show FCM fingerprinting to be a robust method for monitoring and quickly diagnosing microbial water quality, although it lacks the taxonomical depth of 16S rRNA gene metabarcoding. Daily flushing of the IWS section revealed an increase in nitrate levels in the bulk water, together with a decrease in ammonia and nitrite concentrations, suggesting nitrifying activity in biofilms exposed to the stagnant waters in the pipe. Concurrently, the relative abundance of ammonia oxidizing and nitrite oxidizing bacteria in biofilm samples taken from both CWS and IWS test sections increased at the end of the experimental test program, but more so in CWS sections. Hence CWS hydraulic conditions are more conducive to the growth of nitrifiers in biofilms, while daily supply periods under IWS may slow the development of nitrifying biofilms in the long run. We conclude that drinking water delivered through IWS presents a significantly different microbiome from CWS in both the outlet bulk water and pipe biofilm.