Efficient control of pathogenic bacteria, specifically Legionella pneumophila, is one of the main concerns when operating industrial cooling towers. Common practices to limit proliferation involves use of disinfectants, leading to formation of disinfection by-product and increase in water corrosiveness. A disinfectant-free Legionella control method would make the industry more environmentally friendly. A pilot-scale cooling tower (1 m3/h) operated with demineralized water was used to investigate the potential of high-pH conditioning as a disinfectant-free alternative for control of L. pneumophila and other pathogens. One control experiment was performed under standard full-scale operation involving sodium hypochlorite dosage. Thereafter 3 alkaline pHs of the cooling water were tested: 9.0, 9.4 and 9.6. The tests lasted between 25 and 35 days. The cooling water from the basins were analysed for total cell count by flow cytometry, L. pneumophila concentration by plate count and occasional qPCR analyses targeting the mip-gene, bacterial and eukaryotic community analyses with 16S and 18S rRNA gene amplicon sequencing, relative abundance of eukaryotic to prokaryotic DNA by qPCR of the 16S and 18S rRNA gene. The L. pneumophila analyses showed considerable growth at pH 9.0 and pH 9.4 but was maintained below detection limit (< 100 CFU/L) at pH 9.6 without disinfection. Interestingly, the results correlated with the overall abundance of protozoa in the water samples but not directly with the relative abundance of specific reported protozoan hosts of Legionella. The pathogenicity based on 16S rRNA gene amplicon sequencing of the cooling water DNA decreased with increasing pH with a strong decline between pH 9.0 and pH 9.4, from 7.1% to 1.6% of relative abundance of pathogenic genera respectively. A strong shift in microbiome was observed between each tested pH and reproducibility of the experiment at pH 9.6 was confirmed with a duplicate test lasting 80 days. High-pH conditioning ≥ 9.6 is therefore considered as an efficient disinfectant-free cooling tower operation for control of pathogenicity, including L. pneumophila.

Understanding the bacterial dynamics in cooling towers is imperative for the assessment of disinfection efficiency and management of microbial risks linked to aerosol formation. The objective of this study was to evaluate the impact of feed water on the cooling water bacterial microbiome and investigate the survival ability of its members when exposed to continuous chlorine disinfection. Water from an industrial cooling water system (2600 m³/h) was collected over a 5-month period at 3 locations along the feed water line and 3 locations in the cooling tower. ATP measurements suggested that the average ATP-per-cell in the cooling tower evolved independently from the average ATP-per-cell in the feed water. Flow cytometry and 16S rRNA gene amplicon sequencing were then combined to quantify the bacterial dynamics in the whole system. A mass balance based equation was established to determine net growth and net decay of the cooling tower bacterial communities in order to evaluate the impact of continuous chlorination (0.35–0.41 mg Cl₂/L residual chlorine). The results indicated that cooling tower main community members were determined by the input feed water microbiome and the bacterial community structure was further shaped by varying decay rates of the microorganisms. Notably, the order Obscuribacterales showed to be growing in the cooling tower in the presence of residual chlorine up to 0.4 mg Cl₂/L, with a recurrent net growth of 260 ± 95%, taking into account the impact of the concentration factor. This conclusion was only possible thanks to the systematic analysis described in this paper and generates discussion about the resistance of Obscuribacterales to residual chlorine. The described mass balance approach provides a high level of understanding on bacterial dynamics and should be considered for future characterization studies of cooling towers in which accurate investigation of microbiome changes is essential.