Pre-processing and visualization of biofilm development in drinking water distribution systems

Abstract

Biofilm develops in drinking water distribution systems (DWDSs) as complex microorganism communities covered by an extracellular polysaccharide which provides them structure, protection against disinfectants and helps retain food. Biofilm poses a serious risk to DWDSs. The presence of substantial and active attached biomass in the inner wall of pipes can protect pathogenic microorganism, lead to the formation of high biocorrosion zones, and consume residual disinfectant. Various studies have been performed in relation to the influence that the characteristics of DWDSs have in biofilm development. Nevertheless, their joint influence has scarcely been studied, due to the complexity of the community and the environment under study. In this paper, an innovative work is carried out with the introduction of pre-processing methodologies as new tools that allow using the knowledge gained on the development of biofilm in DWDSs in a practical and efficient way. We compile currently available information of the physical and hydraulic conditions of the DWDSs that affect biofilm development to study the effect that the joint influence of these characteristics has in biofilm development. This article proposes pre-processing by Machine Learning approaches, preparing a case-study database to get inference results by posterior analyses. This helps to develop a scalable and interesting set of tools to understand biofilm behaviour with respect to their physical and hydraulic environment. Finally, intelligent data visualization techniques are applied to carry out an exploratory analysis of the obtained metadata and to identify possible interesting patterns and groups related to biofilm development in DWDSs.