This paper presents machine learning analysis to understand the factors impacting iron concentrations and discolouration customer contacts in drinking water distribution systems. Fourteen years of network sampling and additional data from a large UK utility were collated, analysed, and interpreted using self-organising maps (SOMs), which include complex network theory (CNT) centrality metrics for the first time, investigating how possible explanatory variables interact. The outputs are used to inform ensemble decision trees for risk estimation of iron exceedance and customer contacts for each of the utility’s DMAs, helping inform proactive maintenance.

Discoloured drinking water, caused by elevated concentrations of organic and inorganic particles, is unacceptable. It occurs due to accumulation and subsequent mobilisation of material from within drinking water distribution infrastructure. Discolouration is currently partially explained by either the theories of cohesive layers or gravitational sedimentation. It is proposed and shown here how the processes behind these two theories both occur and how to integrate them to better explain observed behaviour and inform operational interventions to reduce discolouration. Deficiencies in understanding regarding the process and factors that influence material accumulation are highlighted. Future research addressing these deficiencies will enable determination of long term sustainable management strategies balancing capital investment and operational maintenance to safeguard distribution of high quality drinking water.