water quality, network and hydraulic features to describe root causes to discoloured water
WRc's award-winning approach to reducing customer contacts related to discoloured water, using data analysis and water quality expertise.
More on Asset Resilience serviceswater quality, network and hydraulic features to describe root causes to discoloured water
a diagnostic model to correlate relationships between customer contacts and model features
the most appropriate method to address discolouration using model outputs and engineering expertise
This project won the Institute of Water South West Area Innovation Award 2023.
The independent judges, drawn from Academic, Water Company and Supply Chain commented: “This entry showed how existing data can be used innovatively to model issues around potable water discolouration. Some original thoughts turned into a practical application allowing the resolution of DWI Reg 28 issues. A worthy winner of this year’s submissions.”
Most of the water quality-related contacts received by UK water companies from consumers are concerned with aesthetic issues. Of these, the most frequently reported is brown-black-orange (b/b/o) discolouration. Although the overall number of b/b/o contacts has reduced in recent years, it is still the most significant and persistent aesthetic water quality issue raised by consumers. Events which lead to discoloured water, such as major bursts, are routinely investigated and relatively well understood. However, little research has been completed on the root cause of persistent low-level discolouration issues and which mitigation measures would be most cost-effective
Using our engineering and network water quality expertise, combined with failure analysis techniques, we have defined key network and water quality factors that can influence discoloured water. This is a multi-layered issue, requiring the interrogation of data streams on consumer segmentation, hydraulic, network and water quality information. New thinking and new techniques for data analytics has enabled us to take a data driven approach to identify which of these candidate factors are more strongly associated with discolouration contacts. The application of machine learning techniques has enabled the efficient detection of complex interactions between factors. Of particular interest are those factors that can be influenced by the activities of the water company, providing actionable insights using readily available data.
Companies will be able to use these new insights to tailor operational interventions in specific areas to help reduce the number of customer contacts. A reduction in contacts will reduce company costs and improve the company customer experience score (C-MEX).
Samantha Vince, previous Head of Water Quality
Bristol Water