Real-time sensor measurements for a better water-quality policy

The current, hydraulic models are infrequently applied to water-quality issues in the distribution network; moreover, they are poorly, or only to a limited degree, equipped for this task. The area however does raise some important questions: How hard is the water at the customer’s? What coverage area is affected by a (faecal) contamination event? How well do (self-cleaning) networks perform when it comes to particle fouling or biofilm formation? Thanks to technological developments in the field of sensors, a (near) real-time view of key water-quality parameters will become possible.

This project researches an approach in which the automatic processing of real-time sensor data provides a more complete image of the water quality in the distribution network. The goal is to identify the water origin (in the case of several pumping stations) and to determine the mixing and/or switching system in the distribution network, as well as the residence times in such an area.

Method to determine mixing ratios

To determine the mixing ratios and the water origin with models using EGV measurements, this research will develop a data processing method and apply it to a specific case. The approach is appropriate for possible extensions to other (surrogate) parameters for microbial activity or water colour or odour. Spike tests will give insight into residence times.

Testing with measured conductivity

To test the method, one real and two virtual distribution areas will be set up and measurements collected. Working with the water companies, the researchers will make an inventory of the knowledge about the underlying system (i.e., distribution network and sensors) that is needed to be able to make the determinations. After the method has been elaborated, the project will research the degree to which the water origin, mixing ratios and residence times can be determined using EGV measurements. The residence times can be determined by using dosages of harmless tracer substances and thus generating a measurable EGV pulse. Although a real-time view of residence times is not achievable, this does provide extra information for the validation of the distribution network models.

Improving distribution network models

An important aspect of the research is the use of practical measurements to increase the reliability of the distribution network model by controlling – and, where necessary, adjusting – the valves, consumption and mains topology. After all, a reliable model is essential for a good modelled determination of the water-quality parameters.


Video – 01:29
Water company joint research: Analysis of sensor data