project

Smart Water Networks – 2

Expert(s):
Mario Castro-Gama MSc, Claudia Quintiliani PhD, Bram Hillebrand MSc

  • Start date
    01 Jun 2020
  • End date
    31 Dec 2022
  • Principal
    Joint Research Programme - Distribution
  • collaborating partners

This project within the Joint Research Programme of KWR and the water utilities concerns the question of how a smart network – a combination of a variety of sensors with the right algorithms – can in practice lead to a more reliable, efficient and transparent drinking water provision. The question we will try to answer is whether technology can help change the black box that is the pipe network, into more of a ‘grey box’ or ‘white box’. This could result, for instance, in the rapid detection of leakages, the quick identification of wrong valve positions, the limitation of excessively long water residence times, and the mitigation of water discolouration risk. The project focuses on the application of data obtained from the sensors in the pilots, and of other (freely available) data, for the operational management of the distribution network. This will be done on the basis of several pilots conducted at and with the water utilities.

The project will yield insight into Smart Water Networks – for example, regarding sensor types and numbers, sensor placement, frequency of readings, which extra data support the application, which data analysis methods are appropriate – and what this concretely offers a drinking water utility. The project will also ensure that the findings will be accessible in a pilot at one drinking water utility and made relevant for other drinking water utilities.

Pipe network no longer a black box

The distribution of drinking water through underground pipe networks is a process that is by and large hidden from view. This means that the pipe network can be seen as a reactor, whose internal processes are not all known with precision. The water quantity can for instance be impacted by a leakage or by closed valves. While the water quality can deteriorate because of contact with the pipe wall, the temperature of the surrounding soil, substance permeation from the surrounding soil, leaching of certain pipe materials, pipe-wall biofilm and other factors. These can potentially have a negative impact for example on the water’s microbial safety, odour, taste or colour. The water quantity or quality can also unintentionally be negatively affected during maintenance work on the network, such as repairs of pipe fractures or pipe cleaning. In such cases, the valve positions are not what they were thought to be, or a contamination may be introduced into the network.

In the current situation, the pipe network should therefore be considered a black box for the most part, and potential problems as difficult to resolve quickly and efficiently. The question we will try to answer in this project is whether technology can help us make the pipe network into more of a ‘grey box’ or ‘white box’. This would for instance enable the rapid detection of leakages, the quick identification of wrong valve positions, the limitation of excessively long water residence times, and the mitigation of water discolouration risk.

One refers to ‘smart networks’ in such cases. These are drinking water networks that are equipped with sensors (pressure, volume flow, EGV, temperature, etc.), which allow a drinking water distribution utility to better monitor and manage the network with regard to water quantity (flow, pressure), the condition of the pipes, or water quality. The term can also refer to the use of algorithms to find defects, such as pipe fractures, or for advanced process management. It is as if a nervous system were added to the body of the infrastructure. Smart networks enable drinking water utilities to make ‘smarter’ use of the infrastructure, without the networks themselves becoming smarter (Van Thienen and Blokker 2019).

This Joint Research Programme project concerns the question of how the combination of a variety of sensors, with associated algorithms, can in practice lead to a more reliable, efficient and transparent drinking water provision (Van den Broeke et al. 2019). The project focuses on the application of data obtained from the sensors in the pilots, and of other (freely available) data, for the operational management of the distribution network.

Pilots with different types of sensors and algorithms

The project will make use of various pilot areas of drinking water utilities with smart water meters, sensors and remotely controllable valves in the distribution network. Because not all possible sensors and algorithms can per pilot be studied at the same time, each pilot will produce answers to a portion of the questions. Over a period of three years, six to eight pilots will be conducted, which will in total provide a broader perspective.

Examples of pilots:

  • How can smart technology assist in detecting anomalies (such as leakages) in the pipe network? A leakage is detected by measuring the inflow volume (into the DMA), and comparing it with either the historical inflow or the outflow volume (water meters). This allows for a focus on the daily water consumption, the nightly water consumption, or the water consumption at specific times.
  • How can smart technology assist in locating anomalies (such as leakages and closed valves) in the pipe network? A leakage or a closed valve can cause an anomaly in the water flow in the pipe network. This means that it should be possible to locate a leakage or a closed valve by using pressure meters and/or tracers in combination with water-quality sensors (e.g., EGV), and by comparing the results with historical data or with a calibrated (hydraulic + water quality) network model. In this context, it should be noted that some parameters can be seen as conservative tracers (EGV), and others as non-conservative tracers (e.g., temperature). It is therefore important to dispose of a good model that describes the change in the non-conservative tracer. The current approach to leakage location uses volume-flow and pressure sensors; in this pilot, we examine how other sensors can improve the localisation.
  • How can smart technology assist in managing water-quality incidents? When a water contamination is detected, the response is to issue a boil-water advisory to customers, trace and eliminate the contamination source, and isolate and remove the contaminated water. Contamination detection is a research project in itself, as is the response strategy, in which the effectiveness of the response depends in part on how quickly the contamination is detected and located. For the detection and source tracing one can research whether the combination of information received from several types of sensors or from sensors in two (or more) different network layouts (to be realised by using valves), result in a better detection or localisation of the source. This can be tested by dosing of a tracer and measuring EGV and turbidity or pH; there is no need to carry out a real contamination. In the response strategy, one option is to use valves and fire hydrants to direct the water; it is also important to monitor whether all the contaminated water is in fact removed.

Smart Networks for an efficient drinking water provision and good water quality

The objective of this Joint Research Programme project is to study in practice how extra sensors and data analysis can contribute to a greater understanding of the (hydraulic and water-quality) processes in the distribution network, so that the drinking water provision can be delivered as efficiently and with as little quality loss as possible.  The project will yield insight into Smart Water Networks – for example, regarding sensor types and numbers, sensor placement, frequency of readings, which extra data support the application, which data analysis methods are appropriate – and what this concretely offers a drinking water utility. The project will also ensure that the findings will be accessible in a pilot at one drinking water utility and made relevant for other drinking water utilities. This will enable the drinking water utilities to further shape Smart Water Networks (outcome), which will lead to a better service and lower costs for the drinking water customer (impact).