Age of data driven leak management dawns
Data analytics software provides visualisation of information water companies need to understand events on their networks. (Image source: Ovarro Ltd.)
Global water resources are finite and stretched and the application of advanced leakage management technologies is one way to ensure supply and demand balances are met. Various drivers mean interest in the application of real-time data analysis has quickly gone from it being ‘quite a nice idea’ to ‘we need to do this, we need to do it now’.
In the UK the drivers are mainly regulatory, with strenuous leakage targets imposed on water companies in England and Wales to help close the gap. There are also challenging targets to bring down the duration of customer supply interruptions.
In Australia and parts of the US, leakage management is driven more by drought and availability of water. Utilities and municipalities are doing much more leakage reduction as part of a range of measures required to manage demand, including water restrictions and efficiency measures.
There are many cities where water availability is constrained and leakage management and improved visibility and management of the network is necessary to keep taps running. In 2018 Cape Town nearly hit day zero when residents experienced such a severe drought that there was justifiable concern that the taps would run dry.
Through a combination of very extreme water restrictions, a lot of focus on managing the water network and a bit of luck with the rain arriving just in time, Cape Town eventually managed to keep the network running. Moreover, there is a clear economic argument for reducing leakage where the cost of producing water outweighs the cost of finding leaks.
The good news is the increased availability and falling cost of real-time leak detection technology. As digital solutions become more readily available, that makes them more affordable for a wider range of situations.
Water utilities typically use flow and pressure data from night flows as good indicators of leakage events and to identify which particular areas might be experiencing leaks. As their approach moves into real-time data analysis, companies have the potential to locate and remedy the larger leaks before the customer even becomes aware of them.
Data analytics provides water utilities with the tools to run a range of statistical models which can be used for a variety of purposes in water network management. The software provides visualisation of the network that was not previously available and is making easily digestible information available on dashboards and in apps.
As well as traditional flow and pressure logging, utilities are carrying out more acoustic logging of noise data, which helps to locate the leaks as well as work out the rate of water loss. New analytical techniques like machine learning and artificial neural networks (ANN) are being used to process data in smarter ways.
These technologies use programmed algorithms that evolve and improve as they process more and more data over time. That intelligence is then used to make increasingly accurate predictions.
ANNs can tell quite quickly if something unusual is happening in the water network and reduce the need for analysts to review it. It is by bringing the full suite of technologies together – measuring pressure, flow, acoustics and anything else relevant and available – that operators can get a broader view of potential water loss and risk to supply continuity.
The UK has seen a significant shift to permanent installations of acoustic logging technologies where information is being acquired proactively. In other parts of the world, more traditional flow and logging equipment is still being installed and there is increased sectorisation of the network to locate leakage down to a specific area.
Acoustic logging technology is not new, ground microphones and hydrophones have been used for some time. Traditionally this has been on a lift-and-shift basis – the network operator identifies an area with leakage, acoustic logging equipment is installed and once the leak is found, the kit is moved to a new site.
As this acoustic logging technology advances and costs fall, the direction of travel is towards more permanent monitoring. This has allowed UK utility Anglian Water to install over 3,200 of fixed network hydrophone acoustic loggers and results have been very positive. Leakage teams have been able to accurately pinpoint leaks and they are experiencing very few false-positive readings.
Having loggers distributed throughout the network and listening in at different points makes it possible to measure the time taken for noise to reach different loggers. This technique, known as correlation, makes for greater accuracy in pinpointing leaks. During trials, Anglian Water have found that correlation can work over longer distances than previously thought, so fewer loggers are required to cover the whole network.
James Hargrave, regional operational leakage manager, Anglian Water said, "The results we achieve show that working collaboratively and thinking differently complements the technology and data platforms to deliver a holistic and scaleable solution."
Some UK companies are exploring best use of these techniques following trials over the last few years and there will be wider rollouts in the next five-year asset management period (AMP7) which started on 1 April 2020. Fixed network acoustic logging features in utilities’ business and water resource plans and chief executives are saying that this is what will help them achieve the stringent targets set by the regulator.
Looking ahead, the uptake of remote monitoring is expected to increase and the human resource challenge experienced during the coronavirus pandemic is a good example of why organisations must prioritise and accelerate their plans. During times when companies are very reluctant, or the situation makes it impossible, to send people out to work on the network and risk the health and safety of their employees, so new ways of carrying out operations will have to be found.
There are always tasks that must be done on site and others that can be carried out remotely and companies are trying to prioritise remote working where that is feasible. It is difficult to predict how that might impact on long-term trends, but given that remote working is already underway for equipment monitoring and meter reading, it is likely to accelerate in leak detection.
Even before Covid-19 there was a perfect storm driving leakage toward digital solutions – water scarcity, tightening budgets, carbon reduction, rising customer expectations and importantly, the availability of better and more affordable technologies.
The water industry has necessarily been conservative and risk averse and has not always moved as fast as others in embracing new technology. These solutions are now becoming more widespread and where the risk is greater, whether that is from water shortages or sending operatives out on site, the balance shifts.
Given their position supplying clean wholesome water to their customer base, utilities should be aiming for a point where they can notify the customer in advance of any issue and provide advice. The water crisis may be deepening, but the increased availability of proven technologies that capture the power of data means that a more efficient and proactive service is within reach.
Source: Ovarro Ltd.