The Smart Water Network: A technical view
A Smart Water Network is a collaboration between device, data and human interaction. To enable this requires two-way near real time networks with field sensors, measurement and control devices; along with analytical software and services to provide visualisation and data to people. Such technology enables utilities to remotely and continuously monitor and diagnose problems, prioritise and manage maintenance issues as well as utilising data to optimise all aspects of the water network.
Integrating data from a “Smart Water Network” into other aspects of the business provides greater value. Consider the impacts on workforce management, asset management and procurement by automation and enabling the delivery of real-world actionable intelligence to where it is needed including control rooms, enterprise applications and direct to the workforce’s mobile devices. The following diagram highlights such interactions.
Figure 1 - Broad Scope of a 'Smart Water Network' interactions
As a result, Smart Water Networks offer water utilities a tremendous opportunity to improve productivity and efficiency whilst enhancing customer service. They also have incredible potential to help alleviate impending water scarcity.
We should also consider the positive impact on a utilities business operating model and opening up the opportunity to move towards a true Digital Utility. The diagram bellow shows how the operating model can address greater integration and collaboration through the evolving digital capabilities now available in the marketplace.
Figure 2 – Example of the impact of ‘Smart Water Network’ on the business operating model
There are key facets of a water system that are unique by their combination to a utility. Consider the geographically distributed nature of a water network with much of it in locations where power supply and communications are unavailable. The cost and very long service life expected of many of its components make it a particularly challenging environment for applying instrumentation and control. Add to this:
- the aging nature of many underground assets causing repeated leaks and bursts and
- a need to reduce energy consumption.
The need for better control could never be greater, hence the regulatory pressure being applied to water companies to invest in asset replacement and technology to improve their performance.
A smart water network consists of several layers of technology identified in the diagram below
Figure 3 - Layered View
At the bottom is the existing physical network of pumps, pipes, valves, plants, reservoirs, water sources and outfalls, not forgetting the consumer’s own plumbing.
Moving up through sensing and control into the data and IT elements through to the enabling decision making; we recognise the requirement for a coherent security posture across the whole network and supporting end to end capabilities.
Water companies have for many years applied instrumentation and control systems to their networks; In plant rooms, where power and telecommunications are possible, most have SCADA solutions deployed on a plant by plant basis. Challenges with power and location have prevented wider implementation of new technologies; pipes are buried under streets and pavements and are difficult and expensive to access and there is no inherent power supply in pipes as there is with electric utilities.
Recent technology improvements now allow instrumentation to move out of the plant rooms and into the dispersed network and it is this that is enabling the move to truly ‘Smart Water Networks’.
Telemetry relies on instrumentation and telecommunications. In the past this would have required PSTN circuits and mains power for each sensor. Development of Lithium batteries with 15-year service life and mobile radio technologies such as LTE Narrow Band and LoRA now enable communication with devices in remote and hard to reach locations including meter wells and basement cupboards. This further drives the ability to have mass deployments of devices such as Smart Water Meters and street by street leak detection, which would have been impossible with hard wired circuits or traditional private radio.
Collecting telemetric data is only part of the solution, on its own without the means to aggregate and analyse it the floods of data produced would soon overwhelm a network management team. How many companies receive so many alarms that all they can manage is keep up with cancelling them?
Fortunately, modern data science comes to the rescue with data integration (Data Lake) and analysis techniques some of which leverage Artificial Intelligence.
Such techniques form the engine room of a ‘Smart Water Network’ by collating the masses of incoming data and reducing it to actionable information. Typical examples include:
- Optimising maintenance schedules across a fleet of pumps by analysing vibration, energy and usage.
- Working out the root cause of customer complaints of contaminated water, or multiple sensor alerts, by analysing recent activity on the network such as valve operations and pipe maintenance.
- Optimising network configuration for forecast weather events in real time.
Mathematical modelling and digital technology advancements now allow a virtual clone to be created that represents every component in the water network, from individual pipes through to pumping and treatment stations. This virtual clone can take information from sensors deployed around the water network so that the virtual clone is a digital representative of the water network in near real time, basically a Digital Twin of the Water Network.
When you have a Digital Twin a world of possibilities opens up. You can analyse historical data at the macro and micro level across the network and use the digital twin to do a “what if” analysis and test changes before you make them. With further enhancements in Artificial Intelligence and Machine Learning, and gaining the trust of operations management staff, digital twin based analytics can learn how to respond to events in real time based on the knowledge built up in the digital twin model. It can then start to suggest remedial actions or even automate them once confidence has built up.
Output from the analytics process can be used to provide automated feedback into SCADA control systems in real time or management reports to assist in future planning. Integration with IT solutions such as Asset Management and Procurement will allow those systems to provide better support and optimise their associated procedures to reduce cost, all within strict cybersecurity control.
The ultimate goal of applying such technology is to enable leaks to be found and fixed, avoid pipe bursts and sewer floods and to reduce the carbon footprint of a network. It can also help capture knowledge to reduce dependency on the expertise of an aging workforce and provide customer related data that can improve service and reduce customer debt.
DXC identifies with a Smart Water Network market as an integration of smart hardware devices, smart technology solutions, and services into the legacy water distribution infrastructure. The primary focus is to increase service quality and efficiency addressing the changing demands for water and long-term sustainability and has potential to intelligently operate, provide decision support information and react in real time.
DXC have considerable experience building and enhancing Digital Twin models and we partner with fluidics engineering specialists when more complex modelling is required. By working with our partners, water companies, other industries and start-ups and by collaborating with regulators and academia, DXC is developing solutions to enable the water companies to respond to the growing demands whilst driving even greater efficiency.