• Home
  • Blog
  • How artificial intelligence is improving a city’s water supply

How artificial intelligence is improving a city’s water supply

Barcelona, the capital of Catalonia in Spain, is turning to artificial intelligence to avoid water supply problems like those that plagued Cape Town in 2018.

The city’s water supply authority, Aigües de Barcelona, is working with the Barcelona Supercomputing Center (BSC) on measures to improve water supplies and increase the efficiency of desalination operations.

In one joint project, the BSC modelled the network of more than 100,000 water supply pipes across Barcelona to see which ones were most likely to fail.

The work is helping Aigües de Barcelona carry out preventive maintenance to stop breakages before they happen, avoiding losses from the more than 187 cubic hectometres of drinking water the city uses every year.

Another project, due to go live in 2019, will see the BSC using artificial intelligence to help improve the operation of a critical drinking water treatment plant, in the town of Sant Joan Despí near Barcelona.


Reverse osmosis

The Sant Joan Despí installation uses a desalination process called reverse osmosis to purify water from the nearby Llobregat River and its aquifer. The process involves pumping water through a membrane to purify the liquid.

The pumps have to work harder, and use more energy, as the membranes get dirty. But cleaning the membranes all the time is not an option because it is costly and prevents the plant from working. The project aims to find an optimum midpoint.

To do this, the BSC is analysing the output of dozens of sensors in the Sant Joan Despí plant, looking at 10-minute interval data over a period stretching back nine years.

This should help Aigües de Barcelona to improve its maintenance planning while reducing the energy use and carbon footprint of the desalination plant.

“To begin with, we’ll help Aigües de Barcelona to plan ahead,” said Fernando Cucchietti, head of the data analytics and visualisation group at the BSC. “But the aim is to eventually do monitoring in real time.”

Water company

Elsewhere, he added, the BSC is working with Aigües de Barcelona to improve the way the water company’s operations centre deals with alerts.

The BSC has also helped Aigües de Barcelona to render its data in visual form, so that management can see trends and patterns more easily. “Aigües de Barcelona is working hard to stay at the forefront of innovation in the water industry,” Cucchietti said.

It is not the only water body going down this route, either. Gary Wong, water industry principal at the application software company OSIsoft, says water authorities can “benefit hugely” from artificial intelligence.

“Water utilities today are in a bind,” he says. “They provide an essential service and one that most people take for granted, but it’s far from easy.”

Even basic improvements to the level of analysis used across water infrastructure and energy use can lead to significant improvements, he says. Manila, for example, has been able to cut leakage in half thanks to improved analysis of sensor data.


About the author: Jason Deign writes as a correspondent for the International WaterCentre, charged with exploring water challenges and the ways these challenges are managed around the world.

Stay in touch

Sign up for our monthly newsletter