Southern Water is investing £15 million in artificial intelligence and 22,000 state-of-the-art monitors to dramatically cut the impact of fatbergs and other blockages that currently cause hundreds of pollution incidents every year.

The software behind the system is provided by Stormharvester - machine learning software developed in the UK powers the system and alerts the water company's control to risks before they become incidents.

The vast majority of sewer blockages are caused by ‘unflushables’ like wet wipes and plastics, as well as fat, oil and grease, gathering in pipes. These blockages can lead to environmentally damaging flooding outdoors and deeply distressing flooding inside homes and properties.

Tackling this issue is a priority for Southern Water - the innovative technology is expected to cut pollution incidents by up to 40%. This equates to around 500 fewer internal floods between now and 2025, and about 7,000 fewer external floods during the same period.

Southern Water said the roll-out of Smart Sewers builds on its existing improvement in tackling external sewer flooding - the utility said it had reported a total of 3,944 incidents last year – well below its target of 4,141. 

The monitors, which are being installed across high-risk areas of the 39,500km sewer network, communicate digitally with our Operational Control Centre, where technicians are warned about potential blockages forming long before any risk of a pollution spill from a manhole, or flooding directly into homes, schools, businesses or any other property.

Alex Saunders, Southern Water’s Head of Wastewater Networks, said:

“We already have a good record on cutting pollutions from sewers thanks our Network Protection Team. There is also growing awareness among customers that wet wipes and other unflushables combine with fat, oil and grease, incorrectly disposed of down drains, to form fatbergs. These block sewers and can cause wastewater to back up the system.

“This revolutionary technology will mean we can respond proactively instead of waiting for sewers to block. Spills from manholes are ugly, unpleasant and damage the environment. Internal sewer flooding is incredibly distressing. We believe the deployment of this system will have a real impact on both of these types of incident.”

The monitors and software in the control centre use ‘machine learning’ to understand the normal behaviour of sewers in both dry and wet weather, and then automatically flag when unusual flows are noted. They measure the height of wastewater under manholes.

Teams can then quickly be dispatched to clean sewers with high powered water jets, while investigations can also be launched to track the potential source of problems.

The water company, which on an average day supplies 556 million litres of water and treats 13371 million litres of wastewater, said the £15 million project is already having an impact.

Southern Water has also already deployed 7000 acoustic loggers across its drinking water network which can detect and pinpoint leaks before they become bursts.