Nov 15, 2019 1:00 AM

To stop London running out of water, a crack squad is hunting down the city's mega leaks

London loses more than half a billion litres of water to leaks every day, and things are projected to get worse. Can an army of water whisperers track down stubborn leaks?

olando_o / Getty Images / WIRED

London is leaking. Over the last year more than 500 million litres of water spilled out of the capital’s ageing water infrastructure every day. Thames Water – the utility company that supplies water to most of London’s 8.8 million residents – manages almost 32,000km of aged pipes, 60 per cent of which were laid more than 60 years ago. The oldest are more than 150 years old.

And the strain on London’s creaking pipes is set to get even worse. London’s population is the fastest-growing in the UK, estimated to increase by 8.8 per cent in the decade leading up to 2026. The increasing population and relatively rain-free climate (compared to the north of the country) make the southeast of England the most water stressed part of the UK, with the Environment Agency predicting that, unless things change, the capital will face major water shortages by 2050.

But an army is mobilising against London’s looming water shortage. Every day, between 7:30pm and 6:00am the next day, squads of water whisperers attempt to locate the sources of hidden leaks. Armed with eclectic toolkits stuffed with archaic listening sticks and pipe-diving cameras, these detectives in high vis jackets are the last line of defence in London’s battle against leaky pipes.

Detecting a leak can sometimes feel like chasing a ghost. “The majority of leaks you’re never going to see,” says John Joe Healy Matthew, a reduction supervisor for Hydrosave, the contractor that handles Thames Water’s leak detection in north London. Big leaks that fill streets with water and send people stripping supermarket shelves of bottled water make headlines, but the real damage happens below the surface. On October 8, a dramatic leak in north London left cars partially submerged and led to homes being evacuated, but such blockbuster spills account for just two per cent of all leaks.

“For a long time the best way we had of detecting leaks was when a customer said they weren’t getting water, or when there was water coming out of the ground,” says Richard Collins, lecturer in water engineering at the University of Sheffield. In 1995, six years after the UK water industry was privatised, total water leakage in England and Wales peaked at over five billion litres per day. After that point, Collins says, water companies’ focus shifted from tackling visible leaks to dealing with invisible drips and spills.

But finding them is a tricky business. To help narrow down where water is being lost, water firms divide their fiefdoms into chunks of between 1,500 and 4,500 houses called district metered areas (DMAs). North London alone is divided into 600 of these DMAs, each equipped with water metres that measure how much water enters that zone. By analysing how much water is being pumped into the area at night (when water consumption is lowest), leak detection engineers can spot a spike in demand that might indicate a leaky pipe.

Then, after narrowing down the leak using further metered areas, the leak detectives can head to the streets to find its exact location. The goal? Avoid the bogeyman of the leak detection world: the dry hole.

Repairing leaks is costly and disruptive. One engineer estimated that calling out a repair squad to dig up the road costs £7,000 in labour and disruption. That’s why the leak detectives – who label a one square metre box around the suspected leak – are keen to avoid situations when repair teams turn up only to find out that the leak was elsewhere. Thames Water sets itself the target of one dry hole for every 20 digs. Any more and the cost of disruption and labour starts to outweigh the benefits of repairing the leak in the first place.

To avoid these dry holes and pinpoint the leak, the water detectives are finding themselves turning to an increasingly sophisticated set of tools. Once the engineers have isolated the leaking pipe, they use acoustic sensors to find its precise location. Engineers place these sensors which listen out for the hissing of a leak along a length of pipe and then their results are processed by an algorithm that determines how long the sound takes to travel from its source to the sensor, and from that derives the location of the leak.

Sometimes the best acoustic sensor is a pair of ears. Most leak detection crews still carry listening sticks – lengths of thin metal piping topped with a wooden ear cone. Since pipes are usually buried about a metre underground – to avoid the stress of frost – the sound from a leak conducts through the tarmac and, at close range, is easily picked up by a detective wielding a listening stick. More modern electronic versions use microphones and headphones to amplify the sound of the leak and make it trackable from further away.

But not every leak is noisy enough to be picked up by a listening stick. While smaller leaks are squeezed out under high pressure with plenty of noise, bigger leaks can gush on undetected. Plastic pipes, used to replace the older cast iron network, are more reliable but conduct less noise, which makes divining the location of these leaks particularly tricky.

When acoustic methods of tracing leaks fail, it’s time to get out the camera. Alan Hill has worked in leak detection for more than 30 years, 17 of them with Hydrosave, and for the last three years he’s been equipped with an underwater camera designed to tease out the most stubborn of leaks.

Armed with 100 metres of cabling, Hill’s job is to pilot his cameras – the prosaically-named JD 7 Investigator and its bigger sibling, the Investigator Plus – through pipes until he can see and hear the leak through his laptop. Pushing cameras through metres of pipe caked with calcium deposits is an unforgiving task. There is no way to control the direction of the camera, so getting started requires bashing the camera off the bottom of the pipe and hoping it richotes in the right direction. Other camera operators, he says, have caught the cable on the sharp rim of a broken pipe and severed it altogether.

When I accompany Hill on a leak-detecting mission on a residential street just south of central London, on a mid-October night cloaked in a light rain, he’s not having much luck. While Hill feeds the cable down the length of pipe, a colleague is crouched next to a laptop watching it inch its way painfully through the calcified pipe, lit by a ring of LEDs around the camera lens. The mystery leak, the crew surmises, probably isn’t in the pipe that the camera has been snaking its way through, but was likely on a tributary that used to run to a property but was now ejecting its contents into the soil beneath London’s streets. When people demolish a building they occasionally forget to properly terminate the pipe that supplied it with water, Matthew says, and instead hammer a kink in to stop the water flowing out. But these kinks sometimes fail years later and lead to mystery leaks.

Undeterred by the phantom leak, Hill opens up his laptop to show me a video of one of his recent success – a 2017 leak he uncovered under Oxford Street. Leakage reduction engineers compare the size of leaks like keen fisherman sizing-up the weight of their catch. “There’s a good one at Harvey Nichols – seven litres a second,” one engineer chips in. “Last year we had a ten litres a second,” adds Matt Thomas, an operations manager at Hydrosave. Over a day that leak would spill 860,000 litres beneath the streets of London – the same amount of water used by 13,000 showers lasting 8.5 minutes each.

Despite repairing an average of 1,400 leaks a week, Thames Water missed its leak reduction targets in 2018-19 by 72 million litres per day. In June 2018 the utility firm agreed to pay back £120 million to its customers for its failure to meet its leakage reduction targets. The amount of leak reduction, Collins says, is usually directly linked to the amount that firms put into finding and repairing the leaks. In 2019, Thames Water signed three contracts worth a total of £200 million over eight years to pay for leak detection and fixing.

But sometimes leak detection isn’t an engineering problem alone. It can also be a data challenge. “We know how much water goes into the system but don’t have a great deal of certainty about how customers are using water and when,” says Collins. Only half of all UK homes use water metres and the uptake of smart metres – which take water use readings every hour or so – is tiny compared to the energy and heating industry.

Water metres also have the added bonus of reducing overall water usage. In England and Wales, homes with a water metre use on average 13 per cent less water than unmetered homes. Knowing how people were using water would help companies like Thames Water tell the difference between a water leak and a household where someone had switched to a night shift. It also might help water companies tweak the supply of water to suit demand. When more water flows through a pipe, the pressure increases, which also increases the amount of leakage in a pipe. If companies knew when water demand was lowest, they could reduce supply to match that demand and cut down on the amount of leakage.

If water scarcity does start to hit London, as the Environment Agency predicts, then one temporary solution might be to pipe water from the northeast of England, where water is more abundant.

The driest part of the UK also happens to be the most heavily populated, which piles the pressure on water supplies in the southeast. “What you really want is to get water from the northeast or northwest where there is relatively low density and you get lots of rain,” says Collins. Schemes to pump water from the north to the south of England have been previously been considered by the Environment Agency but so far plans have been rejected as being much more costly than just developing more infrastructure in the southeast. Water is heavy and pumping it over long distances would only ever be a fallback in desperate times, Collins says.

A better solution would be to reduce leakage and demand to try and conserve the water already within the system. Some utility firms are experimenting with sensor-filled pipes that detect changes in water temperature and pressure that might indicate the presence of a leak. And researchers are also experimenting with systems that could predict when pipes are about to burst.

Really tackling water scarcity might mean changing attitudes so people acknowledge that water is a scarce resource. Cities such as Las Vegas have gone through periods of water shortages and residents adapted their behaviour to fit their environment. “People used to have the expectation of having a grass lawn in the desert which is ridiculous,” Collins says. Now it’s much more common for people in Las Vegas suburbs to have gardens of scrub or desert. But in the UK, we’re still not used to thinking about water as a scarce resource. And no matter the amount of money poured into detecting and fixing leaks, that’s something that might not change until things are almost too late. “People will change if they have to,” says Collins. “I’m just not sure if they have to.”

Comments