Ride out the storms – TfL’s flood prevention strategy

12 September 2016

As global warming exerts a palpable influence on the UK’s climate, Transport for London is looking to safeguard its warren of stations and tunnels from an increased risk of flooding. Greg Noone talks to Ian Uttley, a senior drainage engineer with London Underground, about the methodology behind the organisation’s latest flood-risk assessment for the network, and how it might be protected in the future.

London has more than one river running through it. In addition to the Thames, half a dozen ancient watercourses and canals lie bricked over and half-forgotten beneath the city. One of the most prominent, the Westbourne, was dammed in 1730 on the orders of Queen Caroline to form the Serpentine and beautify Hyde Park, only to be diverted when the river water began to pollute the lake. The legacy of that decision might still be glimpsed on the ceiling of Sloane Square station, where a giant iron siphon conveys the Westbourne up and over the District and Circle lines. The pipe itself is a repair job; in June 1885, its predecessor burst open after a bout of heavy rainfall, showering the platform in three feet of sewage.

Ten minutes away from this monument to London’s ancient watercourses is the department for pumps and drainage at Transport for London (TfL). For the most part, the desk of its senior drainage engineer, Ian Uttley, would remain indistinguishable from the rest of the open-plan workspaces were it not for a few scattered clues: a shiny, incongruously large steel screw lying on its side; an enormous, coffee-stained clay mug inscribed ‘Drain Me’; and, nearby, some ribbed black piping shielding the thin trunk of a yellowing potted palm tree.

Uttley is the co-author of the ‘London Underground Comprehensive Review of Flood Risk,’ or LUCRFR for short – “not particularly lovingly,” he says, somewhat dolefully. The project was established in 2014 as the first comprehensive review of flood risks on the London Underground network since the ‘Flood Mitigation Project’ (FMP) in 2001. When the summary conclusions of the investigation were released in April, multiple news outlets seized on the report’s assessment that 57 tube stations were at a high or increased risk of flooding, the subtext therein that the effects of global warming were now being felt on London’s transport system.

That isn’t necessarily untrue, Uttley admits, but it somewhat obscures the devil in the detail of a project like LUCRFR. Outsized in its ambition and level of intricacy, the project has allowed London Underground to precisely quantify flood risk across one of the largest and most complex mass-transit networks in the world.

“Improvements in hydraulic modelling software and the computational ability of hardware, and so on, allowed us to examine certain hazards for the first time, and other hazards with a much finer resolution,” says Uttley.

Model risk

The engineer ascribes his entry into the profession to serendipity. “I started working when I was 18, and I did day release and three evenings a week for four years in civil engineering with a structures bias,” he says. “And I suppose it was my first role in which the sponsorship on the vocational course was available as part of my job. I’d have to say that was probably accidental, that I fell into that role.”

Nevertheless, hydraulic engineering became his métier. Having begun his career in coastal defence and working for water authorities at the district council level, Uttley joined the engineering team behind the Ashford section of the Channel Tunnel rail link in 1997, specialising in designing watercourse culverts and track-drainage systems. Seven years later, he was recruited as a lead design engineer at London Underground’s PPP, before joining TfL in 2010. 

“It was fairly soon after joining that I was asked to produce the briefing paper putting forward our plans for how we would manage flood risk across the business,” says Uttley. “It was reasonably well received, but in a funding-constrained environment, it didn’t immediately receive much attention. It took a couple of external jolts to the business – one in the shape of the Central Line flooding that occurred in June 2012, and then Hurricane Sandy, which hit the New York subway network in October of that year – to really wake up the business.”

It wasn’t that TfL was particularly worried about a tidal surge; while the New York subway system has to permanently factor the Atlantic Ocean into its flood risk surveys, the last serious inundation of the London Underground from fluvial flooding occurred in 1928 and, since then, the construction of the Thames Barrier has put paid to all but the most serious surge events from the North Sea. Nevertheless, experience has taught that pluvial flooding and burst water mains have the potential to do just as much damage to individual stations and assets across the network. In 2012, it took just one burst main to flood the section of line between Mile End and Stratford stations with nearly three million litres of water. The Central line – one of London’s busiest – was paralysed for over 26 hours.

As it currently stands, if a major flooding event were to strike the London Underground network, the responsibility for dealing with it at the site level is largely devolved to individual station managers, who, in turn, consult established procedures laid out in emergency planning documents. Those measures are generally only activated upon notification from the London Underground Control Centre and typically consist of flood boards.

Warning times can vary. If the event is fluvial, the two days it would take for the Thames catchment area to fully overflow means that the requisite evacuation of the network could be undertaken at a pace one might define as leisurely. If severe rainfall is predicted, however, even the Met Office’s most advanced computers often cannot provide a window of more than a few hours. In that sense, Uttley explains, “there may be some potential for safety consequences in relation to extreme pluvial events in a local context.”

Extra measures

Until LUCRFR was conceived in 2011, TfL was constrained in its understanding of the susceptibility of its major network assets to flooding, in that the FMP had only examined and mitigated safety consequences in line with the organisation’s statutory obligations. This limitation was exacerbated by the fragmentation of the FMPs available archives, with interested station managers and hydraulic engineers having to effectively “fish out a document associated with their particular location, review what the exposure was in 2001 and update it”, according to Uttley.

Now, after five years of survey work at over 800 sites, including stations, shafts, tunnels, depots and sidings, LUCRFR not only allows TfL to consider the safety and business consequences for the network, but has also fully updated the schemata provided by the FMP and united them into what the organisation calls the Flood Risk Management Web Geographic Information System (FRM WebGIS.)

“It allows the user to interrogate lots of different geospatial datasets (such as pluvial flooding, tidal flooding, and London Underground asset locations and heights) and overlay, interrogate and effectively intersect them easily,” Uttley explains. “So, in terms of dynamically responding to threats, we’ve delivered a system to the business that allows it to indulge in an active form of flood-risk management.”

LUCRFR also allowed Uttley and his colleagues to confirm the order of precedence for flood risks across the network. Burst water mains were determined to account for 58% of total risk exposure; pluvial for 27%; fluvial for 13%; and direct connections – man-made buried assets that could flood a network tunnel or shaft – for 3%. Within that, Uttley’s team could quantify the risk associated with each asset under examination through hydraulic modelling, which – for the first time – could also take into account the risk from pluvial flooding.

“The volume of water that might enter a station was considered, as well as where that water might go and how many trains might be delayed or stranded,” explains Uttley. “There’s a monetary value associated with each of those consequences. You can multiply that value by the average frequency of the event, then multiply the damages by this figure to get a risk value in monetary terms.”

It has resulted in some startling placements for major TfL assets, with London Bridge, King’s Cross and Waterloo topping the list of sites judged to be most at risk. “If we look at the number-one site, for example – London Bridge – the total risk value is £403,000 a year, of which £126,000 results from burst water mains and £277,000 from pluvial,” says Uttley.

I hesitatingly mention that London Bridge is the station I use to get home, starting to imagine waves gently lapping St Thomas Street.

“I’ll try to disabuse you of that notion,” he laughs, reminding me that the costliest predictions for certain sites are partly down to their high rate of use by passengers rather than their innate level of vulnerability. “Naturally, we’re going to examine them in more detail.”

Phase two

Since the publication of the final report in April, the second phase of the project has begun in earnest. This is intended to diminish the margin for error in many of the assessments of individual sites, which, given the sheer scale of LUCRFR, could often only be reached through the extrapolation of statistically derived data.

“In relative terms, our results are reasonably accurate and can act as the foundation for the next stage of the project,” assures Uttley. “We can say that, in general, all the sites that are at the top [of the risk table] will probably still be at the top if we obtain a much finer resolution.”

Any additional investment in flood defences at the top ten most-affected sites will be dependent on the results of those investigations. However, given the budgetary constraints TfL is likely to face in the wake of the government’s grant reductions and Mayor Sadiq Khan’s freeze on fares, it is unlikely that we will see a sustained effort of the same scale and complexity as phase one in the near future.

“The programme for phase two is three years, but funding is allocated on an annual basis, so we don’t have any guarantee that we’ll get funding in the second year,” Uttley says.

He is open to the possibility that the assumptions in the initial model might have been too conservative, rather than stark, but says, “you know, there was an overriding engineering principle at the heart of this analysis, which was that your error should always be on the safe side.” It’s the sort of statement that not only summarises the drivers behind LUCRFR, but also belies its achievement. For the first time, London has a statistical model for flood risk in a format that is easy to interpret, update and share, and, in that sense, is emblematic of a worldly city shrewdly confronting the ongoing threat posed by climate change.

Flood barriers at Shepherd’s Bush Station, London.
A geographical diagram representing flood risk in the City of London.

Privacy Policy
We have updated our privacy policy. In the latest update it explains what cookies are and how we use them on our site. To learn more about cookies and their benefits, please view our privacy policy. Please be aware that parts of this site will not function correctly if you disable cookies. By continuing to use this site, you consent to our use of cookies in accordance with our privacy policy unless you have disabled them.