Disabled passengers face the sharp end of TfL lift failures, with 45 outages lasting more than a week in just 80 days

TfL lifts were out of service for more than a week on 45 occasions in just 80 days, according to analysis by Tube Alerter.

The data shows a network that works most of the time, but fails badly in the worst cases. For passengers who cannot take the stairs, a single broken lift can turn a step-free station into one they cannot use.

Key findings

• Tube Alerter tracked 567 lifts appearing in TfL's lift-disruption and step-free-access data across 201 stations between 9 February and 30 April 2026.
• TfL lifts were available for 96.12% of lift-hours, but that headline figure hides severe local failures.
• Among resolved outages, 45 lasted more than a week and 21 lasted more than two weeks. A further 12 outages were still ongoing at the time of analysis.
• The median outage was 5.7 hours, but the mean was 30.8 hours, dragged upwards by a long tail of prolonged failures.
• 26 step-free stations in TfL's lift dataset have only one recorded lift. Where that lift is required for the step-free route, a single failure can remove step-free access entirely.
• Barking's single lift was recorded as unavailable for the equivalent of 60 of the 80 days analysed.
• Wimbledon Park's single lift failed 38 times in 80 days, roughly once every 50 hours.
• Some step-free closures were not caused by lift faults at all, but by staff unavailability.
• Other major cities, including New York, Washington and Toronto, publish lift or elevator availability as a performance metric. London does not publish an equivalent historic station-by-station reliability view.

The basic network number looks respectable. Across the period analysed, TfL lifts were available for 96.12% of possible lift-hours, with an average of 22 lifts unavailable at any moment. The typical outage was also not especially long: 77% were fixed within a day, and the median outage lasted under six hours.

But averages are the wrong way to understand accessibility. A wheelchair user does not experience "96% uptime" across the network. They experience the one lift they need, at the station they are trying to use, at the moment they arrive.

"A broken lift can seem trivial to many people, but it can completely change my day. When a lift is out at King's Cross, navigating its maze of tunnels is incredibly stressful, and has previously forced me to take a far longer journey."A passenger who relies on step-free access, and whose experience helped prompt Tube Alerter's analysis.

That is where the data becomes much less reassuring. Among resolved outages, Tube Alerter recorded 45 lasting more than a week and 21 lasting more than two weeks. A further 12 outages were still unresolved at the time of analysis. The longest recorded outage, at West Silvertown DLR station, lasted 1,574 hours, or 65.6 days. TfL said this was due to lift replacement works between November 2025 and April 2026, with alternative route and assistance information displayed at the station and through TfL journey-planning channels.

Single-lift stations: where one failure can mean total loss

The clearest accessibility risk is at single-lift stations. Tube Alerter identified 26 step-free stations in TfL's lift dataset with only one recorded lift. Where that lift is required for the step-free route, there is no redundancy: if the lift goes, step-free access goes with it.

Barking is the clearest example. Its only lift was recorded as unavailable for 1,442 hours in the 80-day period, equivalent to 60 days. As of 30 April, the latest outage had been running for almost 39 days.

Tube Alerter's archived TfL message for the station stated: "No Step Free Access - Step free access is not available due to a faulty lift." TfL said Barking station is managed by c2c, not TfL, but that it works with c2c and other operators to help resolve lift issues.

For passengers, the distinction is unlikely to matter much. Barking appears in TfL's lift-disruption data and is served by TfL services. If the only lift is unavailable, a journey that looks step-free can become impossible without rerouting or assistance.

Nor was Barking an isolated case. Among single-lift stations, Tube Alerter recorded 982 hours of downtime at Hendon Central, 643 at Hounslow West, 551 at Queen's Road Peckham, 421 at Caledonian Road & Barnsbury, and 308 at Wimbledon Park.

Wimbledon Park: 38 failures in 80 days

At Wimbledon Park, the problem was not one long breakdown, but repeated failure.

The station's only lift was recorded as unavailable 38 times in 80 days, roughly once every 50 hours. In several cases it returned to service, only to fail again within a few hours.

For passengers who rely on the lift, that can be harder to plan around than a single long closure. A lift that is working in the morning may not be working by the time they travel home.

Multi-lift stations and the route question

West Silvertown shows how a station can remain partly open while still failing a specific accessible journey. Lift 2, serving step-free access towards Woolwich Arsenal, was recorded as unavailable from 13 February to 20 April. The opposite-direction lift remained available, so this was not a total station closure. But for passengers needing step-free access in that direction, the disruption lasted more than two months.

Not every lift outage makes a station completely inaccessible. At larger interchanges, another route may still be available. But for passengers who rely on a specific lift, the practical question is simpler: can they complete their journey, or not?

The worst cases are not abstract engineering failures. They are failures of a public promise. A step-free symbol on a map is only meaningful if the lift required for that journey is working when the passenger arrives.

When the problem is staffing, not the lift

Some cases raise an even sharper question. At Bromley-by-Bow and Finchley Central, Tube Alerter's archived messages attributed loss of step-free access to "unavailability of station staff", not a mechanical lift fault.

That makes the problem broader than maintenance. Step-free access can disappear because a machine breaks, but it can also disappear because the staffing model needed to support access is not there. Disability News Service previously reported that staff shortages had repeatedly shut down step-free access at Tube stations, showing this is not a purely theoretical risk.

TfL confirmed that staff unavailability can lead to lifts being taken out of service, saying staff are trained to assist customers if a lift fails. It said it is working to prevent lifts being removed from service due to staff shortages and has an ongoing recruitment campaign to fill vacancies.

Self-reporting tech is helping, but rollout is partial

TfL has started to address part of the problem. In December 2025, it said 93 lifts at 28 Tube stations had been fitted with self-reporting technology, meaning lifts can report themselves as out of service without waiting for station staff to log the fault. TfL said this should update journey-planning tools more quickly and help staff seek assistance sooner.

Tube Alerter's data suggests this technology may matter most for the worst cases. At the 10 publicly confirmed self-reporting stations in the dataset, the median outage was similar to other stations, 5.2 hours versus 5.9. But the mean outage was much lower: 11.0 hours compared with 33.0 hours.

That does not prove self-reporting causes faster repairs, because those stations are skewed towards central London. But it is consistent with the idea that quicker reporting helps prevent faults sitting unresolved for days.

The transparency gap

The international comparison makes London look less transparent than it should. New York's MTA publishes subway elevator availability as a formal performance figure. Its 2024 annual report put subway elevator availability at 97.44%, while New York City's Accessible NYC 2025 report said elevators had a 97.6% availability rate through 30 November. The MTA also publishes elevator and escalator availability as open data.

Washington's Metro is even clearer. Its FY24 Service Excellence Report says rail elevators had 98.2% availability, against a target of 97.7%, with 314 out of 320 elevators in service at any time. It also reports that mean time to repair reached just over five hours in May.

Toronto's TTC also reports elevator availability. Its January 2025 KPI report says subway elevator availability for November was 98.6%, against a 98% target, and defines the measure as the percentage of total available elevator service hours during subway service. It also publishes live elevator status through service alerts, screens, and its Lift Line.

Other systems publish partial or different accessibility information. Paris's RATP publishes live elevator status by line, Tokyo's Toei Subway says all stations have at least one barrier-free route from street to platform, and Singapore publishes detailed rail reliability reporting, although not a directly comparable lift-availability figure. Hong Kong MTR includes passenger lift reliability among its annual operating metrics.

These are not perfect like-for-like comparisons. Different cities count different things. Some include planned works, some exclude long-term replacement projects, some report escalators with lifts, and some focus on live status rather than historic reliability. But that is precisely the point. London has live lift disruption data, but passengers do not get an easy public view of which lifts fail most often, how long they stay broken, or which step-free stations are most fragile.

TfL also has no maximum allowed time for station lift repairs. In a December 2025 FOI response, TfL said most lift contracts have a 99% asset availability KPI, with financial penalties if contractors miss targets, but there is no maximum repair time for an individual lift.

A network can meet a headline target while a small number of passengers face weeks of disruption at the station they actually need.

TfL response

A TfL spokesperson said: "We recognise how important lifts are to our customers and apologise for the disruption caused to customer journeys when lifts are unavailable. Lift unavailability can be caused by unexpected failures, which we work hard to fix and return to service as quickly as possible. Staff unavailability can also lead to lifts being taken out of service as our staff are trained to assist customers should a lift fail. We are working hard to prevent lifts being removed from service due to staff unavailability and have an ongoing recruitment campaign to fill outstanding vacancies. Although our lifts are in service more than 96 per cent of the time, we strive to always keep them in service to ensure that customers can travel across our city."

TfL said that where a London Underground lift is out of service, staff will help customers plan an alternative step-free journey and, if there is not a reasonable alternative route, book a taxi at no cost to the customer. On the DLR, where most stations are unstaffed, passengers can use emergency alarms on platforms to seek assistance, and signage is being rolled out next to lifts with alternative route information and contact details.

Conclusion

The data does not show a network-wide collapse. Most lifts work most of the time. But it does show a gap between headline availability and passenger reality, made harder to scrutinise because TfL does not publish historic lift reliability by station. For disabled passengers who depend on a single lift, "most of the time" is not enough when the one they need is out of service for days or weeks.

Further reading

Data

Every Station, Every Year: London's Ridership Data

Annual entry and exit figures for 515 stations from 2017 to 2025, with covid recovery and weekend vs weekday charts.

Data

TfL Fare Tracker: 21 Years of Tube and Bus Prices

How TfL fares have changed from 2005 to 2026, with every zone, cap, and bus fare in one place.

Analysis

When Will the Tube Be Fully Automated?

The Victoria, Central, Jubilee and Northern lines already run automatically. What's holding back the rest of the network?