Dozens of flights have been cancelled and many others delayed after a computer failure at the headquarters of air traffic control company Nats.
For a time there were no flights able to take off or land at some UK airports and although Nats were able to resolve the problem by around 4pm, delays and cancellations were expected to drag on into the evening.
Transport Secretary Patrick McLoughlin said the disruption was unacceptable.
He went on: “Any disruption to our aviation system is a matter of the utmost concern, especially at this time of year in the run up to the holiday season.
“Disruption on this scale is simply unacceptable and I have asked Nats for a full explanation of this evening’s incident. I also want to know what steps will be taken to prevent this happening again.”
Today’s problem at Nats’ state-of-the-art £700 million centre at Swanwick is by no means the first glitch at the Hampshire centre.
Having resolved the problem, which started around mid-afternoon, Nats said: “It will take time for operations across the UK to fully recover so passengers should contact their airline for the status of their flight.
“We apologise for any delays and the inconvenience this may have caused.”
Heathrow said 50 flights had had to be cancelled at the west London airport and warned that this figure could rise and that delays could persist. A spokesman added that there could be delays lasting into tomorrow.
Problems were also reported further afield, with delays of 15 minutes affecting some flights at Leeds Bradford Airport.
Later, Nats said: “We are investigating the cause of this fault but can confirm that it was not due to a power outage.”
Airports as far north as Aberdeen and Edinburgh were hit by the computer problem. Other airports that reported delays included Birmingham, Manchester, Luton and Bristol.
There were also hold-ups at Stansted and Gatwick but these two major south east England airports appeared to have escaped more lightly than Heathrow where the rate of 80 to 90 flights handled every hour means even a comparatively short systems failure leads to severe disruption.
Which? executive director Richard Lloyd said: “The Denied Boarding Regulations provide a means for consumers to claim compensation for delays and cancellations.
“However if an airline is able to demonstrate the cancellation or delay was due to extraordinary circumstances then they don’t have to pay out.
“Passengers may however still be entitled to refreshments, free phone calls and overnight accommodation depending on how long the delay is, how far you are flying and whether the flight is to or from an EU or non-EU airport or on an EU or non-EU carrier.”
The air traffic control centre at Swanwick has endured a chequered history.
Politicians often boast about big projects coming in “on time and on budget”. In Swanwick’s case, the project was very much late and very much over budget. And today’s problem is the latest in a series of technical hitches that have plagued plane passengers in recent years.
With air traffic controllers (ATC) cooped up in a comparatively old-fashioned headquarters in West Drayton in west London, it was decided in the late 1980s that it would be a good idea to have spanking new ATC centre at Swanwick.
Originally the cost was going to be £132 million and the move-in date was scheduled for 1997.
But various software problems that had to be overcome meant the cost rocketed and the switch over from West Drayton was constantly delayed.
Eventually Swanwick opened in January 2002 - at a cost of £700 million.
One of the most recent computer problems was just over a year ago when a technical problem on December 7 - a busy Saturday just before Christimas - led to much flight disruption.
Professor Ian Allison, head of the school of computing science and digital media at Robert Gordon University in Scotland, spoke of today’s problem.
He said: “This failure is most surprising as these systems have high levels of resilience built into them. Software is written to a high level of engineering to ensure the safety of passengers.
“On this occasion it seems to be a server failure - i.e. an infrastructure failure on which the software is running. I am sure the systems will have been designed to cope with this type of failure - normally to ‘fail over’ safely to another mirror server.”
He went on: “So it seems to have been a major incident that caused the contingency plans to fail as well as the primary system. “