A police helicopter crashed into the Clutha Vaults Bar in Glasgow on 29 November 2013. The pilot, two police officer passengers and seven in the bar were killed. The Air Accidents Investigation Branch published its final report last week. Relatives of those who died had been briefed in advance. They said that they were doubly disappointed.

First, because the AAIB Inquiry report concluded that:

the investigation could not establish why a pilot with over 5500 hours flying experience in military and civilian helicopters, who had been a Qualified Helicopter Instructor and an Instrument Rating Examiner, with previous assessments as an above average pilot, did not complete the actions detailed in the Pilot’s Checklist Emergency and Malfunction Procedures for the Low Fuel 1 and Low Fuel 2 warnings.

The low fuel warnings came from the tanks that directly supplied the two engines. The pilot acknowledged them five times, but did not turn on the transfer pump from the main tank, which contained fuel, or land within ten minutes. He should have done both. The warnings started sounding twenty minutes before the crash. One of the engines flamed out because of fuel starvation forty seconds before the crash, the other eight seconds before.

Second, the relatives were unhappy because the main recommendations of the investigation, that flight data recorders should be installed on small police and medical helicopters, do not carry the immediate force of law. But the AAIB does not make the rules. They come from the Civil Aviation Authority and the European Aviation Safety Authority. And there are political tensions. The Scottish government is putting pressure on the recipient of the AAIB report, the UK Secretary of State for Transport, regarding the implementation of the recommendations.

The public has a touching faith in the ability of inquiries to answer questions, learn lessons and prevent disasters being repeated. But sometimes they can’t do the first because the evidence is missing. The Clutha helicopter crash may well join the 1906 Grantham railway accident as an exhaustively investigated but unresolved – and unresolvable – mystery.

The night mail from London to Edinburgh was due to stop at Grantham at 11 p.m. on 19 September 1906. The points north of the station were set to divert another train to the Nottingham line. The signals were red. But the train roared through, derailed and caught fire. The driver and fireman, eleven passengers and a postal worker died. The driver had 18 years driving experience and had been in sole charge of the two-year-old engine since its construction. There were plenty of rumours: he was drunk, or had gone mad, or had collapsed, or had started a fight with the fireman. But the signalman in the box that the train passed just before the crash had seen them standing motionless on either side of the footplate, looking ahead. As L.T.C. Rolt said in his classic 1955 book about rail disasters, Red for Danger, ‘what precisely took place on the footplate of Number 276... is a question that Sherlock Holmes himself could not answer.’ It will probably be the same regarding events in the police helicopter on 29 November 2013. Even a flight data recorder (FDR) might not have provided the necessary information.

The BEA Trident that crashed 120 seconds after taking off from Heathrow on 18 June 1972 (the deadliest civil aircraft disaster in the UK, with 118 deaths) had an FDR. It showed that the plane was stalling, that the warning system had operated three times, but was turned off 22 seconds before the aircraft dived into the ground. The reason why remains a mystery. There were four pilots in the cockpit. The Trident FDR did not record images or audio. Even if it had, the only words might well have been ‘oh shit!’ just before the crash. It could have been the same over the Clutha.

As for learning lessons, we do that well, but we’re good at forgetting them too. I investigated the 1996 E. Coli O157 outbreak in central Scotland. The bacterium killed 17. The rules for butchers were changed. But in 2005 another butcher was ignoring them. There was another outbreak. A little boy died. I chaired another inquiry, and concluded: ‘I had hoped that the lessons of the shocking event in 1996 would stay in people’s minds. But comparison of the failures that led to this outbreak in South Wales with those in the outbreak in Scotland showed that this has not been the case.’