The Aberfan Disaster: How Known Springs, Ignored Slips and Tip 7 Killed 144 People

At Aberfan, near Merthyr Tydfil in South Wales, a colliery spoil tip had been built on a mountainside directly above the village, on top of a natural spring. On 21 October 1966, after three weeks of heavy rain, roughly 140,000 cubic yards (110,000 m3) of saturated waste broke away and flowed downhill as a slurry. It engulfed Pantglas Junior School and a row of houses, killing 116 children and 28 adults.

The Aberfan disaster was not a freak event. The water that triggered it had been marked on maps since 1874. An earlier tip on the same hillside had slid in 1944. The tip that failed, Tip 7, had moved twice in 1963. Residents had complained for years about black, greasy floodwater coming off the slopes. Each signal was recorded and then explained away, deferred or treated as someone else's problem.

The questions an engineer would ask of any failure apply here. What was the hazard? Who knew about it? What did the warnings look like, and why did nobody act on them? The answers below follow the account of the disaster on Wikipedia, which draws on the official inquiry.

A village built around a colliery

Aberfan sits toward the bottom of the western slope of the Taff Valley, on the eastern slope of Mynydd Merthyr, about four miles (6 km) south of Merthyr Tydfil. When John Nixon and partners sank the Merthyr Vale Colliery on 23 August 1869, Aberfan was two cottages and an inn used by local farmers and bargemen. By 1966 it had about 5,000 people, and most of them worked in the coal industry.

The British coal industry was nationalised in 1947, and from then on the colliery was run by the National Coal Board (NCB). Regulation was the job of HM Inspectorate of Mines. The inspectors, however, had worked as engineers in the coal industry and were former NCB employees. The regulator was staffed by people who came from the organisation it was meant to regulate. That matters when you look at how the warning signs were handled.

The layout of the valley set up the hazard. The River Taff runs north to south through the village. On the western outskirts, the bed of the disused Glamorganshire Canal and a disused railway embankment run alongside the river. The first spoil from the mine went onto the lower slopes, east of the canal. In the 1910s, tipping moved to the western slopes, above the canal line and above the village.

Seven tips above one village

By 1966 there were seven spoil heaps on the hillside, holding about 2.6 million cubic yards (2.0 million m3) of waste. Tips 4 and 5 were conical mounds at the top of the slope. Tip 4 was already misshapen from an earlier slip. The other five sat lower down. All seven were directly above the village.

Only Tip 7 was still in use in 1966. Work on it began in 1958. By the time of the disaster it stood about 111 feet (34 m) high and held 297,000 cubic yards (227,000 m3) of spoil. Its make-up mattered:

• Volume of tailings: about 30,000 cubic yards (23,000 m3) of the tip was tailings.
• What tailings are: waste from the chemical extraction of coal, made of fine particles of coal and ash.
• How they behave wet: when saturated, they act much like quicksand.

So the one active tip contained a layer of fine material that loses its strength when wet. In contravention of the NCB's own procedures, that tip was also partly built on ground where springs came to the surface.

Water: the hazard everyone knew about

Water controls whether a tip stays put. Tips 4, 5 and 7 at Aberfan had all been sited on streams or springs. This was not hidden knowledge. Local people knew about the springs, and they had appeared on Ordnance Survey and Geological Society maps since 1874, nearly a century before the collapse.

The climate made it worse. Aberfan gets a lot of rain, about 60 inches (1,500 mm) a year on average. In 1960 it got 70.5 inches (1,790 mm), the wettest of the years leading up to the disaster. Tip 7 was a pile of loose spoil, with a quicksand-prone layer, sitting on springs in one of the wetter parts of Britain.

The hillside was also sending water into the village. Between 1952 and 1965, the Pantglas area flooded badly at least 11 times. Residents said the floodwater was black and left a greasy residue when it went down. People complained to Merthyr Tydfil County Borough Council, and the council wrote to the NCB about it between July 1963 and March 1964. The people living below the tips were reporting, in plain language, what the water coming off them looked like.

The warnings: Tip 4 in 1944, Tip 7 in 1963

Aberfan had already had a tip failure. Tip 4 was used from 1933 to 1945 and was started on boggy ground between two streams. When it was planned, the Merthyr Tydfil borough engineer judged that, despite the site, it was unlikely to avalanche. In the early 1940s the ground moved, and a drainage channel was dug in early 1944. In November 1944 part of the tip slid 1,600 feet (490 m) down the mountain. It stopped about 500 feet (150 m) above the village.

The pattern was a tip on wet ground, ground movement, a drainage fix, and then a large slide anyway. It stopped short of the houses, which may be why it was treated as survivable rather than as proof of what these tips could do.

Tip 7 then gave its own warnings. In May 1963 it shifted slightly. In November 1963 there was a bigger slide. The NCB said this was not a 'slide' but a 'tailings run', meaning tailings running off the surface of the tip, and that the tip's stability was unaffected. The NCB stopped putting tailings on Tip 7 but kept tipping ordinary spoil on it. Calling the event a surface run-off rather than a slope failure meant the tip could stay in use with only a small change to how it was operated.

21 October 1966

After three weeks of heavy rain, Tip 7 was saturated. Water had built up inside it, fed by the rain and by the spring underneath. The tip then gave way suddenly. About 140,000 cubic yards (110,000 m3) of spoil came down the hillside as a slurry, not as a slow creep, and poured onto the Pantglas area of the village.

The main building it hit was Pantglas Junior School, where lessons had just started. Five teachers and 109 children died there. In total, 116 children and 28 adults were killed as the slurry buried the school and a row of houses.

Each element of the failure had been seen before. The springs were on the maps. Saturation followed heavy rain, which was normal for the area. Fine tailings turn to slurry when wet, which the NCB knew well enough to stop tipping them on Tip 7 three years earlier. And the slope had already moved twice. What was new was that the hazards all came together, at full scale, during a school day.

The inquiry and who was held responsible

The official inquiry was chaired by Lord Justice Edmund Davies. Its report put the blame squarely on the NCB, and it named the organisation and nine of its employees.

The NCB's chairman, Lord Robens, was criticised for making misleading statements. He was also criticised for not being clear about what the NCB had known about the springs on the hillside. Given that the springs had been on public maps since 1874, the question was never whether the information existed. It was whether the organisation had used it.

Even so, neither the NCB nor any of its employees was prosecuted, and the NCB was not fined. For anyone who thinks about how institutions learn from failure, this is the uncomfortable part. A public inquiry found the operator responsible, named individuals, and no legal penalty followed.

Clearing the remaining tips, and who paid

The Aberfan Disaster Memorial Fund (ADMF) was set up on the day of the disaster. It received nearly 88,000 donations, totalling £1.75 million.

The other tips were still standing above the village after the collapse. Residents had to fight for a long time to get them removed. The NCB and the government resisted on the grounds of cost. In the end, the clearance was paid for partly by a government grant and partly by £150,000 forcibly taken from the memorial fund. Money given by the public for the bereaved and the community was used to remove a hazard created by the organisation that had just been blamed for the deaths.

It took decades to put that right:

• 1997: the British government paid the £150,000 back to the ADMF.
• 2007: the Welsh Government gave £1.5 million to the fund and £500,000 to the Aberfan Education Charity, to make amends for the money wrongly taken.

The long tail of harm

The death toll does not show the full cost. Many people in the village developed medical problems because of the disaster. Half of the survivors have had post-traumatic stress disorder at some point in their lives.

Any honest account of a failure has to include this. The event itself took a morning. The fight to clear the tips, the use of the memorial fund, and the long-term health effects lasted for decades. A failure analysis that stops at the moment of collapse leaves out most of what the failure cost.

What to watch for

Aberfan was a civil and mining failure, not a software one, and the comparison only goes so far. You cannot roll back a slope, and the stakes here were children's lives. Readers who want the geotechnical detail should go to the Edmund Davies inquiry report itself, rather than relying on any summary. Still, the way the warnings were handled is familiar to anyone who has run a post-incident review:

• Known hazards treated as background. The springs were on maps from 1874. Information that is always available tends to stop being noticed.
• Renaming an incident to shrink it. Calling the November 1963 movement a 'tailings run' allowed a slope failure to be closed with a small change to operations while tipping continued.
• A near miss mistaken for safety. Tip 4 stopped 500 feet above the village in 1944. Where it happened to stop was luck, not a safety margin.
• Oversight drawn from the overseen. When inspectors are former staff of the operator, they tend to share its assumptions about risk.
• Outside reports dismissed. Eleven floods and repeated complaints about black, greasy water were evidence from the people with the most to lose.
• Cost as an argument after the event. Even after 144 deaths, removing the remaining tips was resisted on cost grounds.

Look for these patterns in your own systems, before something forces you to.