That aside, Fukushima being the concrete example, you can prepare for a once a century earth quake, and 3 meter tsunami, and then nature throws you a 3.5m tsunami.
Ironically the nuclear power station closest to the epicenter of the earthquake that ultimately destroyed the Fukushima Daiichi NPP did exactly this. Two of the reactors will soon be restarted.
From wikipedia, Onagawa Nuclear Power Plant:
The town of Onagawa to the northeast of the plant was largely destroyed by the tsunami[17] which followed the earthquake, but the plant’s 14 meters (46 ft) high seawall was tall and robust enough to prevent the power plant from experiencing severe flooding. Yanosuke Hirai, who died in 1986, is cited as the only person on the entire power station construction project to push for the 14.8-meter breakwater. Although many of his colleagues regarded 12 meters as sufficient, Hirai’s authority eventually prevailed, and Tōhoku Electric spent the extra money to build the 14.8m tsunami wall.
…
Following the tsunami, two to three hundred residents of the town who lost their homes to the tsunami took refuge in the Onagawa nuclear plant’s gymnasium, as the reactor complex was the only safe area in the vicinity to evacuate to, with the reactor operators supplying food and blankets to the needy.[25] At the time Reuters suggested that the Onagawa nuclear power plant may demonstrate that it is possible for nuclear facilities to withstand the greatest natural disasters, and to retain public trust.[5]
Where Tohoku Electric Power Co. increased the height of their sea wall, Tepco took an opposite approach at Fukushima Daiichi and reduced the height of theirs in order to save money. In the end, the Fukushima disaster didn’t really kill that many people (other than fatalities from the evacuation itself, and certainly in contrast to the massive death toll from the direct effects of the Tsunami), but it did create a mess.
That said, even Chernobyl doesn’t come remotely close to the #1 killer in power generation accidents, however.
From Wikipedia, 1975 Banqiao Dam failure:
The dam collapse created the third-deadliest flood in history which affected a total population of 10.15 million and inundated around 30 cities and counties of 12,000 square kilometers (or 3 million acres), with an estimated death toll ranging from 26,000 to 240,000.[1][3][4][5][6] The flood also caused the collapse of 5 million to 6.8 million houses.[5][7] The dam failure took place when many people were preoccupied with the Cultural Revolution[4]
Sadly, a different attitude was taken during its construction - a strong contrast to the one taken towards Yanosuke Hirai at Onagawa, with disastrous consequences:
Chen Xing, then chief engineer of the dam projects, opposed the ideas of constructing too many dams as well as prioritizing the goal of “retaining water”.[2][5][10] He pointed out that the local geographical conditions made it unreasonable to overly emphasize the reservoir’s function of water storage, because otherwise there was risk of creating serious floods and other disasters such as alkalinization of farm land.[5][6][8] Nevertheless, Chen’s warning was ignored and he was criticized for being a “Rightist” and “Opportunist”; he was subsequently removed from his post and was sent to Xinyang[5][6][8]
The pattern I’m seeing here is not with a particular technology when it comes to prevention of catastrophic failures, but a preference for emphasis on ‘safety culture’.
Another nuclear example, though in the context of the UK’s rushed program to acquire nuclear weapons:
From Wikipedia, Windscale fire:
The presence of the chimney scrubbers at Windscale was credited with maintaining partial containment and thus minimizing the radioactive content of the smoke that poured from the chimney during the fire. These scrubbers were installed at great expense on the insistence of John Cockcroft and were known as Cockcroft’s Folly until the 1957 fire.[42]
Combined with an appropriate response to the disaster: (contrast to Soviet officials not screening milk post-Chernobyl yielding high rates of thyroid cancer)
In the days following the disaster, tests were carried out on local milk samples, and the milk was found to be dangerously contaminated with iodine-131.[65]
It was thus decided that consumption of milk from the surrounding area should be stopped, and eventually restrictions were put in place on the consumption of milk from the 200-square-mile (520 km2) area surrounding the piles.[66] Milk from about 500 km2 of nearby countryside was destroyed (diluted a thousandfold and dumped in the Irish Sea) for about a month.[7] However, no one was evacuated from the surrounding area.
Still, some fatalities:
Their study concluded that because the actual amount of radiation released in the fire could be double the previous estimates, and that the radioactive plume actually travelled further east, there were likely to be 100 to 240 cancer fatalities in the long term as a result of the fire.[3][2]
But nothing close to a similar energy-related incident of the same era:
From Wikipedia, Great Smog of London:
A period of unusually cold weather, combined with an anticyclone and windless conditions, collected airborne pollutants—mostly arising from the use of coal—to form a thick layer of smog over the city. It lasted from Friday 5 December to Tuesday 9 December 1952, then dispersed quickly when the weather changed.[3]
…
More recent research suggests that the total number of fatalities may have been considerably greater, with estimates of between 10,000 and 12,000 deaths.[1][2]