Last week’s crisis at the Fukushima nuclear power plant in Japan saw radioactive water leak again from the crippled facility, raising fears that groundwater flowing into the Pacific Ocean could be contaminated. The Japanese government also raised the international incident level – the scale used to assess nuclear accidents - from one to three out of seven. The original nuclear meltdown following the 2011 Japanese earthquake was scaled seven.
Even if Fukushima was ultimately caused by the 2011 earthquake and ensuing tsunami, accidents such as this beg the question: can nuclear energy ever be truly safe?
There are three reasons to think that nuclear accidents are common, and could increase - and it’s not because of the technology. Let’s have a look at the evidence.
Lessons from history
In the early 1980s, Yale sociologist Charles Perrow argued that the partial meltdown of a nuclear reactor at Three Mile Island was a “normal accident”. The crux of his argument was that complicated technological systems have unavoidable problems that can’t be designed around.
Perrow’s argument — still relevant today — rested on three pillars. First, people are fallible, even at nuclear reactors. Operator error is still a very common factor in incidents and accidents.
Second, big accidents almost always have very small beginnings. Nuclear power plants are so complex that relatively simple things — shirt tails, fuses, light bulbs, mice, cats, and candles — can disrupt the entire system.
And finally, many failures are those of organisations more than technology. Given the right event, all these factors can lead to system-wide failure. Perrow concludes that such high-tech, dangerous systems are hopeless and should be abandoned, as the inevitable risks of failure outweigh any conceivable benefits.
Nuclear reactors do have inherent advantages over fossil fuels, but Perrow’s argument raises serious questions about nuclear safety.
Even so, Perrow was writing in the 1980s. Surely things have improved since then? Well, perhaps not.
If you consider the full range of incidents and accidents reported on the International Nuclear Event Scale, there have been hundreds of events over the past few decades. One peer-reviewed study identified 105 nuclear accidents totalling U$176.9 billion in damages and 4,231 fatalities worldwide from 1952 to 2011. The International Atomic Energy Agency also reports no less than 2,400 separate incidents since the organisation began collecting data in the 1950s.
Most of these incidents involved no major releases of radiation or fatalities. But three emerging trends still cause reason for grave concern.
First, major modern nuclear power accidents are no longer one-off events. Instead, they can span years or even decades, creating a sort of “continuous accident”.
The infamous Chernobyl nuclear disaster may have started on April 25 1986, but it continued into the early 1990s. Secrecy, further accidents, and wildfires in the exclusion zone meant that exposure to dangerous levels of radiation weren’t controlled immediately.
We can see this same “continuous” trend with the accident at Fukushima. The triple meltdown itself at Fukushima in March 2011 was just the beginning.
In March 2013 a power outage left four underground spent fuel pools without fresh cooling water for several hours. The same month, it surfaced that a TEPCO crew laying down rat-proof netting caused another outage. In April 2013 regulators discovered that thousands of gallons of radioactive water had seeped into the ground from a leaking system of plastic sheeting.
In May, a fire broke out near Fukushima Unit 3 — ostensibly caused by cardboard boxes catching flame. And most recently in August 2013, regulators announced that 300 tons of radioactive water was found leaking from storage tanks.
New designs, new problems
There is some evidence that newer reactor designs and systems are more prone to accidents. Dennis Berry, Director Emeritus of Sandia National Laboratories, explains that the problem with new reactors and accidents is twofold: scenarios arise that are impossible to plan for in simulations, and people make mistakes.
As he put it:
Fabrication, construction, operation, and maintenance of new reactors will face a steep learning curve: advanced technologies will have a heightened risk of accidents and mistakes. The technology may be proven, but people are not.
Former nuclear engineer David Lochbaum has noted that almost all serious nuclear accidents have occurred when operators have little experience with a plant. This makes new systems incredibly risky.
Lochbaum cites numerous historical examples of nuclear reactor accidents, including Three Mile Island and Chernobyl, which suffered accidents immediately or soon after opening. Only Fukushima seems to have defied the trend; it was opened in 1971 and continued operating until the 2011 earthquake.
The third problem is electric market restructuring. This puts more pressure on nuclear operators to keep costs low, potentially compromising safety.
The problem is, as former Nuclear Regulatory Commission chair Peter Bradford states, “nuclear energy can be cheap, or it can be safe. But it can’t be both.” And even then, “there’s always the possibility somebody will cut a corner”.
For example, the pressure to build new generators on existing sites to avoid finding new locations can increase the risk of catastrophe, since there is a greater chance that one accident can affect multiple reactors.
Nuclear waste storage is also becoming more dangerous, with many spent fuel pools packed with more fuel rods to keep costs low, making them hotter and denser. Operators have to add boron to water pool to absorb neutrons, increasing the risk of chain reaction, or criticality, accidents.
The industry has also been trying to tinker with reactor sizes and promote designs that operators have little experience with, making operator training a factor. Some of these new reactor designs use more fuel and create more heat, meaning they have bigger cores containing larger quantities of dangerous fissionable materials, increasing the magnitude of any accident that could occur.
These factors are worrying (to say the least) given the severity of what a single, serious accident can do. Too bad it seems a matter of when, not if, we will see more of them in the future.