Let’s Build a Nuclear Reactor! Accidents Will Happen

Related articles

There have been three major accidents at commercial nuclear reactors – Three Mile Island (1979), Chernobyl (1986), and Fukushima (2011). Let’s take a look at each of these accidents to see what happened.




Before we do the analysis, you need to understanding coolants, moderators, reactor controls, and containment. Not sure, then begin here, where I explain those terms. 




How a reactor plant works

Pressurized Water Reactor (PWR)

Boiling Water Reactor (BWR)

These two drawings show the two most common types of reactors used to produce electrical power. If you take a close look, you’ll see only one fundamental difference between these two designs.

  • In the PWR, the water that turns to steam and spins the turbines is brought to a boil in a steam generator
  • The BWR design boils the water in the reactor core itself. 

Big-picture-wise, a BWR is a simpler design but is more likely to have radioactive contamination in the steam side of the plant. The reactor at Three Mile Island was a PWR, while the Fukushima reactors were BWR designs. Chernobyl was neither of these designs – those design decisions had enormous consequences. 

In both cases, the reactor heats water as it passes through the core. More precisely, heat is exchanged in the reactor core, cooling the reactor and warming the coolant. In the case of a BWR, the water is heated to boiling in the reactor vessel itself. Water in a PWR follows a more circuitous path. The water carrying the heat it picked up in the reactor passes through a nest of thousands of tubes to transfer that heat to water on the other side of those tubes, causing it to boil – the steam generator. 

In both cases, the steam spins turbines that extract every last bit of energy before the steam passes into a condenser. There, the steam passes through another nest of cooling tubes returning the steam to the liquid state. The now-liquid water is pumped back to be heated and start the cycle again.

Chernobyl – Problematic moderators, control rods, and containment 

Both Western-style nuclear reactors and Chernobyl used water for cooling. But Chernobyl’s reactor was designed to produce both heat and an isotope of plutonium, Pu-239, used in creating nuclear weapons. Chernobyl used graphite for moderating neutrons because graphite is a better moderator than is water. Moderation slows down the neutrons, allowing for more fission and producing more plutonium. 

The problem with this design is that a loss of coolant won’t shut down the reactor because the graphite moderator is still present. In Chernobyl, when they lost their cooling water, the graphite kept moderating neutrons, causing fission. This kept pumping energy into the core for some time, raising the temperature and leading to the tremendous steam explosion that blew the reactor apart. 

Control rods

Another problem was Chernobyl’s control rods – the bottom several inches not only didn’t absorb neutrons but were also made of graphite. When the operators went to insert the control rods, power actually increased at first. Had the rods continued driving into the core, the neutron-absorbing part would have entered, and power would have started to drop. However, the rods stuck with the moderator portion inserted and the absorber part still mainly outside the core. This helped to drive power even higher at just the time the operators needed to shut it down.

Western-style reactors don’t use graphite in their control rods, so as soon as the rods start driving into the core, power begins to drop. Thus, even if the control rods stick, reactor power will still be lower than when the reactor SCRAM (the emergency shutdown of the reactor) was initiated.


Chernobyl’s last problem was a lack of containment. I visited one of these reactors in Lithuania in 2003 or so and was amazed that we could walk directly into the Reactor Hall without going through any containment structure at all – it was more like walking into a warehouse (albeit with a 20-foot diameter circle on the floor) than into a containment structure. I asked what the big circle on the floor was and was told it was the “bioshield” – a 7-meter-thick plug that sat on top of the reactor core itself, protecting those working in the hall from the radiation emitted by the fissions taking place below. 

Unfortunately, while Western-style containment structures are designed to withstand a lot of punishment from within and without, the structure enclosing the Chernobyl reactor provided no more protection – in either direction – than the walls of a home. The steam explosion triggered by the rapid power increase sent debris, including the bioshield, ripping through the walls and roof, opening huge rents through which radioactive smoke poured.

Both the Three Mile Island and Fukushima reactors were Western-design nuclear power plants, so they did not suffer from the design flaws of the Chernobyl plant. So it’s natural to wonder what happened to cause them to meltdown – if for no other reason than to try to figure out what sort of risk might be posed by these reactors.

Three Mile Island

Events began at the Three Mile Island when a pressure relief valve lifted and then stuck. Rather than briefly reducing pressure slightly, there was, instead, the equivalent of a small hole in the reactor plant. As the valve continued to bleed pressure (and water) from the reactor, a series of alarms began to sound. Unfortunately, the operators didn’t interpret the readings correctly, largely because this scenario had not been anticipated or trained for. Over time water level in the reactor core dropped to the point at which the fuel began to melt. The zirconium used to clad the fuel started to react with the water to form hydrogen, which escaped through the stuck valve and collected in the containment structure.

Hydrogen, of course, is an explosion hazard. The NRC and the operators and engineers realized that it had to be vented to prevent an explosion from damaging the containment. But many fission fragments are either gases or are volatile, so venting hydrogen meant also venting radioactivity. However, the time of the venting could be chosen, preferably when the wind was blowing away from population centers. As a result, nobody outside the plant site received a dose of radiation high enough to cause problems – nobody received a radiation dose high enough to make them. 


The Fukushima reactors were jolted with the strongest earthquake to hit Japan and in the top five globally since accurate records have been maintained. This subjected the reactors to far more stress than they were designed to withstand, and the reactors shut themselves down as they were designed to do. At some point, water began to leak from the reactor plant, threatening to uncover the fuel. With no power available from the disrupted local grid, the plant relied on emergency diesel generators to provide electricity to keep the cooling pumps running and inject water into the plant to keep the core covered. 

Then the tsunami hit, swamping the diesel generator and causing a total loss of power to the site. Water continued leaking from the reactor plant, uncovering the core; the fuel began to heat and eventually melted before emergency responders could start adding seawater to the reactors. 

Like Three Mile Island, water reacted with the fuel cladding to produce hydrogen; unlike TMI, this was not vented, and the containment structures experienced multiple hydrogen explosions. This made it possible for the volatile and gaseous fission products to enter the atmosphere, forming a contamination plume. While most of the contamination drifted out to sea, some drifted northwest, contaminating the land in this direction. Despite this, nobody off-site received enough radiation to cause harm. 

During my trip to Japan a month after the accident, I measured radiation dose rates in several locations where the plume from Fukushima touched down. The dose rates were clearly elevated, ranging from about 10 to 100 times as high as normal background radiation dose rates in those areas. This sounds worrisome.

I also identified the nuclides causing these dose rates – primarily Cs-134, Cs-137, and various isotopes of iodine. Because dose rate drops as nuclides decay, I calculated that the dose to someone who lived every minute of an 80-year lifespan in the “hottest” part of the plume would receive a lifetime radiation dose that was highly unlikely to cause any health problems. The World Health Organization came to the same conclusions – that radiation from Fukushima was unlikely to hurt anybody 

Summing up

Since 1942, there have been only these three major accidents at commercial reactors in what is now nearly 15,000 reactor-years of commercial nuclear power operations (one reactor-year accumulates when one reactor operates for one year). 

Two of these resulted in the destruction of the reactor core, release of a fraction of the radioactivity to the environment, and no anticipated short- or long-term health risks to workers or the public. Chernobyl, the only reactor not of Western design, was far worse. It was a combination of design features at Chernobyl that made the accident worse and more deadly than need have been the case.

It’s tempting to look at the raw statistics and to try to draw conclusions. While Russian-designed reactors comprise only 16% of the world’s nuclear power reactors, they have suffered 33% of the major accidents. But that 33% represents a single accident; the numbers are too small to draw any meaningful conclusions other than 

  • nuclear reactors don’t suffer major accidents very often 
  • Western-style reactors don’t (yet) seem to be prone to the sort of accidents that injure or kill people. 

That is why I didn’t mind living within a few hundred feet of the reactor on a submarine or a few hundred yards from a pier filled with nuclear submarines for four years, or just a few miles from a commercial nuclear reactor in Upstate New York.