The basic principle at the core of most nuclear reactors is simple: pack together enough radioactive material of the right type, and you get a chain reaction in which an atom (let’s say uranium) “splits” into two smaller atoms (i.e. undergoes fission), releasing some heat and also some neutrons (particles at the center of atoms); the neutrons can strike nearby uranium atoms and cause them to split as well, leading to a chain reaction that continues to release heat along with the neutrons that sustain it [figure 1, below].
This splitting happens naturally at a low rate in uranium, so if you pack the material tightly enough with the right conditions, the process can start on its own. In fact it has happened spontaneously in nature on rare occasions, for example 1.7 billion years ago in Oklo, Gabon, the right convergence of natural uranium and water led to an underground “reactor” that lasted for over 1000 years and produced about 100 kilowatts (kW) of heat on average, roughly equal to the output of 20 standard residential rooftop solar arrays in midday sun. Alhough 100 kW is small, the energy that can be released from such a process per unit of fuel is enormous – 1 metric ton of typical enriched uranium fuel can release over 1 billion kWh of thermal energy over its useful life in a reactor, as much as would be derived from 160,000 metric tons of coal.
Building a device that releases this huge store of energy is quite straightforward. Making such a device both safe and economical is the technical challenge engineers and scientists have labored over for the past 60 years. Additionally, engineers must contend with the problem of nuclear waste disposal and how to prevent undesired parties from using the same technology needed for a benign energy system to instead make a weapon. Each of these topics is complex and deserving of multiple textbooks, but here we briefly overview the technical aspects of plant design, fuel cycles, and waste as a primer for reading some of the articles in this review.
Basic Plant Design
At a high level, all a nuclear power plant is doing is carrying out the chain reaction described above in a controlled way, and then using the resultant heat to produce electricity. Typically, electricity is generated by using the heat to produce steam that turns a generator, in much the same way as in a coal plant or concentrating solar power array.
Figure 2 [above] shows a typical modern “Pressurized Water Reactor” (PWR), with three “loops” of water. The first loop passes through the reactor and picks up heat from the chain reaction, but is so pressurized that it does not actually boil. The water pipes carrying this hot water then pass through a steam generator, where water from a separate loop vaporizes to steam. Note that the water coming directly from the reactor core, containing radioactive elements, ideally never comes in physical contact with the water being turned to steam, it just passes its heat along and heads back to the reactor core. The hot steam then turns a turbine to generate electricity, and later comes into contact with pipes from a third loop carrying cold water. The cold water cools down the steam and condenses it back into liquid water, so it can then flow back to the steam generator and be vaporized again. The cooling loop, several steps removed from the actual nuclear reactions, either passes through an iconic cooling tower (like the one displayed on the cover of this publication) or an external water source like the ocean or a river, releasing the heat into the air or water, but not releasing any physical material from the nuclear reaction.
Of course, the details are more complex, especially what is happening inside the reactor itself. All uranium is not equally useful for sustaining a chain reaction – the most abundant isotope, U238, is fairly difficult to use, while the much less common U235 is more desirable. Natural uranium found today contains around 99.3% U238 and just 0.7% U235, which under most conditions is not enough to carry out a chain reaction as neutrons released by the fissioning (splitting) of one U235 atom are not likely to collide with another U235 atom in time. To run most modern nuclear reactors, the uranium either needs to be “enriched,” by increasing the fraction of U235, or needs to be immersed in a strong “moderator,” a substance that makes neutrons bump into other uranium atoms at a higher rate, thus making a chain reaction more likely. Water, the typical working fluid in reactors as described above, is not a very strong moderator, meaning that the uranium has to be slightly enriched in standard plant designs, usually to 3% U235. However, other configurations are possible – Canada did not want to enrich nuclear material, so instead built the CANDU fleet of plants using deuterium oxide (“heavy water”) which is a much stronger moderator than H2O, allowing even natural uranium to carry out a chain reaction. This eliminated the need for enrichment facilities to increase the fraction of U235 in fuel, but required facilities to produce heavy water instead.
Controlling A Chain Reaction, and Its After-Effects
One obvious question: if a chain reaction is happening in the reactor, releasing ever more heat and neutrons, how do we keep the reaction from “running away” and becoming so hot it melts the reactor? Modern reactors use three main strategies: 1) they are designed with a negative feedback loop, where the reactor becoming hotter slows down the reaction for reasons we will not describe here, 2) they are designed with a “negative void coefficient,” meaning that the reaction slows down or stops if the pressurized water coolant is lost; thus, if the reactor starts to overheat and vaporizes the water, the reaction is slowed or halted, and 3) they use “control rods,” physical rods made of some neutron-absorbing material that can be inserted amongst the fuel rods, absorbing enough neutrons to halt the process. These processes have been very reliable – there have been no major accidents at plants with the above three safety measures.
But there certainly have been accidents at nuclear power plants. They usually involve “decay heat,” which is heat that is released even after the chain reaction has ceased. This heat comes from the continued breakdown of unstable atoms produced in the reaction, and can be of considerable magnitude. A full day after a reaction has been halted, a typical reactor will still be producing 10 Mega Watts (MW) of heat. This is enough to heat all of the water in the “first loop” by over 750 C per day, and would quickly start melting through the reactor vessel and/or start causing explosions if the rest of the loops were not running to draw the heat away. This was the problem at Fukushima – the reaction was halted, but without electricity, the cooling loops could not keep running and the reactor eventually overheated. Managing decay heat is thus one of the central problems addressed in new reactor designs, which brings us to the next section, a brief review of new designs being considered.
Improving Plant Design
So far we have reviewed the predominant type of reactor in the world today, the Pressurized Water Reactor using enriched uranium. There are other types, such as the CANDU reactors with heavy water mentioned before, and “boiling water reactors” that allow the first loop of water to boil rather than keeping it liquid with high pressure. But most of the basic principles are the same. To use nuclear industry parlance, all reactors of these types are usually categorized as Generation III, or III+ if they have slightly improved safety and/or performance.
Do we need to improve on this plant design? In some countries, namely China and South Korea, new Generation III and III+ plants are being built fairly economically (roughly cost-competitive with other options) and are deemed safe enough. In the West, however, most countries either deem them unsafe or struggle to build them economically, for a variety of reasons.
Especially given growing interest in low-carbon electricity, much attention is being given to new reactor and plant designs. These are too varied and detailed to treat in depth, but they usually involve some of the following three: 1) improved safety, 2) reduced cost, and 3) reduced waste.
“Passively safe” is a term associated with next-generation plant designs, ideally meaning a plant design where decay heat is handled passively and does not rely on active engineering systems that could fail. A simple example would be to have the reactor resting in a huge pool of coolant all the time, so large that even in the event of indefinite power outage the coolant reservoir is able to handle the decay heat. Costs can be reduced by reducing the complexity of plant design, or by operating at higher temperatures to allow better thermal efficiency in electricity generation. Wastes can be reduced in several ways, such as by modifying the nuclear chain reaction to produce less stable radioactive byproducts, resulting in less total waste with shorter lifetimes.
Some proposed designs attempt to combine multiple improvements, for example small modular reactors (<300 MW) could be significantly safer due to their small size and easier thermal management, and could reduce costs by being easier to assemble in factories with less time for costly on-site construction. Of course, only time and experience will tell if their costs would actually be lower, or whether smaller economies of scale or other factors would make them more expensive. Most proposed designs trade off between safety, cost, or wastes, for example “fast neutron reactors” can significantly cut waste generation but are usually more costly, or supercritical water reactors that could reduce costs but may not offer much additional inherent safety. But all of these designs are very far from commercial licensing, probably on the order of a decade or longer, and significant financial investment and patience will be required to develop them further and determine with more certainty if any offer a more appealing set of traits than current Generation III reactors.
In the final section of this brief overview, we will examine the basics of the nuclear fuel cycle as it exists in most countries with PWR’s. Natural uranium is mined and sent to a fuel enrichment and fabrication facility. There it is separated into two streams – one enriched in U235, usually to around 3%, and another very depleted in U235, which is usually discarded. Unfortunately, the same equipment used to enrich the uranium to this level for nuclear power can also be used to enrich it further, closer to 90% U235, to make weapons-grade material, leading to ambiguities over whether some countries are enriching uranium for civilian or military purposes.
The enriched fuel can then be used in PWR’s, where it serves as fuel until the level of fissionable isotopes becomes very low again. Notice that the spent fuel leaving the plant now has quite a variety of radioactive products, formed through various reactions happening inside the reactor. The diversity of these wastes adds to the challenge of waste management, as some have half-lives of only several years while others have half-lives of many thousands of years.
Also notice that the spent fuel contains a significant amount of plutonium. This plutonium could also be used as fissionable material in a reactor, so many countries choose to “reprocess” their waste by extracting the plutonium and mixing it with depleted uranium to make more reactor fuel. This process tends to reduce the volume of waste and could be advantageous if uranium were in short supply or expensive, but for now uranium seems relatively abundant and inexpensive, and the reprocessing itself has proven expensive. Pure, fissionable plutonium created through reprocessing also leads to concerns about safety, weapons proliferation, and terrorism. However, despite these concerns, most countries using nuclear energy routinely reprocess their fuel, with the US being a notable exception mostly due its policies that attempted to “lead by example” in reducing weapons proliferation in the 1970’s.
As with plant designs, there are ways to improve on the current fuel cycle. One high level improvement would be to form a “closed” rather than “open” fuel cycle by utilizing different kinds of reactors that generate as much fissionable materials as they consume. Another is to use “fast reactors,” described earlier, to reduce the amount and lifetime of wastes. There are also possible geopolitical improvements, for example a global fuel cycle where a few agreed-upon countries supply fuel and accept waste from other countries. This would allow some countries to have nuclear power plants while never enriching fuel or handling their waste, and for countries like the US to have an easier waste disposal solution. Like the new reactor designs, though, these changes would take a very long time, easily beyond a decade, so if countries or the world decide they are desirable they will require patience.
 Source: Intel Education Resources. http://inteleducationresources.intel.co.uk/examcentre.aspx?id=278
 Source: US National Nuclear Regulatory Commission. http://www.nrc.gov/admin/img/art-students-reactors-1-lg.gif