If you read some of the previous articles in this series (chemical/nuclear explosions and nuclear weapons), it will be easy to appreciate the enormous amount of energy available in a nuclear reaction. A very small amount of fuel delivers a vast amount of useable energy. A nuclear bomb releases all of this energy in one gigantic explosion. Nuclear power plants, on the other hand, harness the same energy by carefully controlling the reaction rate and capturing the heat emitted by the fission. But is it safe? Can a nuclear plant blow up like a nuclear bomb? What's a meltdown and when can it happen? What do we do about the inevitable radioactive waste? For these reasons, nuclear power is highly controversial around the world. In order to decide for ourselves what our position is on nuclear energy we need to learn how these reactors work.
There are significant similarities between a nuclear reactor and a nuclear weapon and there are critical differences as well. A nuclear weapon relies on a run-away nuclear reaction, whereas the nuclear reaction rate in a reactor is highly controlled.
Can A Reactor Blow Up?
The nuclear fission reaction that occurs in nuclear weapon is the same cascading reaction that takes place inside a nuclear reactor used for energy production. However, the designs of the two devices are vastly different. A nuclear weapon is carefully designed to maximize an uncontrolled chain reaction, releasing as much energy as possible as quickly as possible, before the weapon itself explodes and stops the chain reaction. A nuclear power reactor's reaction is controlled so that it releases a steady supply of energy over time. While a reactor can overheat and undergo meltdown, a very dangerous situation, it is impossible for a power reactor to undergo a nuclear explosion (though it can undergo a regular explosion as a result of steam pressure or hydrogen gas build-up). The fission chain reaction itself, whether it's in a nuclear weapon or a reactor, is the same reaction. Free neutrons emitted by one fission initiate other fissions in the material, and so on.
The reaction must be kept critical or it will spontaneously slow down and stop. What does critical mean? I offer a long explanation in the article Nuclear Weapons: Understanding Binging Energy about three quarters down the article starting at "as you can see . . . " but for this article all you need to know is that critical means that the nuclear reaction is self-sustaining where there is no increase or decrease in power, temperature or neutron population. Supercritical means the reaction is increasing in power, temperature and neutrons; subcritical means that it is decreasing.
This kind of reaction naturally fluctuates - it grows or shrinks exponentially and the trick for a reactor is to hold that reaction at a fairly constant rate. To do this, a reactor must be able to slow the reaction process down to the point that it can be controlled, whereas in a bomb you want the reaction to be run-away. In a reactor, the criticality itself must have a slowed down time-scale and the secret to that is to make use of delayed neutrons versus prompt neutrons, something we will explore in detail. A reactor is always kept just at criticality - in reality it is always fluctuating between delayed-supercritical and subcritical, but the important point (and the key reason why a reactor will never blow up in a nuclear explosion) is that it is always below prompt-critical. A nuclear weapon must be at or above prompt-critical in order to detonate. This means that for each fission event, one or more immediate or prompt neutrons is emitted, causing an additional event, which causes a very rapid exponential increase in fission, and therefore in power, heat and neutron number.
This being said, a reactor can experience an event that is akin to a pre-detonation in a nuclear bomb, where a low-powered uncontrolled chain reaction explosion occurs in one very small section and this would cause a lot of damage and a meltdown. This is not a full-scale nuclear explosion and it has never happened in practice. Even the Chernobyl disaster, which involved a run-away chain reaction, a meltdown and a low-powered steam explosion and fire (the last two of which threw a tremendous amount of radioactive debris into the air), did not include a nuclear explosion of any kind. A steam or hydrogen explosion with the release of radioactive fission products into the air is most often the greatest safety concern for any nuclear reactor.
Power Reactors Versus Research Reactors
There are two basic types of nuclear reactor based on type of use - power reactors, which most of the focus is on here, and research reactors. Research reactors differ from power reactors in that they use uranium that is more highly enriched - usually around 20% enrichment, but some reactors use very highly enriched weapon-grade 93% uranium. These reactors are designed differently from power reactors because they have a much higher core power density. Whereas power reactors are sources of useful heat, research reactors are most often used as fast neutron sources.
A brief mini-lesson on neutrons: All these neutron terms - delayed, prompt, fast and thermal (slow) - can be very confusing. After all a neutron is just a neutron. The terms delayed and prompt require a more involved explanation, given above, and we will talk more about them later. The terms fast and thermal, however, simply refer to the neutron's kinetic energy. A lot of kinetic energy means it is traveling fast. A fast neutron has an energy of around 1 MeV (million electron volts) and it is traveling at about 10% the speed of light (20,000 km/s), whereas a thermal neutron (we will be talking about both of these in detail) has an energy of about 0.025 eV, equivalent to about 2.2 km/s. That's about 10,000 times less energy than a fast neutron.
The free neutrons that research reactors produce are used for neutron scattering experiments and for the testing of new materials and they are often also used for the production of radioisotopes for medical and industrial use. Research reactors use either fast or thermal neutrons. Power reactors are used for electricity production, heat generation and for submarine propulsion and they usually but not always involve thermal neutrons.
An example of a research reactor is the NRU thermal reactor at Chalk River. It's one of several thermal neutron research reactors in the world. CNL's (Canadian Nuclear Laboratories) Chalk River Laboratories in Ottawa, shown below, contains Canada's only major neutron beam source and material testing reactor, and it is one of the two largest producers of medical isotopes in the world, which are used for diagnostic applications as well as cancer treatments. Research here also focuses on increasing knowledge of the effects of radiation on humans to ensure worker safety at nuclear facilities around the world. A few decades ago my dad worked here.
Universities in various provinces in Canada have research reactors, about half of which are SLOWPOKE reactors. The University of Edmonton (where I got my degrees and worked) has a SLOWPOKE 2 fast neutron research reactor that was developed in the late 1960's by Atomic Energy of Canada Limited (AECL). It uses highly enriched uranium (93%; sourced from the U.S.) as its fuel. It produces fast neutrons for radioisotope production, neutron activation (elemental) analysis, research and teaching. This compact low energy reactor was specifically designed for Canadian universities, using a beryllium reflected core with very low critical mass. It produces a very high neutron flux. The core is only 22 cm x 22 cm and it sits in a pool of light water 2.5 m in diameter and 6 m deep. The pool is large enough that the core can be cooled by natural convection. It has a high degree of inherent safety because it can regulate itself passively. The chain reaction slows down when the water heats up or when it starts to bubble (boil).
The U of A website is disappointingly (and perhaps reassuringly?) spare in detail. I was unable to find out if there are any plans underway to update this reactor into a more secure low-enriched (20%) uranium version, a process that is just getting underway for Canadian SLOWPOKE reactors.
A larger SLOWPOKE 3 reactor was designed in the 1980's. This reactor could supply one hundred times more power, enough to be used for a district heating (networked hot water) system for a remote community that currently relies on fossil fuels such as oil or natural gas for heating. The expected market for this system so far hasn't materialized and the single reactor that was built at Whiteshell Laboratories in Manitoba was shut down, likely because the prices of oil and especially natural gas have been relatively low.
Thermal Reactors Versus Fast Reactors
Another way of comparing reactors is by the way they carry out the fission reaction. Most power reactors are thermal reactors, which means they use slowed or thermal neutrons to maintain the fission chain reaction in their fuel. These reactors use a neutron modulator, a material that slows free neutrons emitted during fission until they are thermalized, which means the neutrons have the same average kinetic energy (temperature) as the surrounding particles in the core. The neutrons are slowed in order to increase their cross section, or probability of fissioning the fissile fuel (usually uranium-235), and to reduce their chance of being captured by fissionable but not fissile uranium-238, a process that would take the neutrons out of play. A fissile material is capable of sustaining a nuclear chain reaction (with neutrons of any energy). A fissionable material is capable of undergoing fission only by capturing a high-energy fast neutron, and once captured it cannot sustain a fission chain reaction.
Mined uranium naturally contains far more U-238 than U-235. Enriching uranium increases the percentage of the U-235 isotope present. U-235 is the fissible isotope, and when it fissions, it emits fast neutrons. These neutrons are more likely to pass right through the fuel than to interact with it and produce new fissions. In order to keep the fission reaction going, some kind of moderating material is required in order to slow these fast neutrons down to the point where they are likely to interact with other nuclei. Several materials can be used, most commonly regular (light, 1H20) water, but solid graphite, heavy water (2H20), and less commonly beryllium and other materials can be used. All of these substances have low mass (they contain fairly small atoms), high scattering cross section and low absorption cross section. This means that neutrons are scattered rather than absorbed, as they collide elastically with nuclei in the moderator. The neutrons bounce off nuclei but now have less energy, so this process in effect distributes their kinetic energy around until they reach thermal equilibrium with the moderator. Some reactors are more thoroughly thermalized than others. For example, in the NRU research reactor at Canadian Nuclear Laboratories in Chalk River, nearly all the fission reactions are produced by thermal neutrons, while in a pressurized water power reactor (we will discuss) a portion of the fissions are produced by fast neutrons. In a theoretical supercritical water power reactor (will also discuss), more than half the fissions may be produced by fast neutrons and this would technically make these reactors fast neutron reactors rather than thermal reactors.
Fast reactors in use today use no moderator at all. They rely on fission produced by fast neutrons to sustain the chain reaction. As we might suspect, ordinary natural unenriched uranium won't cut it in most fast reactors. The cross section is too small and the fission chain reaction cannot be sustained. A fast reactor can make use of non-fissile U-238 but it requires a more substantial proportion of fissile U-235 in its fuel as well. In these reactors, up to 20% of the fissions can come from the fission of U-238, which is not fissile at all with thermal neutrons. Fast neutrons, moving at around 10% the speed of light, have a very low chance of producing fission in neighbouring U-235 nuclei (low cross section), so richer fissile material (a higher density of U-235) is needed in order to maintain the chain reaction. Some of these reactors use highly enriched weapon-grade uranium, which is both very expensive (a very expensive to build enrichment plant is required) and presents a security issue. However, a fast reactor can be designed that actually produces more fissile material than it consumes. It breeds fuel in other words and is therefore called a fast breeder reactor. For example, it can breed fissile plutonium by fissioning non-fissile U-238.
Several breeder reactor prototypes have been built around the world since the 1960's but only three are currently in use. The Superphénix reactor in France, brought online in 1984 and closed in 1998, could produce 20% more fuel than it consumed, and optimum breeding allowed about 75% of the energy of natural uranium to be used compared to just 1% used in a standard light water reactor. The fast neutrons used in most breeder reactors have enough energy to fission heavy medium and long half-life fission products (these are transuranic actinides on the periodic table) - these isotopes contribute significant radiotoxicity to ordinary spent nuclear fuel for over 500,000 years. In contrast, the fission products in breeder reactor spent fuel tends to remain radiotoxic only for a few hundred years, a huge advantage.
All reactors breed some new fissile material. The ratio of new fissile material to fissile material that is "burned" is called the conversion ratio. A conversion ratio of 1 means that the reactor is breaking even - it produces as much fissile material as it consumes. Light water reactors have a conversion ratio of 0.6. Pressurized heavy water reactors (such as CANDU reactors) have ratio of 0.8. In a breeder reactor the ratio is over 1. Eventually a breeder reactor will produce enough new fuel to supply a starting fuel load for another reactor. These reactors have a very high neutron economy and in principle they can be used to create new fissile fuel or run long-term without refueling or they can be used to burn nuclear waste.
Radioactive Waste: A Growing Problem
When we think about breeder reactors, it brings to the front the challenges of dealing with radioactive waste. What is it? Nuclear fuel usually consists of metal rods that enclose stacked up ceramic pellets, which are made of compacted uranium oxide powder that is sintered at high temperatures, though there are exceptions. The metal varies with the design of the reactor, as does the kind of fuel used, but the metal is often a zirconium alloy. This alloy has a very low cross section for thermal neutrons, a good thing. It is also very hard and it is very corrosion-resistant. Two fuel bundles from a CANDU power reactor are shown below right. The CANDU, a heavy-water pressurized power reactor, invented in Canada, is so far our only type of power reactor.
In a natural uranium reactor such as the CANDU, the fissile component in the fuel starts off at 0.71% U-235 and ends up with almost the same amount of fissile material, 0.50%, which is now composed of 0.23 % U-235 and 0.27% Pu-239. What makes the waste "waste" is not the lack of fissile material but the build-up of neutron absorbing fission products, which make sustaining the fission reaction impossible. These percentages are small and it doesn't seem like a lot of radioactivity on paper, but if you stood unshielded a few metres from spent nuclear fuel that was just removed from the reactor you would receive a lethal dose in a few seconds. There would also be a tremendous amount of heat given off. Because most isotopes in the nuclear waste have a short half-life, decay heat from the waste decreases fairly rapidly and exponentially. The graph below shows how fast decay heat decreases over 10 days after total shutdown, using two different models.
The word "meltdown" is not a technical word but in practice it refers to the core melting and partially or completely collapsing. The temperature has risen enough that least one nuclear fuel element exceeds its melting point. Often the core cladding is breached as well, letting radioactive materials breach containment and escape into the environment. Lava-like core material, called corium, can react with oxygen or steam chemically, releasing even more heat into the system, and it can react with boric acid as well if that is used as an emergency coolant (more about this later). It may also release volatile elements into the air. The zirconium alloy metal rods can oxidize under extreme heat, releasing explosive hydrogen gas. The corium temperature can reach as high as 2400°C, and at this temperature when it comes into sudden contact with water, steam is explosively released, sending shards of core everywhere and possibly further damaging the core containment. Thermal decomposition of the concrete containment in contact with corium produces water vapour and carbon dioxide, which can further react with metals in the corium and create even more explosive gas. Core meltdown is clearly a very dangerous, and dangerously unpredictable, situation.
Even after full reactor shutdown, there is still a lot of heat production because the decay of the products will still be going on, contributing about 7% of full reactor power. This feature, common to all reactors, can have dire consequences. In the Fukushima Daiichi nuclear disaster, for example, residual decay heat from the core, after complete shutdown (the fission chain reaction was stopped), rose after a loss of coolant flow. Fuel rods exposed to air reacted chemically with it producing hydrogen gas, which is highly explosive when mixed with air. The inevitable resulting hydrogen explosions blew radioactive material everywhere, contaminating hundreds of square kilometres of land and part of the ocean.
CANDU Reactor: Waste Problem and Possible Solution
Most nuclear power reactors use low-enriched uranium as fuel but the heavy water design of the CANDU reactor means that it can use fuel with a lower percentage of fissile uranium than light water reactors. It can even use recovered uranium from spent light water reactor fuel. It can also burn a mixture of uranium and plutonium oxides (MOX fuel) as well as the plutonium from dismantled nuclear weapons. These are all big advantages of this kind of reactor. Still, the CANDU is not a breeder reactor. It creates a significant amount of radioactive waste that has to be stored somewhere. Currently Canada stores all of its spent nuclear fuel at the reactor sites in pools of water (for 10 years) and then in dry cask storage. As of 2011, Canada had 2.2 million spent fuel bundles in storage, each containing 20 kg of spent fuel, translating into about 44,000 tonnes of heavy metal waste. There are plans for an eventual deep geologic repository below the water table but nothing has been built as of yet.
Now for the bad news: Although decay heat drops exponentially as soon as the bundles are removed from the reactor and during this time much of the radioactive material has decayed into safe stable elements, heavy transuranium actinides, with half-lives of up to 25,000 years remain and become the dominant source of decay heat after about 100 years dry storage. Spent CANDU fuel remains radiotoxic for about 1 million years. Perhaps a better way of dealing with this waste, rather than storage, is to reuse it in future fast neutron reactors. A fast reactor can potentially convert the actinides into other fission products that have much shorter half-lives. The use of fast neutron reactors would not only reduce the effective half-life of the radioactive waste but it could make use of it to generate an enormous amount of additional electricity. A very interesting 2012 paper written by Dr. Peter Ottensmeyer for the Ontario Centre for Engineering and Public Policy claims a reduction in radiotoxicity to about 300 years (fast neutrons can fission the transuranic actinides) is possible, that is until the waste reaches the background level of natural uranium. This paper also claims that Canada's current 44,000 tonnes of total spent fuel could ultimately be converted into $48 trillion dollars worth of (non-carbon) electricity in fast neutron reactors because there is still significant fissile energy present in the waste that the fast neutron reactor can use. Understandably on reading this, one can tell that he is impatient with policymakers to get on this, rather than continue to store the waste above ground or spend resources to transport it and put it deep underground.
How A Nuclear Power Reactor Works
All nuclear power reactors utilize the same basic plan. Heat from nuclear fission is transferred to a fluid, which flows through a turbine, making it turn. The turbine then either drives a submarine's propellers or it spins an electrical generator, as shown in this brief 1-minute video produced by the Tennessee Valley Authority (requires the instalment of a VLC player).
Reactor Type Based on Coolant Used
In addition to categorizing nuclear reactors by their function - power versus research, or whether they use fast or thermal neutrons, reactors are categorized based on the kind of coolant they use. The following are some of the most common reactor designs in use. There are many designs and some are not included here, such as the gas-cooled reactor, of which Great Britain currently has two advanced designs and the still very experimental molten salt reactor.
Pressurized Water Reactor
A controlled uranium fission chain reaction produces lots of heat. In a pressurized water reactor, water is kept under pressure in the primary coolant loop (the orange loop in the diagram below). This water absorbs fission heat by thermal conduction through the core cladding. Now extremely hot but not able to boil as it's pressurized, the water is then pumped through a heat exchanger, which is also a steam generator (blue container with the orange tube in it). The water in this secondary system is completely separate from the pressurized core coolant water and in the diagram it is vastly simplified - the steam created is actually pumped through thousands of tubes. This secondary coolant water (blue) evaporates into pressurized steam, which drives the turbine (grey rotor assembly) and that spins the generator. This arrangement, using two completely separated fluids, ensures that the secondary coolant never becomes radioactive.
The animation below shows how energy is transferred through a typical pressurized water reactor. Primary coolant is shown in orange and secondary coolant is shown in blue
The steam turbine drives an electrical generator connected to the power grid, which distributes electrical energy. The secondary coolant is then cooled and condensed back into liquid water and then pumped back into the steam generator.
Most Western nuclear power plants, like the reactor just described, are pressurized water reactors that use ordinary (light) water as the primary coolant. The CANDU power reactor, mentioned earlier, is an exception because it uses pressurized heavy water. We will explore how it works in a moment but first I'd like to compare two additional kinds of light water reactors. One uses boiling water and the other, which is still in conceptual stage, uses supercritical water.
Boiling Water Reactor
The boiling water reactor, shown below, is a bit simpler in design than the pressurized water reactor. In this case, water is heated by thermal conduction from the reactor core (red vertical fuel rods) where it boils into steam (blue to violet transition) and then the steam is used directly to drive a steam turbine (green box). The steam is condensed back into water (the grey pipe contains cold water) and returned to the core. There is no secondary coolant and no pressurizer.
|Robert Steffens (alias RobbyBer 8 November 2004), SVG: Marlus_Gancher, Antonsusi (talk) using a file from Marlus_Gancher;Wikipedia|
The Fukushima 1 nuclear power plants, which went into meltdown and exploded, were boiling water reactors. They experienced a so-called double-failure event in which the reaction could no longer be properly limited or moderated (power was lost to the water pumps) combined with a complete emergency core cooling system failure. Water vapour pressure continued to increase, heated by the nuclear fuel, and the Mark 1 containment that was used in these reactors failed, which allowed the release of highly pressurized radioactive steam. Modified containment is now designed to release steam in a controlled manner should such a double scenario happen again, with the addition of activated carbon filters to trap radioisotopes as the steam leaves.
A disadvantage of the boiling water reactor is that the newest designs have control rods that are inserted from the bottom of the reactor (the diagram above shows this new design). In most other reactor types, the rods are held above the reactor by electromagnets so if power is lost the rods fall into the reactor and the reaction is stopped. In the boiling water reactor two hydraulic power sources can drive the rods into the reactor in an emergency, but the system relies on at least one of them to work.
Supercritical Water Reactor
A third type of light water reactor is the supercritical water reactor, which is currently still in the concept stage of development (it's categorized as a Generation IV reactor).
What is supercritical water? Supercritical water, not to be confused with a supercritical mass of nuclear fuel, is water that is at its critical point. Above a specific temperature and pressure, the physical properties of water change very dramatically and these, now supercritical properties, can be utilized. To understand what critical point is, lets first consider the physical phases of water. A general phase diagram is shown below right (ignore the plasma phase for our purposes).
A supercritical water reactor operates at higher pressure and temperature than pressurized water reactors and it has a direct once-through cycle like boiling water reactors. Its advantage is that it has a much higher thermal efficiency while keeping to a simple design, which looks a lot like the boiling water reactor plan except the pressure is very high and the water phase is different, shown below.
fast neutron reactor has advantages such as higher core power density and more efficient use of nonfissile uranium-238 (which is by far the more abundantly available isotope of uranium). The fast neutrons also split (radioactive) actinides and transmute long-lived radioactive fission products into isotopes with shorter half-lives. By doing this, this reactor will not only reduce the radiotoxicity of the nuclear waste it produces but it cuts short the waste's radioactive lifetime as well.
And it's highly efficient. A supercritical water reactor can in theory use almost all of the fuel present. Conventional fast reactors have the disadvantage that they are expensive to build and they require much more costly highly enriched fuel (some need weapons-grade fuel, which adds a security concern as well). The supercritical water reactor will likely need higher fuel enrichment as well but nowhere near weapon-grade. A disadvantage is that supercritical water is more corrosive than ordinary water, so all the core materials including the cladding will have to be of a higher standard (materials that resist corrosion but also do not absorb many neutrons). Hydrogen can also be added to the water to reduce its corrosiveness. Higher temperature and pressure in the core in general will mean increased mechanical and thermal stress on vessel materials as well.
On the other hand, supercritical water has far better heat transfer ability so less of it is needed, and this means a smaller core and a smaller containment structure is needed. However, in an accident, there is less water available to act as an emergency coolant. Temperature in the core could shoot up very high very fast, requiring cladding that will not readily melt and well-designed and redundant emergency cooling systems.
These are a couple of the challenges involved in developing this kind of reactor. Despite them, the reactor's minimal radioactive waste, low cost and simple design place this reactor high on the list of potential future design candidates.
The supercritical water reactor is just one of many theoretical reactor designs currently being researched, all of them categorized as Generation IV reactors. A simple comparison of reactor generations is shown below. CANDU, developed in the 1970's, is Generation II, for example.
Heavy Water Pressurized Reactor: CANDU
Edmonton is currently home to Alberta's only nuclear reactor - it's the research reactor SLOWPOKE 2, described earlier. In addition to research reactors, Canada has 20 nuclear power reactors - in Ontario and Quebec, and one in New Brunswick, accounting for about 15% of Canada's total electricity production. They are all CANDU-6 reactors, Canadian-invented pressurized heavy water reactors, which evolved from prototypes invented during and just after WW II to explore nuclear energy. While Canada lacked any expensive uranium enrichment facilities, it does have uranium ore, so a design evolved from the experimental ZEEP heavy water prototype, which used unenriched uranium and was designed by Canadian, British and French scientists as part of an effort to produce plutonium for nuclear weapons (Canada was part of the Manhattan Project). It then went through several earlier CANDU-type designs before the current CANDU was reached.
A more advanced Generation III+ CANDU reactor was designed but none were built. It was to be a light water cooled reactor that uses a separate heavy water moderator which is much like current CANDU reactors (described below). The key difference is that this reactor would use low enriched uranium fuel rather than natural unenriched uranium, which means more of the fuel would be "burned," reducing the amount of nuclear waste In 2007, Alberta Energy planned to use this newer prototype to process oil sands in Northern Alberta, but it was later scrapped. In 2011, SNC-Lavalin took over AECL and killed this reactor design. Some sources online claim that the design was unoriginal and flawed. Now, Toshiba has developed a small reactor, which Alberta plans to power oil sands extraction. Expected to come online in 2020, the Toshiba 4S, is a micro liquid sodium-cooled fast neutron reactor and it would replace the use of natural gas in oil sands extraction and therefore reduce carbon emissions. Not surprisingly there is significant concern among Albertans at the thought of having nuclear reactors dotting the northern landscape. I wonder if this will have any impact on the American decision whether or not to go forward with the Keystone Pipeline.
Meanwhile, several of SNC-Lavalin's new advanced fuel CANDU reactors might be sold to China. This design can use recycled uranium from light water reactors (as well as more abundant thorium) as fuel, so China could recycle the spent fuel from its existing 22 nuclear reactors and possibly from some of the additional 26 reactors now under construction.
A basic schematic diagram of a CANDU reactor is shown below.
|Emoscopes;Wikipedia; the legend is a screen capture from the Wikipedia page|
CANDU reactors use natural unenriched uranium as fuel and heavy water (deuterium oxide) as the moderator. Designed much like a light water pressurized reactor, the use of heavy water, while much more expensive than light water, is a big advantage because it is a far better moderator (a ratio of 11,449 compared to 246 for light water). The light hydrogen in light water is very good at slowing fast neutrons into thermal neutrons. Because the proton is almost the same mass as a neutron it absorbs a lot of kinetic energy when the two collide in an elastic collision. However, light hydrogen also tends to absorb neutrons as well (into a heavy water nucleus), and this is undesirable because it results in fewer neutrons available for fission. Heavy hydrogen is also good at producing thermal neutrons and it has a much lower absorption cross section.
Technically put, the heavy water reactor has a high neutron economy - every neutron emitted is slowed down to the right kinetic energy to maximize the chance of a fission reaction and because it is not absorbed it stays in play, having a good chance of starting another fission reaction. The extra cost of the heavy water balances the savings of being able to use natural unenriched uranium. The use of heavy water allows the reactor to "burn" more uranium, using it more efficiently. However, a larger volume of fuel must go through the system, resulting in a larger volume of spent fuel. That being said, the spent uranium is less radioactive and can be stored more compactly. There is another downside to this - this reactor creates plutonium and tritium (although heavy hydrogen is fairly immune to neutron capture, some is captured to create tritium) as reaction byproducts, and these are two substances used to make boosted fission and fusion bombs. Spent CANDU fuel, once its radioactivity is reduced, can be a security threat.
The Benefit of Delayed Neutrons
The CANDU design has a built-in safety feature in that heavy water moderation stabilizes the fission chain reaction. This is how it works: The heavy hydrogen nucleus in heavy water is made of one proton and one neutron and it has quite low binding energy, which means it can be broken apart fairly easily. Some energetic neutrons and especially gamma rays (from the fission itself and from the decay of fission fragments) break the nuclei apart and add more free neutrons to the mix. The fission fragments have half-lives from seconds to hours to years, so gamma rays are emitted over time. This means that neutrons are jarred from heavy water over time in a more drawn out manner than they would be from light water.
Neutrons emitted over a period of time are called delayed neutrons. They are emitted anywhere from milliseconds to minutes after the fission event. During the fission event itself, the emission of neutrons is prompt - they are emitted almost instantaneously. The high power growth rate of nuclear weapons is hinged upon this fact, but if this happened in a reactor it would be almost impossible to control the reaction rate. An individual free neutron lasts only about one millisecond in the core before it is captured, but the emission of neutrons from the decay of fission products effectively extends that lifetime, affording a controllable overall rate of reaction. To maximize the benefit of delayed neutrons, the fuel is kept at such a level it would be subcritcal without the contribution of decay neutrons. The additional continuous contribution of delayed decay neutrons keeps the fuel just critical, allowing fission to continue but at a manageably slow rate.
The safety feature of this becomes apparent when the reaction accelerates for whatever reason in one part of the reactor. All nuclear reactors have an inherent challenge: the continuously replenished/lost cloud of neutrons inside a reactor core is subject to spatial and temporal fluctuations that are complex and difficult to predict, requiring sophisticated software to carefully control the reaction rate in various parts of the vessel. In operation, reactors are very complex systems. The neutrons interact differently with different isotopes present and the composition of isotopes is constantly changing, so neutron probes and temperature sensors are placed throughout the core. In the CANDU, a fluctuation in one zone will propagate relatively slowly to the rest of the core, and that delay should allow various feedback mechanisms built into the reaction to operate and allow technicians time to respond to the emergency. Still, this brings up the disturbing question of how good is the computer system (of any reactor) and how secure is it? In the case of the CANDU, there are two systems running simultaneously and independently, one as a backup for the other.
Is the CANDU safe?
This design has built-in features, as described above, that work to maintain a safe, steady and predictable fission reaction, but what happens if something goes very wrong and overcomes these features?
This design has a positive void coefficient. This means that if voids (steam bubbles) form in the reactor, its reactivity increases, a potentially devastating flaw in reactor design. In the CANDU, one can picture a nightmare scenario in which heavy water in the core gets too hot or there is a rupture that releases pressure. Sudden boiling would lead the reactor to become even more active, and then more steam is released and so on, creating a potential positive feedback loop leading to meltdown and a steam explosion. The Chernobyl disaster, involving a run-away chain reaction, was the result of a positive void coefficient. In the CANDU, however, there is a deliberately large mass of moderating heavy water in the reactor core so that, if this kind of event were set off, the effects should progress slowly, and that extra time along with the sluggish response of the fission process itself, should afford operators ample opportunity to deal with the problem, though this has not been proven in practice. While there is no international standard for nuclear reactors that prohibits designs with a positive void coefficient, the Chernobyl disaster, caused by a large positive void coefficient, makes some potential buyers and industry watchdogs wary of the CANDU reactor design. The latest advanced fuel CANDU, which may be sold to China, operates with more highly enriched fuel and has a negative void coefficient.
A design aspect that CADNU has going for it is the fact that the fuel bundles are held horizontally in the core. This means that if they get very hot, they will melt and sag and this change in shape will reduce their fission efficiency. There is very little excess reactivity in the core to start with so any deformation of the rods should bring the reaction to subcritical and stop the fission.
To deal with an emergency, the CANDU design has several built-in emergency systems: two independent shutdown systems, an emergency core cooling system and an emergency containment system. Shutdown system 1 uses neutron-absorbing rods that drop by gravity into the core, stopping the fission reaction. Shutdown system 2 injects high-pressure liquid neutron poison into the moderator. The poison is usually boric acid because boron nuclei have a high cross section for neutron capture. When dissolved in the moderator, boric acid provides spatially uniform neutron absorption, and in sufficient quantity it alone will stop the reaction. The emergency core cooling system can re-establish core cooling through high-pressure water injection, medium-pressure water supply from the building's dousing tank, and by recovering water from the building's sump. Again, Fukushima speaks caution. It was a situation where all primary, backup generator and battery power supplies were lost, and where valves to a similar dousing tank were also inexplicably closed and could not be reopened because of the loss of power. As a final safety measure, the CANDU reactor is contained by continuous concrete envelope.
Shortly after the Fukushima disaster, Canada's CANDU reactors underwent an extensive safety re-evaluation by the Canadian Nuclear Safety Commission. The full report can be read here. And here you can read the yearly national industry safety reports from 2008 to 2013.
Fast Breeder Reactors: Promise, Disappointment and New Promise
All large-scale breeder reactors around the world are currently fast breeder reactors, all of which use liquid sodium metal as the coolant, which can be circulated through heat exchangers outside the reactor tank (loop design, right) or inside the reactor tank (pool design, left), as shown below.
Liquid sodium cooled reactor designs have great promise but there is a long list of past failures. Over the decades, many experimental designs, including France's fairly disastrous Superphénix, have been built to use liquid sodium coolant but most of them have since been shut down. Sodium is very hazardous. It is so reactive that when it comes into contact with water it explodes and it burns upon exposure to air. Another serious concern is that liquid sodium can chemically react with many kinds of core cladding. The Monju loop-type power plant in Japan, brought online in 1994, was forced to shut down when a sodium leak caused a major fire. Restarted, it was again forced to shut down after another accident and it is now being decommissioned after costing the equivalent of over 9 billion dollars. Despite past failures, the promise of liquid sodium remains. As mentioned earlier, the Toshiba 4S micro reactor, being considered for processing oil sands bitumen in Northern Alberta, is an advanced liquid sodium fast reactor. An advanced (Generation IV) fast neutron breeder reactor design, called SFR, also calls for liquid sodium, as both the primary and secondary coolant. A diagram of this design is shown below.
Liquid lead also has been, and continues to be, considered as a reactor coolant. It has several advantages such as high neutron reflection and low neutron absorption cross sections and it is an excellent shield against gamma radiation as well. It has a high enough boiling point that it can effectively cool the core even at several hundred degrees C above normal operation conditions. The price to pay for this is that it is understandably difficult to refuel and service a molten lead cooled reactor core. Some Soviet nuclear submarine reactors built in the 1970's were lead-cooled, but other than that none has been used. However, a new Generation IV lead-cooled fast reactor is being developed, which overcomes the maintenance problem because the entire core can be replaced after many years of operation. In addition and unlike other reactor designs, no electricity is required to cool down the reactor after shutdown because of its natural circulation behaviour, a big safety advantage. If liquid lead-bismuth is used as the coolant, it would also quickly solidify in the case of a leak, eliminating the risk of an explosion. Liquid sodium reactors run the risk of a positive void coefficient but lead's nuclear properties eliminate that problem. Finally, lead is not very reactive and it is relatively cheap (bismuth if used, however, is expensive). The diagram below shows what this lead reactor might look like.
The atomic nucleus, when harnessed as a nuclear weapon, has the power to cause unfathomable death and destruction, and that fear understandably spreads to a distrust of nuclear power as well. Several highly publicized nuclear reactor accidents have also driven fear into the world's population. Far less publicized are a stupendous number of military nuclear accidents and near accidents. An example that sticks out in my mind was the manual assembly a critical mass of plutonium during a 1946 demonstration, which killed the Canadian physicist, Louis Slotin, within days. These accidents and various nuclear reactor disasters tell us that our understanding of nuclear science has a long way to go. Yet despite all of that and many financially costly failures in reactor design, the green promise of nuclear energy continues to emerge, ironically, as one of the new technologies that might save mankind and our planet.
Well-designed with safety kept as a high priority, nuclear energy might finally experience its golden age, leaving carbon-emitting non-renewable oil, coal and natural gas technologies to history. I recommend a website called whatisnuclear.com. Created by a nuclear physicist and two nuclear engineers, it offers a series of interesting and easy to read articles written especially for people who want to learn more about nuclear energy and make an informed choice. It would also serve as an excellent primer for science teachers, as I hope this article does.
Next, in the last article in this series explore the potential and challenges of nuclear fusion power.