Nuclear reactor physics
From Wikipedia, the free encyclopedia
|
Most nuclear reactors use a chain reaction to induce a controlled rate of nuclear fission in fissile material, releasing both energy and free neutrons. A reactor consists of an assembly of nuclear fuel (a reactor core), usually surrounded by a neutron moderator such as water, graphite, or zirconium hydride, and fitted with mechanisms such as control rods that control the rate of the reaction.
The physics of nuclear fission has several quirks that affect the design and behavior of nuclear reactors; this article presents a general overview of the physics of nuclear reactors and their behavior.
[edit] Criticality
In a nuclear reactor, most fission events are caused by neutrons impacting nuclear fuel. Hence, the power output (and neutron production) of a nuclear reactor at present depends on the number of neutrons that are already in the core from previous fissions, and on the expected value of how many fissions will occur as a result of each neutron before the neutron is absorbed or lost.
If the rate of production of new neutrons from fission in an assembly of nuclear fuel (a "core") is less than the rate of loss from absorption or escape, then the core is subcritical and will not support a self-sustaining chain reaction. If the rate of production exceeds the rate of loss, then the core is supercritical and the amount of neutrons produced will grow exponentially. The rate of growth depends on the ratio of neutron production to loss, and on the average lifetime of a neutron in the reactor core.
If we write 'N' for the number of free neutrons in a reactor core and 'τ' for the average lifetime of each neutron (before it either escapes from the core or is absorbed by a nucleus), then the reactor will follow differential equation (the evolution equation)
- dN / dt = αN / τ
where α is a constant of proportionality, and dN / dt is the rate of change of the neutron count in the core. This type of differential equation describes exponential growth or exponential decay, depending on the sign of the constant α, which is just the expected number of neutrons after one average neutron lifetime has elapsed:
- α = PimpactPfissionnavg − Pabsorb − Pescape
Here, Pimpact is the probability that a particular neutron will strike a fuel nucleus, Pfission is the probability that the neutron, having struck the fuel, will cause that nucleus to undergo fission, Pabsorb is the probability that it will be absorbed by something other than fuel, and Pescape is the probability that it will "escape" by leaving the core altogether. navg is the number of neutrons produced, on average, by a fission event -- it is between 2 and 3 for both 235U and 239Pu.
If α is positive, then the core is supercritical and the rate of neutron production will grow exponentially until some other effect stops the growth. If α is negative, then the core is "subcritical" and the number of free neutrons in the core will shrink exponentially until it reaches an equilibrium at zero (or the background level from spontaneous fission). If α is exactly zero, then the reactor is critical and its output does not vary in time (dN / dt = 0, from above).
Nuclear reactors are engineered to reduce Pescape and Pabsorb. Small, compact structures reduce the probability of direct escape by minimizing the surface area of the core, and some materials (such as graphite) can reflect some neutrons back into the core, further reducing Pescape. Light metals such as aluminum that are not strong neutron absorbers are used to build the structure of reactor cores.
The probability of fission, Pfission, depends on the nuclear physics of the fuel, and is often expressed as a cross section. Reactors are usually controlled by adjusting Pabsorb. control rods made of a strongly neutron-absorbent material such as cadmium or boron can be inserted into the core: any neutron that happens to impact the control rod is lost from the chain reaction, reducing α. Pabsorb is also controlled by the recent history of the reactor core itself (see below).
[edit] Starter sources
The mere fact that an assembly is supercritical does not guarantee that it contains any free neutrons at all. At least one neutron is required to "strike" a chain reaction, and if the spontaneous fission rate is sufficiently low it may take a long time (in 235U reactors, as long as many minutes) before a chance neutron encounter starts a chain reaction even if the reactor is supercritical. Most nuclear reactors include a "starter" neutron source that ensures there are always a few free neutrons in the reactor core, so that a chain reaction will begin immediately when the core is made critical. A common type of neutron source is a mixture of an alpha particle emitter such as 241Am (Americium-241) with a lightweight isotope such as 9Be (Beryllium-9). Once the chain reaction is begun, the starter source is removed from the core to prevent damage from the high neutron flux in the operating reactor core.
[edit] Subcritical multiplication
Even in a subcritical assembly such as a shut-down reactor core, any stray neutron that happens to be present in the core (for example from spontaneous fission of the fuel, from radioactive decay of fission products, or from a neutron source) will trigger an exponentially decaying chain reaction. Although the chain reaction is not self-sustaining, it acts as a multiplier that increases the equilibrium number of neutrons in the core. This subcritical multiplication effect can be used in two ways: as a probe of how close a core is to criticality, and as a way to generate fission power without the risks associated with a critical mass.
As a measurement technique, subcritical multiplication was used during the Manhattan Project in early experiments to determine the minimum critical masses of 235U and of 239Pu. It is still used today to calibrate the controls for nuclear reactors during startup, as many effects (discussed in the following sections) can change the required control settings to achieve criticality in a reactor. As a power-generating technique, subcritical multiplication allows generation of nuclear power for fission where a critical assembly is undesirable for safety or other reasons. A subcritical assembly together with a neutron source can serve as a steady source of heat to generate power from fission.
Including the effect of an external neutron source ("external" to the fission process, not physically external to the core), one can write a modified evolution equation:
- dN / dt = αN / τ + Rext
where Rext is the rate at which the external source injects neutrons into the core. In equilibrium, the core is not changing and dN/dt is zero, so the equilibrium number of neutrons is given by:
- N = τRext / ( − α)
If the core is subcritical, then α is negative so there is an equilibrium with a positive number of neutrons. If the core is close to criticality, then α is very small and thus the final number of neutrons can be made arbitrarily large.
[edit] Neutron moderators
To improve Pfission and enable a chain reaction, uranium-fueled reactors must include a neutron moderator that interacts with newly produced fast neutrons from fission events to reduce their kinetic energy from several MeV to several eV, making them more likely to induce fission. This is because 235U is much more likely to undergo fission when struck by one of these thermal neutrons than by a freshly-produced neutron from fission.
Neutron moderators are materials that interact weakly with the neutrons but absorb kinetic energy from them. Most moderators rely on either weakly bound hydrogen or a loose crystal structure of another light element such as carbon to transfer kinetic energy from the fast-moving neutrons.
Hydrogen moderators include water (H2O), heavy water(D2O), and zirconium hydride (ZnH2), all of which work because a hydrogen nucleus has nearly the same mass as a free neutron: neutron-H2O or neutron-ZnH2 impacts excite rotational modes of the molecules (spinning them around). Deuterium nuclei (in heavy water) absorb kinetic energy less well than do light hydrogen nuclei, but they are much less likely to absorb the impacting neutron. Water or heavy water have the advantage of being transparent liquids, so that, in addition to shielding and moderating a reactor core, they permit direct viewing of the core in operation and can also serve as a working fluid for heat transfer.
Crystal structure moderators rely on a floppy crystal matrix to absorb phonons from neutron-crystal impacts. Graphite is the most common example of such a moderator. It was used in Chicago Pile-1, the world's first man-made critical assembly, and was commonplace in early reactor designs including the Soviet RBMK nuclear power plants, of which the Chernobyl plant was one.
[edit] Moderators and reactor design
The amount and nature of neutron moderation affects reactor controllability and hence safety. Because moderators both slow and absorb neutrons, there is an optimum amount of moderator to include in a given geometry of reactor core. Less moderation reduces the effectiveness by reducing the Pfission term in the evolution equation, and more moderation reduces the effectiveness by increasing the Pescape term.
Most moderators become less effective with increasing temperature, so under-moderated reactors are stable against changes in temperature in the reactor core: if the core overheats, then the quality of the moderator is reduced and the reaction tends to slow down (there is a "negative temperature coefficient" in the reactivity of the core). Water is an extreme case: in extreme heat, it can boil, producing effective voids in the reactor core without destroying the physical structure of the core; this tends to shut down the reaction and reduce the possibility of a fuel meltdown. Over-moderated reactors are unstable against changes in temperature (there is a "positive temperature coefficient" in the reactivity of the core), and so are less inherently safe than under-moderated cores.
Most reactors in use today use a combination of moderator materials. For example, TRIGA type research reactors use ZrH2 moderator mixed with the 235U fuel, an H2O-filled core, and C (graphite) moderator and reflector blocks around the periphery of the core.
[edit] Delayed neutrons and controllability
Fission reactions and subsequent neutron escape happen very quickly; this is important for nuclear weapons, where the object is to make a nuclear core release as much energy as possible before it physically explodes. Most neutrons emitted by fission events are prompt: they are emitted essentially instantaneously. Once emitted, the average neutron lifetime (τ) in a typical core is on the order of a millisecond, so if the exponential factor α is as small as 0.01, then in one second the reactor power will vary by a factor of (1+0.01)1000, or more than ten thousand. Nuclear weapons are engineered to maximize the power growth rate, with lifetimes well under a millisecond and exponential factors close to 2; but such rapid variation would render it practically impossible to control the reaction rates in a nuclear reactor.
Fortunately, the effective neutron lifetime is much longer than the average lifetime of a single neutron in the core. About 0.65% of the neutrons produced by 235U fission, and about 0.75% of the neutrons produced by 239Pu fission, are not produced immediately, but rather are emitted by radioactive decay of fission products, with an average lifetime of about 15 seconds. These delayed neutrons increase the effective average lifetime of neutrons in the core, to nearly 0.1 seconds, so that a core with α of 0.01 would increase in one second by only a factor of (1+0.01)10, or about 1.1 -- a 10% increase. This is a controllable rate of change.
Most nuclear reactors are hence operated in a prompt subcritical, delayed critical condition: the prompt neutrons alone are not sufficient to sustain a chain reaction, but the delayed neutrons make up the small difference required to keep the reaction going. This has effects on how reactors are controlled: when a small amount of control rod is slid into or out of the reactor core, the power level changes at first very rapidly due to prompt subcritical multiplication and then more gradually, following the exponential growth or decay curve of the delayed critical reaction. Further, increases in reactor power can be performed at any desired rate simply by pulling out a sufficient length of control rod -- but decreases are limited in speed, because even if the reactor is taken deeply subcritical, the delayed neutrons are produced by ordinary radioactive decay of fission products and that decay cannot be hastened.
[edit] Reactor poisons
Any element that strongly absorbs neutrons is called a reactor poison, because it tends to shut down (poison) an ongoing fission chain reaction. Some reactor poisons are deliberately inserted into fission reactor cores to control the reaction; boron or cadmium control rods are the best example. Many reactor poisons are produced by the fission process itself, and buildup of neutron-absorbing fission products affects both the fuel economics and the controllability of nuclear reactors.
[edit] Long-lived poisons and fuel reprocessing
In practice, buildup of reactor poisons in nuclear fuel is what determines the lifetime of nuclear fuel in a reactor: long before all possible fissions have taken place, buildup of long-lived neutron absorbing fission products damps out the chain reaction. This is the reason that nuclear reprocessing is a useful activity: spent nuclear fuel contains about 99% of the original fissionable material present in newly manufactured nuclear fuel. Chemical separation of the fission products restores the nuclear fuel so that it can be used again.
Nuclear reprocessing is useful economically because chemical separation is much simpler to accomplish than the difficult isotope separation required to prepare nuclear fuel from natural uranium ore, so that in principle chemical separation yields more generated energy for less effort than mining, purifying, and isotopically separating new uranium ore. In practice, both the difficulty of handling the highly radioactive fission products and other political concerns make fuel reprocessing a contentious subject. One such concern is the fact that spent uranium nuclear fuel contains significant quantities of 239Pu, a prime ingredient in nuclear weapons (see breeder reactor).
[edit] Short-lived poisons and controllability
Short-lived reactor poisons in fission products strongly affect how nuclear reactors can operate. Unstable fission product nuclei transmute into many different elements (secondary fission products) as they undergo a decay chain to a stable isotope. The most important such element is Xenon, because the isotope 135Xe, a secondary fission product with a half-life of about 9 hours, is an extremely strong neutron absorber. In an operating reactor, each nucleus of 135Xe is destroyed by neutron capture almost as soon as it is created, so that there is no buildup in the core. However, when a reactor shuts down, the level of 135Xe builds up in the core for about 9 hours before beginning to decay. The result is that, about 6-8 hours after a reactor is shut down, it can become physically impossible to restart the chain reaction until the 135Xe has had a chance to decay over the next several hours; this is one reason why nuclear power reactors are best operated at an even power level around the clock.
135Xe buildup in a reactor core makes it extremely dangerous to operate the reactor a few hours after it has been shut down. Because the 135Xe absorbs neutrons strongly, starting a reactor in a high-Xe condition requires pulling the control rods out of the core much farther than normal. But if the reactor does achieve criticality, then the neutron flux in the core will become quite high and the 135Xe will be destroyed rapidly -- this has the same effect as very rapidly removing a great length of control rod from the core, and can cause the reaction to grow too rapidly or even become prompt critical.
135Xe played a large part in the Chernobyl accident: about eight hours after a scheduled maintenance shutdown, workers tried to bring the reactor to a zero power critical condition to test a control circuit, but because the core was loaded with 135Xe from the previous day's power generation, the reaction rapidly grew uncontrollably, leading to steam explosion in the core, fire, and violent destruction of the facility.
[edit] Uranium enrichment
While many fissionable isotopes exist in nature, the only usefully fissile isotope found in any quantity is 235U (see, e.g., the League of Women Voters' "Nuclear Waste Primer"); about 0.7% of the uranium in most ores is the 235 isotope, and about 99.3% is the inert 238 isotope. For most uses as a nuclear fuel, uranium must be enriched - purified so that it contains a higher percentage of 235. Because 238U absorbs fast neutrons, uranium nuclear weapons require their uranium fuel to be ~90% 235U to work. Nuclear reactors with water moderation can operate with only moderate enrichment of ~5% 235U. Nuclear reactors with heavy water moderation can operate with natural uranium, eliminating altogether the need for enrichment and preventing the fuel from being useful for nuclear weapons; the CANDU power reactors used in Canadian power plants are an example of this type.
Uranium enrichment is extremely difficult, because the chemical properties of 235U and 238U are identical, so physical processes such as gas diffusion or mass spectrometry must be used to separate the isotopes based on their slightly different mass. Because enrichment is the main technical hurdle to production of nuclear fuel and simple nuclear weapons, enrichment technology is politically sensitive.
[edit] Oklo: a natural nuclear reactor
Modern deposits of uranium contain only up to ~0.7% 235U (and ~99.3% 238U), which is not enough to sustain a chain reaction moderated by ordinary water. But 235U has a much shorter half-life (700 million years) than 238U (4.5 billion years), so in the distant past the percentage of 235U was much higher. About two billion years ago, a water-saturated uranium deposit (in what is now the Oklo mine in Gabon, West Africa) underwent a naturally occurring chain reaction that was moderated by groundwater and, presumably, controlled by the negative void coefficient as the water boiled from the heat of the reaction. Uranium from the Oklo mine is about 50% depleted compared to other locations: it is only about 0.3% to 0.7% 235U; and the ore contains traces of stable daughters of long-decayed fission products.
