Introduction

One of the big debates of our age is the safety of nuclear power: how do nuclear reactors compare to nuclear bombs, what do we do with the nuclear waste, and how dangerous is radiation? The textbook gives a broad overview of these issues and in this handout we will dig a little deeper into nuclear physics and the working principles of the world's nuclear reactors.

The 2 basic necessary items for a nuclear reactor are:

1. a core where the nuclear reaction takes place and heat is generated

2. a coolant, which transfers the heat from the core to a system that turns the turbine, which drives an electric generator.

In some reactors, the core coolant is used to drive the turbine directly (e.g., Boiling-water reactors), whereas in others a heat exchanger is used, and clean hot water drives the turbine (e.g., Pressurized hot water reactors). Some military and most commercial British reactors use a gas as coolant. Most powerplants are built near a river or lake to have water available as a coolant for the condensor section (remember Carnot), and all powerplants (including nukes) create thermal pollution.

The Nuclear Reactions

Most commercial reactors are based on the principle of fission: splitting a large nuclide into smaller fragments. Careful addition of the masses of the fission products (FP) shows that during the fission reaction some mass is lost and converted to thermal and radiation energy (E=mc2). The basic fission reaction is Fuel + n ==> FP + xn + heat + radiation, where the term xn is the # of neutrons (n) released. The radiation consists largely of gamma rays (g), but many of the FP are radioactive (ra) and may also emit a or b particles. The ra decay of the FP also generates heat, so the fuel assembly has many different heat sources. About 450 different nuclides are generated as FP and their respective daughters. The fuel can be U or Pu, but not all isotopes of these elements are fissile. The nuclides 239Pu and 235U are fissile by n collision. The likelihood that fission will occur upon collision with a neutron is expressed by the nuclear cross section, given in the units "barns". The larger the cross section, the more likely it is for that nuclide to catch an n and fission.

The newly generated n can fission neighbouring nuclides and a chain reaction mechanism is set up. The reaction will evolve exponentially if the # of neutrons that is generated is > 1 and all these n's are used for fission reactions. In natural environments, 235U occurs on earth today at an abundance of 0.7 % of total Uranium; the rest is mainly 238U. 239Pu only occurs in trace quantities in the earth and is largely person-made from 238U through n capture. The n's released during 235U fission are high-energy n's (fast neutrons) and both 238U and 235U have similar cross sections for fast n's: the first will absorb the n and make 239U, whereas 235U will fission. If natural U were to be used in a simple reactor, the nuclear reaction would almost stop immediately because 238U gobbles up all available n and this absorption creates no new n. The newly made 239U isotope decays through a rapid step of b decay (T1/2 = ~ 9 hours) to 239Np and then to 239Pu, which has a T1/2 of 24,400 years. 235U has a much larger cross sections for slow neutrons (also known as thermal neutrons) than 238U, and so the reactions in most power generating reactors are based on thermal neutrons: the fast n's are slowed down with a moderator. In a mechanistic sense, we can see moderators as light elements that take over part of the momentum from the flying n's during collisions (like billiard balls). The moderator should not be n-hungry and so we do not want an effective n absorber.

Nuclear Reactor Types

The commonly used moderator nuclides are 1H, 2D and 12C, the first two usually in the form of water (H2O and D2O, resp. light and heavy water), and C as graphite. H is slightly hungry for n (to make 2D) but D is an ideal moderator, but expensive to make (from ocean water). All commercial US reactors use light water as a moderator, Canadian and some military US reactors use heavy water as a moderator, and British and some US and Russian military reactors use graphite. Because light water is a slight neutron poison, some n will disappear into the moderator. A self-sustained chain reaction has h = { #of n produced / # of n consumed } ~ 1. The n economy is thus of great importance to run a reactor at some steady pace, and the different n sinks have to be known quite well to tune the speed of the chain reaction. Sinks for n are the light water moderator, other capture reactions with FP, and to much lesser degree capture by 238U. On average, 2.5 n are freed during a fission of 235U, and some of these will strike another 235U. In US reactors with light water, the fuel has to be enriched in 235U to about 3 % to make the chain reaction self-sustaining. The reactor has control rods, which are efficient n poisons (Cd or B) to regulate the reaction speed, which will vary when the fuel is new and once it gets older (more 235U used up).

When the n's are not slowed down by a moderator, abundant new 239Pu will form from n capture by 238U, and such a system is called a fast breeder reactor (used in France). Not only rabbits are fast breeders, but the fast in the nuclear sense refers to the energy (speed) of the n, not the speed of breeding. These reactors are liquid-metal cooled, with a conventional core of 235U burning, and the heat is carried off by liquid sodium, which is neither a moderator nor n-eater (but reacts explosively with air or water!). The fast neutrons travel through the sodium and are captured by a blanket of 238U where new fuel in the form of 239Pu is grown. The different types of fuel and blanket from the French reactors are reprocessed, that is, the 235U and 239Pu are extracted and used as new fuel, and the remaining 238U is prepared in a new blanket for breeding (Mama nuke puts the blankies over her little breeders). In the USA, no reprocessing of commercial fuel is carried out, but old 238U was sold on the heavy metal market to make tank-penetration rocket tips for the Gulf war! The Canadian reactors (CANDU-type) use heavy water as moderator, and so these reactors can run on un-enriched (natural) Uranium. The UK reactors have a massive graphite core as moderator and are cooled by a mixture of Helium and CO2, which doesnot eat n's. These reactors run at a high temperature and are thermodynamically efficient.

The US commercial reactors are thermally stable: in case of loss-of-coolant accidents, the reactor looses its coolant and also its moderator, and the reaction shuts off automatically. The ra decay of FP will still generate heat, which may lead to partial melting of the core. The core could melt through the floor and hit the water table, and the resulting explosion could disperse the ra debris. This extreme scenario is probably not very likely. In the Three Mile Island accident, which was caused by a string of severe operator errors and mechanical malfunctions, the core had started to melt, but the emergency cooling system kept it from further melting, and the dome retained all the stuff inside. UK reactors are thermally neutral: loss of coolant will not influence the nuclear reaction rate, but emergency cooling systems are essential.

Nuclear weapons versus power generators

The difference between a nuclear weapon and nuclear power reactor should now be clear: reactors have fuel that is not 235U-enriched enough to cause a bomb-like explosion and without moderator, they will not run at all. Nuclear bombs consist of >90% 235U or 239Pu and do not need a moderator, and the whole chain reaction happens in a few milliseconds. All thermal energy is generated in a short time and in a small spot: a strong explosion and intense "heat concentration" results, which is very destructive. A critical mass is a certain mass of 235U that will have a sustained chain reaction. A nuclear bomb is exploded by moving several subcritical masses quickly together to create a supercritical mass, and BOOM.

The CANDU reactor and many Russian reactors can be used for the breeding of bomb-grade 239Pu while being operated as commercial power plants. Commercial US reactors use the fuel rods for 1-2 years, and refuelling is a total operational shut down procedure for more than 4-5 months. The initial 239Pu is formed from 238U during the first weeks of operation, but after a month the 239Pu concentration becomes high enough that also 240Pu formation will occur through n capture. So "older" fuel rods have more 240Pu than 'younger' ones. 240Pu tends to autofission and nuclear bombs with too much 240Pu may not explode fully because they go off before supercriticality is reached (they fizzle). Pure, bomb-grade 239Pu can thus only be grown from 238U by short-term exposure to n's. In US reactors, all fuel rods are standing in a single pool of water, and one cannot move a single rod out after a few weeks while the plant is under load. In CANDU reactors and Chernobyl - type reactors, the fuel rods can be replaced one by one while the reactor is operating. Each fuel rod is surrounded by its own coolant pipe and each rod can be individually removed. This creates less down time and continuous operation in general, but also allows to remove fuel rods after 2-3 weeks when there is still very little 240Pu. So terrorists are unlikely to steal spent fuel-rods from commercial US reactors for bomb-grade Pu, because it stinks. CANDU reactors, which have been sold worldwide, can be used for the production of bomb-grade Pu.

The Chernobyl reactor and its 1986 accident is a complex story by itself. The reactor is a burner with a graphite moderator and individually cooled fuel rods with light water as a coolant. This is a thermally unstable design: with loss of coolant, the nuclear reaction will speed up because the moderator remains in place whereas a neutron poison (light water) is removed. At routine conditions, the core runs already very hot (700 C) and the graphite core has to be shielded from air by an inert gas otherwise it would self-ignite. During the sad 1986 accident, some operators (not nuclear engineers!) carried out a test on running the reactor when the coolant flow was interrupted. The reactor core became so hot that the graphite caught fire, and the volatile elements from the fuel rods were carried with the hot plume into the air. There was no protective dome and the roof collapsed during the explosion. The control rods became stuck in their tracks when the graphite core ignited and the reaction became uncontrollable. The core temperature remained below 2800oC, because no a emitters (U) were found in the plume. The heroic efforts of Russian firefighters in the plant, who received terrible doses of body radiation, prevented further major meltdown. The main contaminants were 131I and 137Cs, which were spread all over Europe. The nuclear plant was subsequently covered with boron (a n poison) and lead (absorbs radiation) and layers of concrete (entombment). Why did they have such a rotten design for a reactor: the fuel rods could be removed at any time under load, so 239Pu (without that bad 240Pu) could be extracted from them for bomb manufacturing, without a shutdown. The necessary working space above the core, needed to manipulate fuel rods with remote handling cranes, prevented construction of a dome.

Radio-active exposure and waste disposal

The main routinely released ra effluents of nuclear reactors are a ra isotope of Krypton, a noble gas, and Tritium (3H). Isotopes from the fuel rods are not supposed to reach the area outside the dome, although spent fuel rods are usually stored for cooling in a swimming pool-like facility outside the reactor . The problem in the USA is that we do not have a permanent storage facility for spent fuel rods, and we do not reprocess them. We are stuck with them and ra waste disposal is a major problem. Reasonably safe solutions seem to be present and possible sites for disposal are:

deep sea beds - red clays in the deep sea

salt domes- water-poor quiet environments

granites- extensive massive plutons e.g., in New Hampshire.

ash flow tuffs Yucca Mountain, NV - voluminous volcanic rock sheets

A small Dutch political party once proposed that Holland should be rented out to store all the world's nuclear waste. All Dutch people would leave and live on the 'French Riviera', living from the rental income. Good business sense, but storage in the wettest country in the world is not a very safe solution. A repository should have strict safeguarding that groundwater would stay out of it and in that sense, the salt domes and ash flow tuffs in Nevada may be our best bets. The ra isotopes of nuclear waste would be "fixed" in borosilicate glass or in ceramic compounds similar to natural minerals that have carried ra elements for billions of years (SYNROC). The glass would be loaded into a stainless steel cannister, mantled by concrete and wheeled by rail car into the catacombs of Yucca Mountain. The air should be circulated to carry off the heat from ra decay. An expensive undertaking, but possible, and probably better than local storage in swimming pools on reactor sites. The main risks of permanent storage may lie in the transport to the repository, whereas terrorist activities and natural disasters pose a small but real risk at the site.

Radiation dosage and human tolerances are an intensive field of study, but safe dosage is difficult to determine. The exposure of people in Nagasaki and Hiroshima gave the high-level end of radiation damage from radiation exposure, but low-level exposure damage and its relation to cancer is difficult to determine. The units for radiation are discussed in the book, and the most common unit is the REM. Natural background radiation is 100-250 millirem/ year, dependent on U and Th concentration in local rocks and/or Radon the accumulation in houses. Jet flights at great altitude cause additional exposure from the solar wind. Individual body dosage (e.g. for nuclear plant workers) is set at a maximum of 500 millirem/year; a dosage of 100 REM causes radiation sickness, and a 500 REM exposure is deadly for 50% of the population.