Nuclear Plant

A nuclear power plant or nuclear power station is a thermal power station in which the heat source is a nuclear reactor. As is typical in all conventional thermal power stations the heat is used to generate steam which drives a steam turbine connected to an electric generator which produces electricity.

• Most nuclear electricity is generated using just two kinds of reactors which were developed in the 1950s and improved since.
• New designs are coming forward and some are in operation as the first generation reactors come to the end of their operating lives.
• Over 11% of the world's electricity is produced from nuclear energy, more than from all sources worldwide in 1960.

This paper is about the main conventional types of nuclear reactor. For more advanced types, see Advanced Reactors and Small Reactors papers, and also Generation IV reactors.

A nuclear reactor produces and controls the release of energy from splitting the atoms of certain elements. In a nuclear power reactor, the energy released is used as heat to make steam to generate electricity. (In a research reactor the main purpose is to utilise the actual neutrons produced in the core. In most naval reactors, steam drives a turbine directly for propulsion.)

The principles for using nuclear power to produce electricity are the same for most types of reactor. The energy released from continuous fission of the atoms of the fuel is harnessed as heat in either a gas or water, and is used to produce steam. The steam is used to drive the turbines which produce electricity (as in most fossil fuel plants).

The world's first nuclear reactors operated naturally in a uranium deposit about two billion years ago. These were in rich uranium orebodies and moderated by percolating rainwater. The 17 known at Oklo in west Africa, each less than 100 kW thermal, together consumed about six tonnes of that uranium. It is assumed that these were not unique worldwide.

Today, reactors derived from designs originally developed for propelling submarines and large naval ships generate about 85% of the world's nuclear electricity. The main design is the pressurised water reactor (PWR) which has water at over 300°C under pressure in its primary cooling/heat transfer circuit, and generates steam in a secondary circuit. The less numerous boiling water reactor (BWR) makes steam in the primary circuit above the reactor core, at similar temperatures and pressure. Both types use water as both coolant and moderator, to slow neutrons. Since water normally boils at 100°C, they have robust steel pressure vessels or tubes to enable the higher operating temperature. (Another type uses heavy water, with deuterium atoms, as moderator. Hence the term ‘light water’ is used to differentiate.)

Components of a nuclear reactor

There are several components common to most types of reactors:

Fuel. Uranium is the basic fuel. Usually pellets of uranium oxide (UO₂) are arranged in tubes to form fuel rods. The rods are arranged into fuel assemblies in the reactor core.*

* In a new reactor with new fuel a neutron source is needed to get the reaction going. Usually this is beryllium mixed with polonium, radium or other alpha-emitter. Alpha particles from the decay cause a release of neutrons from the beryllium as it turns to carbon-12. Restarting a reactor with some used fuel may not require this, as there may be enough neutrons to achieve criticality when control rods are removed.

Moderator. Material in the core which slows down the neutrons released from fission so that they cause more fission. It is usually water, but may be heavy water or graphite.

Control rods. These are made with neutron-absorbing material such as cadmium, hafnium or boron, and are inserted or withdrawn from the core to control the rate of reaction, or to halt it.*  In some PWR reactors, special control rods are used to enable the core to sustain a low level of power efficiently. (Secondary control systems involve other neutron absorbers, usually boron in the coolant – its concentration can be adjusted over time as the fuel burns up.)

* In fission, most of the neutrons are released promptly, but some are delayed. These are crucial in enabling a chain reacting system (or reactor) to be controllable and to be able to be held precisely critical.

Coolant. A fluid circulating through the core so as to transfer the heat from it.  In light water reactors the water moderator functions also as primary coolant. Except in BWRs, there is secondary coolant circuit where the water becomes steam. (See also later section on primary coolant characteristics)

Pressure vessel or pressure tubes. Usually a robust steel vessel containing the reactor core and moderator/coolant, but it may be a series of tubes holding the fuel and conveying the coolant through the surrounding moderator.

Steam generator. Part of the cooling system of pressurised water reactors (PWR & PHWR) where the high-pressure primary coolant bringing heat from the reactor is used to make steam for the turbine, in a secondary circuit. Essentially a heat exchanger like a motor car radiator*. Reactors have up to six 'loops', each with a steam generator. Since 1980 over 110 PWR reactors have had their steam generators replaced after 20-30 years service, 57 of these in USA.

* These are large heat exchangers for transferring heat from one fluid to another – here from high-pressure primary circuit in PWR to secondary circuit where water turns to steam. Each structure weighs up to 800 tonnes and contains from 300 to 16,000 tubes about 2 cm diameter for the primary coolant, which is radioactive due to nitrogen-16 (N-16, formed by neutron bombardment of oxygen, with half-life of 7 seconds). The secondary water must flow through the support structures for the tubes. The whole thing needs to be designed so that the tubes don't vibrate and fret, operated so that deposits do not build up to impede the flow, and maintained chemically to avoid corrosion. Tubes which fail and leak are plugged, and surplus capacity is designed to allow for this. Leaks can be detected by monitoring N-16 levels in the steam as it leaves the steam generator.

Containment. The structure around the reactor and associated steam generators which is designed to protect it from outside intrusion and to protect those outside from the effects of radiation in case of any serious malfunction inside. It is typically a metre-thick concrete and steel structure.

Newer Russian and some other reactors install core melt localisation devices or 'core catchers' under the pressure vessel to catch any melted core material in the event of a major accident.

There are several different types of reactors as indicated in the following table.

Nuclear power plants in commercial operation

Reactor type	Main Countries	Number	GWe	Fuel	Coolant	Moderator
Pressurised water reactor (PWR)	US, France, Japan, Russia, China	277	257	enriched UO2	water	water
Boiling water reactor (BWR)	US, Japan, Sweden	80	75	enriched UO2	water	water
Pressurised heavy water reactor (PHWR)	Canada, India	49	25	natural UO2	heavy water	heavy water
Gas-cooled reactor (AGR & Magnox)	UK	15	8	natural U (metal), enriched UO2	CO2	graphite
Light water graphite reactor (RBMK & EGP)	Russia	11 + 4	10.2	enriched UO2	water	graphite
Fast neutron reactor (FBR)	Russia	2	0.6	PuO2 and UO2	liquid sodium	none
	TOTAL	438	376

IAEA data, end of 2014.  GWe = capacity in thousands of megawatts (gross)

Source: Nuclear Engineering International Handbook 2011, updated to 1/1/12

For reactors under construction: see paper Plans for New Reactors Worldwide.

Fuelling a nuclear power reactor

Most reactors need to be shut down for refuelling, so that the pressure vessel can be opened up. In this case refuelling is at intervals of 1-2 years, when a quarter to a third of the fuel assemblies are replaced with fresh ones. The CANDU and RBMK types have pressure tubes (rather than a pressure vessel enclosing the reactor core) and can be refuelled under load by disconnecting individual pressure tubes.

If graphite or heavy water is used as moderator, it is possible to run a power reactor on natural instead of enriched uranium. Natural uranium has the same elemental composition as when it was mined (0.7% U-235, over 99.2% U-238), enriched uranium has had the proportion of the fissile isotope (U-235) increased by a process called enrichment, commonly to 3.5 - 5.0%. In this case the moderator can be ordinary water, and such reactors are collectively called light water reactors. Because the light water absorbs neutrons as well as slowing them, it is less efficient as a moderator than heavy water or graphite.

During operation, some of the U-238 is changed to plutonium, and Pu-239 ends up providing about one third of the energy from the fuel.

In most reactors the fuel is ceramic uranium oxide (UO₂ with a melting point of 2800°C) and most is enriched. The fuel pellets (usually about 1 cm diameter and 1.5 cm long) are typically arranged in a long zirconium alloy (zircaloy) tube to form a fuel rod, the zirconium being hard, corrosion-resistant and transparent to neutrons.* Numerous rods form a fuel assembly, which is an open lattice and can be lifted into and out of the reactor core. In the most common reactors these are about 4 metres long. A BWR fuel assembly may be about 320 kg, a PWR one 655 kg, in which case they hold 183 kg uranium and 460 kgU respectively. In both, about 100 kg of zircaloy is involved.

*Zirconium is an important mineral for nuclear power, where it finds its main use. It is therefore subject to controls on trading. It is normally contaminated with hafnium, a neutron absorber, so very pure 'nuclear grade' Zr is used to make the zircaloy, which is about 98% Zr plus about 1.5% tin, also iron, chromium and sometimes nickel to enhance its strength. 

Burnable poisons are often used in fuel or coolant to even out the performance of the reactor over time from fresh fuel being loaded to refuelling. These are neutron absorbers which decay under neutron exposure, compensating for the progressive build up of neutron absorbers in the fuel as it is burned. The best known is gadolinium, which is a vital ingredient of fuel in naval reactors where installing fresh fuel is very inconvenient, so reactors are designed to run more than a decade between refuellings. Gadolinium is incorporated in the ceramic fuel pellets. An alternative is zirconium diboride integral fuel burnable absorber (IFBA) as a thin coating on normal pellets.

Gadolinium, mostly at up to 3g oxide per kilogram of fuel, requires slightly higher fuel enrichment to compensate for it, and also after burn-up of about 17 GWd/t it retains about 4% of its absorbtive effect and does not decrease further. The ZrB₂ IFBA burns away more steadily and completely, and has no impact on fuel pellet properties. It is now used in most US reactors and a few in Asia. China has the technology for AP1000 reactors.

The power rating of a nuclear power reactor

Nuclear power plant reactor power outputs are quoted in three ways:

• Thermal MWt, which depends on the design of the actual nuclear reactor itself, and relates to the quantity and quality of the steam it produces.
• Gross electrical MWe indicates the power produced by the attached steam turbine and generator, and also takes into account the ambient temperature for the condenser circuit (cooler means more electric power, warmer means less). Rated gross power assumes certain conditions with both.
• Net electrical MWe, which is the power available to be sent out from the plant to the grid, after deducting the electrical power needed to run the reactor (cooling and feed-water pumps, etc.) and the rest of the plant.*

* Net electrical MWe and gross MWe vary slightly from summer to winter, so normally the lower summer figure, or an average figure, is used. If the summer figure is quoted plants may show a capacity factor greater than 100% in cooler times. Watts Bar PWR in Tennessee is reported to run at about 1125 MWe in summer and about 1165 MWe net in winter, due to different condenser cooling water temperatures. Some design options, such as powering the main large feed-water pumps with electric motors (as in EPR) rather than steam turbines (taking steam before it gets to the main turbine-generator), explains some gross to net differences between different reactor types. The EPR has a relatively large drop from gross to net MWe for this reason.

The relationship between these is expressed in two ways:

• Thermal efficiency %, the ratio of gross MWe to thermal MW. This relates to the difference in temperature between the steam from the reactor and the cooling water. It is often 33-37%.
• Net efficiency %, the ratio of net MWe achieved to thermal MW. This is a little lower, and allows for plant usage.

In WNA papers and figures and WNN items, generally net MWe is used for operating plants, and gross MWe for those under construction or planned/proposed.

Pressurised water reactor (PWR)

This is the most common type, with over 230 in use for power generation and several hundred more employed for naval propulsion. The design of PWRs originated as a submarine power plant. PWRs use ordinary water as both coolant and moderator. The design is distinguished by having a primary cooling circuit which flows through the core of the reactor under very high pressure, and a secondary circuit in which steam is generated to drive the turbine. In Russia these are known as VVER types – water-moderated and -cooled.

A PWR has fuel assemblies of 200-300 rods each, arranged vertically in the core, and a large reactor would have about 150-250 fuel assemblies with 80-100 tonnes of uranium.

Water in the reactor core reaches about 325°C, hence it must be kept under about 150 times atmospheric pressure to prevent it boiling. Pressure is maintained by steam in a pressuriser (see diagram). In the primary cooling circuit the water is also the moderator, and if any of it turned to steam the fission reaction would slow down. This negative feedback effect is one of the safety features of the type. The secondary shutdown system involves adding boron to the primary circuit.

The secondary circuit is under less pressure and the water here boils in the heat exchangers which are thus steam generators. The steam drives the turbine to produce electricity, and is then condensed and returned to the heat exchangers in contact with the primary circuit.

Boiling water reactor (BWR)

This design has many similarities to the PWR, except that there is only a single circuit in which the water is at lower pressure (about 75 times atmospheric pressure) so that it boils in the core at about 285°C. The reactor is designed to operate with 12-15% of the water in the top part of the core as steam, and hence with less moderating effect and thus efficiency there.  BWR units can operate in load-following mode more readily then PWRs.

The steam passes through drier plates (steam separators) above the core and then directly to the turbines, which are thus part of the reactor circuit. Since the water around the core of a reactor is always contaminated with traces of radionuclides, it means that the turbine must be shielded and radiological protection provided during maintenance. The cost of this tends to balance the savings due to the simpler design. Most of the radioactivity in the water is very short-lived*, so the turbine hall can be entered soon after the reactor is shut down.

* mostly N-16, with a 7 second half-life

A BWR fuel assembly comprises 90-100 fuel rods, and there are up to 750 assemblies in a reactor core, holding up to 140 tonnes of uranium. The secondary control system involves restricting water flow through the core so that more steam in the top part reduces moderation.

 

Pressurised heavy water reactor (PHWR)

The PHWR reactor design has been developed since the 1950s in Canada as the CANDU, and from 1980s also in India. PHWRs generally use natural uranium (0.7% U-235) oxide as fuel, hence needs a more efficient moderator, in this case heavy water (D₂O).** The PHWR produces more energy per kilogram of mined uranium than other designs, but also produces a much larger amount of used fuel per unit output.
** with the CANDU system, the moderator is enriched (i.e. water) rather than the fuel – a cost trade-off.
The moderator is in a large tank called a calandria, penetrated by several hundred horizontal pressure tubes which form channels for the fuel, cooled by a flow of heavy water under high pressure in the primary cooling circuit, reaching 290°C. As in the PWR, the primary coolant generates steam in a secondary circuit to drive the turbines. The pressure tube design means that the reactor can be refuelled progressively without shutting down, by isolating individual pressure tubes from the cooling circuit. It is also less costly to build than designs with a large pressure vessel, but the tubes have not proved as durable.

A CANDU fuel assembly consists of a bundle of 37 half metre long fuel rods (ceramic fuel pellets in zircaloy tubes) plus a support structure, with 12 bundles lying end to end in a fuel channel. Control rods penetrate the calandria vertically, and a secondary shutdown system involves adding gadolinium to the moderator. The heavy water moderator circulating through the body of the calandria vessel also yields some heat (though this circuit is not shown on the diagram above).

Newer PHWR designs such as the Advanced Candu Reactor (ACR) have light water cooling and slightly-enriched fuel.

CANDU reactors can accept a variety of fuels. They may be run on recycled uranium from reprocessing LWR used fuel, or a blend of this and depleted uranium left over from enrichment plants. About 4000 MWe of PWR might then fuel 1000 MWe of CANDU capacity, with addition of depleted uranium. Thorium may also be used in fuel.

Advanced gas-cooled reactor (AGR)

These are the second generation of British gas-cooled reactors, using graphite moderator and carbon dioxide as primary coolant. The fuel is uranium oxide pellets, enriched to 2.5-3.5%, in stainless steel tubes. The carbon dioxide circulates through the core, reaching 650°C and then past steam generator tubes outside it, but still inside the concrete and steel pressure vessel (hence 'integral' design). Control rods penetrate the moderator and a secondary shutdown system involves injecting nitrogen to the coolant.

The AGR was developed from the Magnox reactor, also graphite moderated and CO₂ cooled, and one of these is still operating in UK to late 2014. They use natural uranium fuel in metal form. Secondary coolant is water.

Light water graphite-moderated reactor (RBMK)

This is a Soviet design, developed from plutonium production reactors. It employs long (7 metre) vertical pressure tubes running through graphite moderator, and is cooled by water, which is allowed to boil in the core at 290°C, much as in a BWR. Fuel is low-enriched uranium oxide made up into fuel assemblies 3.5 metres long. With moderation largely due to the fixed graphite, excess boiling simply reduces the cooling and neutron absorbtion without inhibiting the fission reaction, and a positive feedback problem can arise, which is why they have never been built outside the Soviet Union. See appendix on RBMK Reactors for more detail.

Advanced reactors

Several generations of reactors are commonly distinguished. Generation I reactors were developed in 1950-60s and only one is still running today. They mostly used natural uranium fuel and used graphite as moderator. Generation II reactors are typified by the present US fleet and most in operation elsewhere. They typically use enriched uranium fuel and are mostly cooled and moderated by water. Generation III are the Advanced Reactors evolved from these, the first few of which are in operation in Japan and others are under construction and ready to be ordered. They are developments of the second generation with enhanced safety. There is no clear distinction Gen II to Gen III.

Generation IV designs are still on the drawing board and will not be operational before 2020 at the earliest, probably later. They will tend to have closed fuel cycles and burn the long-lived actinides now forming part of spent fuel, so that fission products are the only high-level waste. Of seven designs under development, 4 or 5 will be fast neutron reactors. Four will use fluoride or liquid metal coolants, hence operate at low pressure. Two will be gas-cooled. Most will run at much higher temperatures than today’s water-cooled reactors. See Generation IV Reactors paper.

More than a dozen (Generation III) advanced reactor designs are in various stages of development. Some are evolutionary from the PWR, BWR and CANDU designs above, some are more radical departures. The former include the Advanced Boiling Water Reactor, a few of which are now operating with others under construction. The best-known radical new design has the fuel as large 'pebbles' and uses helium as coolant, at very high temperature, possibly to drive a turbine directly.

Considering the closed fuel cycle, Generation 1-3 reactors recycle plutonium (and possibly uranium), while Generation IV are expected to have full actinide recycle.

Fast neutron reactors (FNR)

Some reactors (only one in commercial service) do not have a moderator and utilise fast neutrons, generating power from plutonium while making more of it from the U-238 isotope in or around the fuel. While they get more than 60 times as much energy from the original uranium compared with the normal reactors, they are expensive to build. Further development of them is likely in the next decade, and the main designs expected to be built in two decades are FNRs. If they are configured to produce more fissile material (plutonium) than they consume they are called Fast Breeder Reactors (FBR). See also Fast Neutron Reactors and Small Reactors papers.

Floating nuclear power plants

Apart from over 200 nuclear reactors powering various kinds of ships, Rosatom in Russia has set up a subsidiary to supply floating nuclear power plants ranging in size from 70 to 600 MWe. These will be mounted in pairs on a large barge, which will be permanently moored where it is needed to supply power and possibly some desalination to a shore settlement or industrial complex. The first has two 40 MWe reactors based on those in icebreakers and will operate at a remote site in Siberia. Electricity cost is expected to be much lower than from present alternatives.

The Russian KLT-40S is a reactor well proven in icebreakers and now proposed for wider use in desalination and, on barges, for remote area power supply. Here a 150 MWt unit produces 35 MWe (gross) as well as up to 35 MW of heat for desalination or district heating. These are designed to run 3-4 years between refuelling and it is envisaged that they will be operated in pairs to allow for outages, with on-board refuelling capability and used fuel storage. At the end of a 12-year operating cycle the whole plant is taken to a central facility for 2-year overhaul and removal of used fuel, before being returned to service. Two units will be mounted on a 21,000 tonne barge. A larger Russian factory-built and barge-mounted reactor is the VBER-150, of 350 MW thermal, 110 MWe. The larger VBER-300 PWR is a 325 MWe unit, originally envisaged in pairs as a floating nuclear power plant, displacing 49,000 tonnes. As a cogeneration plant it is rated at 200 MWe and 1900 GJ/hr. See also Nuclear Power in Russia paper.

Lifetime of nuclear reactors

Most of today's nuclear plants which were originally designed for 30 or 40-year operating lives. However, with major investments in systems, structures and components lives can be extended, and in several countries there are active programs to extend operating lives. In the USA most of the more than one hundred reactors are expected to be granted licence extensions from 40 to 60 years. This justifies significant capital expenditure in upgrading systems and components, including building in extra performance margins.

Some components simply wear out, corrode or degrade to a low level of efficiency. These need to be replaced. Steam generators are the most prominent and expensive of these, and many have been replaced after about 30 years where the reactor otherwise has the prospect of running for 60 years. This is essentially an economic decision. Lesser components are more straightforward to replace as they age. In Candu reactors, pressure tube replacement has been undertaken on some plants after about 30 years operation.

A second issue is that of obsolescence. For instance, older reactors have analogue instrument and control systems. Thirdly, the properties of materials may degrade with age, particularly with heat and neutron irradiation. In respect to all these aspects, investment is needed to maintain reliability and safety. Also, periodic safety reviews are undertaken on older plants in line with international safety conventions and principles to ensure that safety margins are maintained.

Another important issue is knowledge management (KM) over the full lifecycle from design, through construction and operation to decommissioning for reactors and other facilities. This may span a century and involve several countries, and involve a succession of companies. The plant lifespan will cover several generations of engineers. Data needs to be transferable across several generations of software and IT hardware, as well as being shared with other operators of similar plants.* Significant modifications may be made to the design over the life of the plant, so original documentation is not sufficient, and loss of design base knowledge can have huge implications (eg Pickering A and Bruce A in Ontario). Knowledge management is often a shared responsibility and is essential for effective decision-making and the achievement of plant safety and economics.

* ISO15926 covers portability and interoperability for lifecycle open data standard. Also EPRI in 2013 published Advanced Nuclear Technology: New Nuclear Power Plant Information Handover Guide.  

See also section on Ageing, in Safety of Nuclear Power Reactors paper.

Load-following capability

Nuclear power plants are essentially base-load generators, ideally running continuously at high capacity. This is because their power output cannot efficiently be ramped up and down on a daily and weekly basis, and in this respect they are similar to most coal-fired plants. (It is also uneconomic to run them at less than full capacity, since they are expensive to build but cheap to run.) However, in some situations it is necessary to vary the output according to daily and weekly load cycles on a regular basis, for instance in France, where there is a very high reliance on nuclear power.

BWRs can be made to follow loads reasonably easily without burning the core unevenly, by changing the coolant flow rate. Load following is not as readily achieved in a PWR, but especially in France since 1981, so-called 'grey' control rods are used. The ability of a PWR to run at less than full power for much of the time depends on whether it is in the early part of its 18 to 24-month refuelling cycle or late in it, and whether it is designed with special control rods which diminish power levels throughout the core without shutting it down. Thus, though the ability on any individual PWR reactor to run on a sustained basis at low power decreases markedly as it progresses through the refueling cycle, there is considerable scope for running a fleet of reactors in load-following mode. European Utility Requirements (EUR) since 2001 specify that new reactor designs must be capable of load-following between 50 and 100% of capacity with a rate of change of electric output of 3-5% per minute. The economic consequences are mainly due to diminished load factor of a capital-intensive plant. Further information in the Nuclear Power in France paper and the 2011 Nuclear Energy Agency report, Technical and Economic Aspects of Load Following with Nuclear Power Plants.

As fast neutron reactors become established in future years, their ability to load-follow will be a benefit.

Primary coolants

The advent of some of the designs mentioned above provides opportunity to review the various primary heat transfer fluids used in nuclear reactors. There is a wide variety – gas, water, light metal, heavy metal and salt:

Water or heavy water must be maintained at very high pressure (1000-2200 psi, 7-15 MPa, 150 atmospheres) to enable it to function well above 100°C, up to 345°C, as in present reactors. This has a major influence on reactor engineering. However, supercritical water around 25 MPa can give 45% thermal efficiency – as at some fossil-fuel power plants today with outlet temperatures of 600°C, and at ultra supercritical levels (30+ MPa) 50% may be attained.

Water cooling of steam condensers is fairly standard in all power plants, because it works very well, it is relatively inexpensive, and there is a huge experience base. Water is a lot more effective than air for removing heat, though its thermal conductivity is less than liquid alternatives.

Helium must be used at similar pressure (1000-2000 psi, 7-14 MPa) to maintain sufficient density for efficient operation. Again, there are engineering implications, but it can be used in the Brayton cycle to drive a turbine directly.

Carbon dioxide was used in early British reactors, and their current AGRs which operate at much higher temperatures than light water reactors. It is denser than helium and thus likely to give better thermal conversion efficiency. It also leaks less readily than helium. There is now interest in supercritical CO₂ for the Brayton cycle.

Sodium, as normally used in fast neutron reactors at around 550ºC, melts at 98°C and boils at 883°C at atmospheric pressure, so despite the need to keep it dry the engineering required to contain it is relatively modest. It has high thermal conductivity. However, normally water/steam is used in the secondary circuit to drive a turbine (Rankine cycle) at lower thermal efficiency than the Brayton cycle. In some designs sodium is in a secondary circuit to steam generators. Sodium does not corrode the metals used in the fuel cladding or primary circuit, nor the fuel itself if there is cladding damage, but it is very reactive generally. In particular it reacts exothermically with water or steam to liberate hydrogen. It burns in air, but much less vigorously. Sodium has a low neutron capture cross section, but it is enough for some Na-23 to become Na-24, which is a beta-emitter and very gamma-active with 15-hour half-life, so some shielding is required. If a reactor needs to be shut down frequently, NaK eutectic which is liquid at room temperature (about 13°C) may be used as coolant, but the potassium is pyrophoric, which increases the hazard.

Lead or lead-bismuth eutectic in fast neutron reactors are capable of higher temperature operation at atmospheric pressure. They are transparent to neutrons, aiding efficiency due to greater spacing between fuel pins which then allows coolant flow by convection for decay heat removal, and since they do not react with water the heat exchanger interface is safer. They do not burn when exposed to air. However, they are corrosive of fuel cladding and steels, which originally limited temperatures to 550°C. With today's materials 650°C can be reached, and in future 800°C is envisaged with the second stage of Gen IV development, using oxide dispersion-strengthened steels. They have much higher thermal conductivity than water, but lower than sodium. Westinghouse is developing a lead-cooled fast reactor concept. While lead has limited activation from neutrons, a problem with Pb-Bi is that it yields toxic polonium (Po-210) activation product, an alpha-emitter with a half-life of 138 days. Pb-Bi melts at a relatively low 125°C (hence eutectic) and boils at 1670°C, Pb melts at 327°C and boils at 1737°C but is very much more abundant and cheaper to produce than bismuth, hence is envisaged for large-scale use in the future, though freezing must be prevented. The development of nuclear power based on Pb-Bi cooled fast neutron reactors is likely to be limited to a total of 50-100 GWe, basically for small reactors in remote places. In 1998 Russia declassified a lot of research information derived from its experience with submarine reactors, and US interest in using Pb generally or Pb-Bi for small reactors has increased subsequently. The Gen4 Module (Hyperion) reactor will use lead-bismuth eutectic which is 45% Pb, 55% Bi. A secondary circuit generating steam is likely.

SALT:  Fluoride salts boil at around 1400°C at atmospheric pressure, so allow several options for use of the heat, including using helium in a secondary Brayton cycle circuit with thermal efficiencies of 48% at 750°C to 59% at 1000°C, or manufacture of hydrogen. Fluoride salts have a very high boiling temperature, very low vapour pressure even at red heat, very high volumetric heat capacity (carry more heat than the same volume of water), good heat transfer properties, low neutron absorbtion, good neutron moderation capability, are not damaged by radiation, are chemically very stable so absorb all fission products well and do not react violently with air or water, are compatible with graphite, and some are also inert to some common structural metals. Some gamma-active F-20 is formed by neutron capture, but has very short half-life (11 seconds).

Lithium-beryllium fluoride Li₂BeF₄ (FLiBe) salt is a eutectic version of LiF (2LiF + BeF2) which solidifies at 459°C and boils at 1430°C. It is favoured in MSR and AHTR/FHR primary cooling and when uncontaminated has a low corrosion effect. LiF without the toxic beryllium solidifies at about 500°C and boils at about 1200°C. FLiNaK (LiF-NaF-KF) is also eutectic and solidifies at 454°C and boils at 1570°C. It has a higher neutron cross-section than FLiBe or LiF but can be used intermediate cooling loops.

Chloride salts have advantages in fast-spectrum molten salt reactors, having higher solubility for actinides than fluorides.  While NaCl has good nuclear, chemical and physical properties its high melting point means it needs to be blended with MgCl₂ or CaCl₂, the former being preferred in eutectic, and allowing the addition of actinide trichlorides. The major isotope of chlorine, Cl-35 gives rise to Cl-36 as an activation product – a long-lived energetic beta source, so Cl-37 is much preferable in a reactor.

All low-pressure liquid coolants allow all their heat to be delivered at high temperatures, since the temperature drop in heat exchangers is less than with gas coolants. Also, with a good margin between operating and boiling temperatures, passive cooling for decay heat is readily achieved. Since heat exchangers do leak to some small extent, having incompatible primary and secondary coolants can be a problem. The less pressure difference across the heat exchanger, the less is the problem.

The removal of passive decay heat is a vital feature of primary cooling systems, beyond heat transfer to do work. When the fission process stops, fission product decay continues and a substantial amount of heat is added to the core. At the moment of shutdown, this is about 6.5% of the full power level, but after an hour it drops to about 1.5% as the short-lived fission products decay. After a day, the decay heat falls to 0.4%, and after a week it will be only 0.2%. This heat could melt the core of a light water reactor unless it is reliably dissipated, as shown in 2011 at Fukushima, where about 1.5% of the heat was being generated when the tsunami disabled the cooling. In passive systems, some kind of convection flow is relied upon.

The top AHTR/FHR line is potential, lower one practical today. See also paper on Cooling Power Plants.

There is some radioactivity in the cooling water flowing through the core of a water-cooled reactor, due mainly to the activation product nitrogen-16, formed by neutron capture from oxygen. N-16 has a half-life on only 7 seconds but produces high-energy gamma radiation during decay. It is the reason that access to a BWR turbine hall is restricted during actual operation.

Nuclear reactors for process heat

Producing steam to drive a turbine and generator is relatively easy, and a light water reactor running at 350°C does this readily. As the above section and Figure show, other types of reactor are required for higher temperatures. A 2010 US Department of Energy document quotes 500°C for a liquid metal cooled reactor (FNR), 860°C for a molten salt reactor (MSR), and 950°C for a high temperature gas-cooled reactor (HTR). Lower-temperature reactors can be used with supplemental gas heating to reach higher temperatures, though employing an LWR would not be practical or economic. The DOE said that high reactor outlet temperatures in the range 750 to 950°C were required to satisfy all end user requirements evaluated to date for the Next Generation Nuclear Plant.

Primitive reactors

The world's oldest known nuclear reactors operated at what is now Oklo in Gabon, West Africa. About 2 billion years ago, at least 17 natural nuclear reactors achieved criticality in a rich deposit of uranium ore. Each operated intermittently at about 20 kW thermal, the reaction ceasing whenever the water turned to steam so that it ceased to function as moderator. At that time the concentration of U-235 in all natural uranium was about three percent instead of 0.7 percent as at present. (U-235 decays much faster than U-238, whose half-life is about the same as the age of the Earth.) These natural chain reactions, started spontaneously with the presence of water acting as a moderator, continued overall for about 2 million years before finally dying away. It appears that each reactor operated in pulses of about 30 minutes – interrupted when the water turned to steam, thereby switching it off for a few hours until it cooled. It is estimated that about 130 TWh of heat was produced. (The reactors were discovered when assays of mined uranium showed only 0.717% U-235 instead of 0.720% as everywhere else on the planet. Further investigation identified significant concentrations of fission products from both uranium and plutonium.)

During this long reaction period about 5.4 tonnes of fission products as well as up to two tonnes of plutonium together with other transuranic elements were generated in the orebody. The initial radioactive products have long since decayed into stable elements but close study of the amount and location of these has shown that there was little movement of radioactive wastes during and after the nuclear reactions. Plutonium and the other transuranics remained immobile.

Sources: 
Wilson, P.D., 1996, The Nuclear Fuel Cycle, OUP.
Scientific American 2005 article on Oklo
Technical and Economic Aspects of Load Following with Nuclear Power Plants, OECD Nuclear Energy Agency (June 2011)