The case against nuclear power: a summary
1. Greenhouse gas emissions from nuclear power
Nuclear power is not a zero carbon-energy source. To generate electricity from uranium, a complex chain of industrial processes is needed to:
- convert uranium ore in the ground into fuel elements for the reactor;
- construct the facilities including the nuclear power plant itself; and
- handle the wastes and to store the wastes in a save geological repository.
Operating a nuclear reactor is the only part of the chain which produces virtually no carbon dioxide.
A complete life-cycle analysis shows that generating electricity from nuclear power emits 20-40% of the carbon dioxide per kiloWatt hour ( kWh) of a gas-fired system when the whole system is taken into account (see Nuclear Power: the Energy Balance by Jan-Willem Storm van Leeuwen and Philip Smith).
The nuclear process chain also emits other greenhouse gases besides carbon dioxide with far stronger global-warming potential such as chloro- and fluorohydrocarbons and probably SF6. These emissions are difficult to quantify from the open literature, but the total emission of carbon dioxide equivalents by a nuclear system will be significantly more than 20-40% of a gas-fired system with the same energy output.back to top
2. Nuclear power in the world energy mix
Nuclear power currently provides a small fraction of total world energy consumption. The conventionally accepted figure is just over 6% of all non biomass fuel. But a more meaningful figure is about 2.5 % (statistical reviews such as BPs convert nuclear electricity into primary energy units tonnes of oil equivalent by multiplying the amount of nuclear electricity by three. This is misleading.)back to top
3. The energy required to extract uranium
Nuclear-power stations are fuelled by uranium, a metal found in the earths crust in various chemical compounds. Uranium is extracted from ore by mechanical and chemical processes, like copper from copper ore. The energy requirements of the extraction process depend on the grade of the uranium ore. For example, the extraction of 1 kilogram of uranium from ore containing 1 kg U per tonne rock consumes ten times as much than from ore of 10 kg U per tonne rock.
Moreover, the extraction yield falls with the ore grade, an unavoidable chemical phenomenon. At high ore grades (10 kg U per tonne rock or more), about 99% of the uranium can be extracted from the rock; but at low grades (e.g. 0.2 kg U per tonne rock) not more than about half of the uranium present in the rock can be extracted in practice.
Both factors increasing specific energy consumption, and falling yield with decreasing ore grade pose a limit below which an uranium-bearing rock can be processed but can no longer be considered a net source of energy. This threshold grade is about 200 grammes U per tonne rock (See here). The specific energy requirements for extraction of an element from a matrix rise exponentially with falling concentration of that element in its matrix. This is inherent to any extraction process. Advanced technology may lower the threshold grade from, for example, 0.02% to 0.015% (200 repectively 150 grammes uranium per tonne rock), but that wouldnt add significant uranium resources to the total energy resources.back to top
4. How much uranium is there?
At present, some 440 nuclear reactors are operating worldwide, with a combined capacity of some 363 GW(e). These reactors require about 67,000 tonnes of natural uranium per year. The present reserves and resources (to 80 US$/kg U) are about 3.5 million tonnes (see here). This , enough to last some fifty years at the current consumption rate.back to top
5. Uranium from granite
To fuel a nuclear reactor with nominal capacity of 1 GW(e) each year, about 162 tonnes natural uranium has to be extracted from earths crust. With an average uranium grade of four grams U per tonne of granite, 4o million tonnes of rock must be extracted to produce 162 tonnes of uranium.
The rock has to be dug up, ground to fine powder and chemically treated with sulfuric acid and other chemicals to extract the uranium compound from the mass. Assumed an extraction yield of 50%, 80 million tonnes granite have to be treated. This is a block 100 metres wide, 100 metres high and three kilometres long each year, for one reactor.
For comparison, a coal-fired power station of equivalent 1 GW(e) power consumes about 2 million tonnes of coal each year.back to top
6. Uranium from seawater
Seawater contains 3.3 milligram uranium per cubic meter seawater. The total volume of seawater of the world is estimated at 1.37 billion cubic kilometers, so the oceans contain some 4.5 billion tonnes of uranium. Technically it is possible to extract uranium from seawater.
To obtain 162 tonnes of uranium, about 170 cubic kilometers of seawater would have to be treated (assuming an rate yield of 30% can be achieved). That would amount to 5,300 cubic meters per second (yielding 5.3 grams uranium per second) continuously during a whole year, for one reactor. Very large energy and chemical inputs would be needed for this process.back to top
7. A nuclear renaissance?
Assume a thousand new nuclear power plants with a combined capacity of 1500 GW(e) will be built during the coming decades, as proposed in a 2003 study from the Massachussetts Institute of Technology. A park of this capacity would supply about 10% of the present world energy consumption (but less than 10% of total world energy consumption by the time the nuclear power comes on line because total consumption will have grown considerably during the intervening period).
Annual uranium consumption will then rise to some 250,000 tonnes and the known uranium reserves and resources will be exhausted in about fourteen years.
The MIT study argues that this would not present a problem as the market mechanism would step in: as uranium became scarcer, its price would rise.
Higher price means that more abundant ores with a lower uranium content would become economically mineable. These ores, such as shales and granites, would last for hundreds of years, says the study.
But this is a fallacy. The sole civil application of uranium is its use in power reactors to generate useful energy. The huge amounts of very lean ores that the MIT study and the World Nuclear Association refer to have grades well below what is known as the energy threshold. This means that these uranium-bearing rocks never will be a net energy source, whatever extraction technology is deployed, as is explained above.
Even if large new rich uranium deposits are found, doubling the known reserves, which is improbable from a geological point of view, the total reserves will last for fewer than thirty years in the scenario of the nuclear renaissance.back to top
Also on nuclear power in openDemocracys climate change debate:
Michael Davies & Antony Froggatt, Nuclear power: still no thanks
8. Breeder reactors
All nuclear reactors currently in regular operation are thermal reactors, most of them light-water reactors (LWRs). This type of reactor can only fission 0.6%-0.7% of natural uranium. Theoretically, a breeder reactor is able to fission about 60% of natural uranium atoms via conversion into plutonium. The high figure of the breeder is the source of old nuclear dreams of the all nuclear society and burning the rocks.
The breeder is not just a reactor type. It is a system or cycle made up of three components. All three components need to operate flawlessly and exactly in tune with each other before any breeding is achieved.
First, the breeder reactor, which generates more fissile atoms (plutonium) from non-fissile uranium-238 atoms than it consumes by fissioning.
Second, the reprocessing of the spent fuel to separate the plutonium and remaining uranium from the fission products and unusable, but nasty and dangerous transuranic elements.
Third, the fuel-fabrication facility to make new fuel elements from the highly radioactive plutonium and recycled uranium from the reprocessing plant. Not one of the three components has ever been demonstrated to operate smoothly, let alone the three components together as a fine-tuned continuously operating system.
Fifty years of intensive research in seven or eight countries, with tens, if not hundreds of billions of dollars invested have so far failed to demonstrate that the breeder cycle is feasible in practice. The same holds true for the thorium cycle, which is even more difficult to develop.
The MIT study The Future of Nuclear Power does not expect breeder reactors will come into operation during the next three decades.back to top
9. Thermonuclear fusion
Thermonuclear fusion is the energy source of the sun. For man-made fusion reactors only the D-T reaction (deuterium-tritium) is practicable. No uranium or plutonium is needed for this kind of nuclear reaction. Deuterium can be extracted from seawater, tritium has to be breeded from lithium. Although the principle of controlled fusion (other than explosions of hydrogen bombs) has been demonstrated, still no reactor exists which produces more energy than it consumes.
For fifty years research on nuclear fusion has been conducted in the United States, European Union, Japan and Russia, with investments of many tens of billions of dollars. A German study (Nuclear fusion status report Arbeitsbericht Nr. 75, Office for TechnologyAssessment, Bundestag, 2002) concludes that the first fusion reactor producing net electricity may be built around 2050.
As this study put it:
To achieve this programme, very substantial scientific and technical challenges must be met. The R&D required will take several decades and demand funding on a large scale. Over almost 50 years in which fusion research has been going on, the difficulties in developing a fusion plant have been repeatedly underestimated, with the result that the horizon for implementation had to be pushed further and further into the future, becoming in effect a moving target.
(see also Thermonuclear fusion; as of March 2003)back to top
10. Nuclear power from the sun
Humanity has a perfectly functioning thermonuclear fusion reactor at its disposal. The reactor delivers its energy to man in a constant, abundant flow of clean, benign electromagnetic radiation, without radioactive wastes and harmful radiation.
Solar energy can be harvested in several ways. In order to get an impression of the potential, we compare the costs of electricity generation by photovoltaic (PV) panels with the costs of nuclear-generated electricity. We start this rough calculation at the nuclear-renaissance scenario of MIT, as described above.
It seems unlikely that the construction costs of new nuclear power plants will be lower than those during the last construction period in the US in the 1980s. Based on those empirical values, the construction costs of 1,500 GW(e) nuclear capacity may be estimated at some $7,500 15,000 billion (with the higher figure being more probable).
$15,000 billion would be sufficient to construct enough PV system to generate about 23 EJ/a (exajoule per annum), calculation based on the current state of technology, costs and efficiency. The total world primary energy consumption in 2003 was around 409 EJ/a, excluding biomass. Accounting for the learning curve effect and the expected doubling of the conversion efficiency of PV systems within two decades, a system with a capacity of some 90 EJ/a can be built with the same investment. This would supply 15-20% of assumed world energy demand by 2030.
A nuclear park with a capacity of 1,500 GW(e) produces about 38 EJ/a electricity (assuming an average load factor of 80%). So the nuclear system would produce more electricity than the PV system (current technology) for the same money? No.
The energy source of the PV system, the sun, is free and has a constant flow and a constant quality. During the lifetime of the system no costs other than for maintenance of the system are required.
Nuclear power, on the other hand, consumes uranium, which has to be mined from ever deeper mines and extracted from ever leaner ores. It produces an ever growing mass of hazardous, radioactive wastes, which have to be packed and placed in a safe repository. The lifetime costs of all processes needed to run the nuclear system, apart from operation and maintenance and to clean up its wastes, surely will rise to a multiple of the initial construction costs.
Assuming that the lifetime costs of the nuclear park are three times the initial construction costs, a PV system (current technology) with a production rate of at least 70 EJ/a could be built for the same costs. In such a large development and construction project the expected higher efficiency and lower specific costs almost certainly will be achieved. If the calculation is based on that figure, a solar nuclear energy conversion system with a capacity of about 270 EJ/a can be built and operated, more than half of the world energy demand.back to top
The sole energy systems having no influence on the climate are based on solar energy like photovoltaics, biomass and wind. The potential of these really sustainable systems is many times the present world energy demand.
In the short term, the largest and cheapest gains in mitigating climate change problems are to be made in energy conservation, especially in transport.
In longer term assessment of energy systems and energy services, all external costs should be internalised, including costs that will be borne decades from now, such as handling of nuclear wastes and dismantling of nuclear-power stations.back to top
The case against nuclear power: a summary
Nuclear power is a small player in the world energy mix. The current contribution of nuclear power to mitigation of anthropogenic carbon dioxide emission is from 2-6% of total man-made emissions from non-biomass resources, depending on the definition used.
Transport contributes well over 30% of total world emissions of carbon dioxide. Doubling the efficiency with which energy is used in transport would be much more effective in reducing greenhouse gas emissions than doubling the size of the worlds nuclear capacity, and much cheaper.
Known uranium reserves will last for about fifty years at the current consumption rate.
If the nuclear share is increased to 10% of the current world energy supply by construction of 1500 GW(e) new capacity (more than four times the current world nuclear capacity), the reserves will last for about fourteen years.
If large new rich uranium deposits will be found, doubling the known reserves, which is improbable from a geological point of view, global reserves will last for less than thirty years.
Uranium-bearing rocks with grades lower than of the leanest ores processed at present, should not be regarded as energy source, because extraction of the uranium from the rock consumes more energy than can be generated by fissioning that uranium in a reactor. The energy threshold is an inherent feature of uranium ores and practically independent on the state of extraction technology.
The economic price mechanism will not open up new large uranium resources that may be considered energy sources, because their grades are below the energy threshold.
Breeder technology has proven to be unfeasible and it never will be feasible.
Fusion for energy generation (other than thermonuclear weapons) has been a moving target for fifty years and will be so for at least the next fifty years.
For the total costs of a new nuclear-power plant, a photovoltaic system can be constructed with an electricity generating capacity of two to seven times as much.
The nuclear industry and its allies are familiar with all these facts. Their real motives may therefore be quite different from a solution to the energy and climate change challenges of the world.
back to top
This article appears as part of openDemocracys online debate on the politics of climate change. The debate was developed in partnership with the British Council as part of their ZeroCarbonCity initiative a two year global campaign to raise awareness and stimulate debate around the challenges of climate change.