European study on future fusion power plants
No permanent storage / attractive safety properties
Analysis of their ecological and economic properties has confirmed favourable results of previous investigations: Present know how indicates that accidents with severe impact on the environment in a fusion power plant are impossible and permanent disposal of waste is not necessary with recycling. The price of electricity will be equivalent to that of other environmentally benign energy technologies.
The aim of fusion research is to reproduce the generation of energy by the sun in a power plant on earth by deriving energy from fusion of atomic nuclei. The fuel is an ionised low density gas, a "plasma", composed of the two hydrogen isotopes, deuterium and tritium. This fuel is confined in a magnetic field and heated to ignite the fusion fire. Above a temperature of 100 million degrees the plasma starts to "burn": The hydrogen nuclei fuse to form helium, thereby releasing neutrons and large quantities of energy. The possibility of a fusion fire providing energy is to be shown by the international ITER (Latin for "the way") test device with a generated fusion power of 500 megawatts. ITER was planned on the basis of the materials and technologies available today, which are not yet fully optimised for fusion. This is the objective of a parallel physics and technology programme. All of this work is preparatory to a demonstration power plant; commercial plants could then supply the grid from the middle of the century.
Four models for a future power plant
The aim of the "European Fusion Power Plant Conceptual Study" is to sound out the economic and ecological properties that can be expected of a future power plant, and the lines of development that afford the greatest prospects. The latest research results were therefore taken as a basis for investigating four different concepts for a fusion power plant: All four models have an electric power of about 1500 megawatts and are of the "tokamak" type like ITER. To illuminate a wide spectrum of physical and technical possibilities, they are each based on different extrapolations of present day plasma physics and technology reaching variously far into the future.
In relation to ITER models A and B are the least far reaching: The assumptions on plasma behaviour, e.g. stability of the plasma, are only about 30 per cent better than the very cautious estimates for ITER. Unlike in ITER, the building material is a low activation steel now being investigated in the European Fusion Programme. The biggest differences relate to technical components of the power plant, e.g. the so called "blanket": This lining of the plasma vessel serves to decelerate the fast neutrons resulting from the fusion process. These transfer their entire kinetic energy to a coolant in the form of heat and also produce tritium as fuel component from lithium.
For these purposes model A is furnished with a liquid metal blanket: It uses a liquid lithium lead mixture for tritium production, and the fusion heat is absorbed and transferred with water. In contrast, model B is fitted with a blanket filled with pebbles of lithium ceramic and beryllium. The helium coolant chosen here allows higher temperatures than water does – up to 500 instead of 300 degrees centigrade – and hence higher efficiencies for the subsequent power production. Both blanket versions are being developed in the European Fusion Programme; test versions are to be investigated in ITER.
Unlike models A and B, the more far reaching model C and the rather more futuristic model D are based on major progress being made in plasma physics. Improved plasma states are combined with more powerful blanket concepts; but these are already being developed in Europe: In the dual coolant blanket of model C the first wall is cooled with helium; most of the heat generated is transported to the heat exchanger by circulation of liquid metal. Silicon carbide inserts insulate the structure from the flowing liquid metal. The higher coolant temperature of about 700 degrees allows more efficient conversion of fusion heat to electricity. Even more advanced in model C is the use of a self cooling blanket, liquid metal (up to 1100 degrees) serving for both cooling and tritium production; the structures consist of silicon carbide.
Considerations of safety are concerned with the radioactive tritium and the high energy neutrons, which activate the walls of the plasma vessel. The consequences of all serious accidents were clarified by analysing the two more contemporary models A and B in greater detail: Sudden and total failure of the cooling system is assumed to cause the accident; the power plant is then left to its own devices without any intervention. Result: Plasma instabilities impairing the operating conditions immediately extinguish the burning process; the residual heat in the walls is not sufficient to impair components severely or even melt them. The power plant does not contain any other energy source that could destroy its containment, which thus always remains intact.
It was then investigated how much tritium and activated material could be mobilised by the temperature rise and escape from the plant. Finally, the resulting radioactive exposure at the power plant perimeter was determined for the most adverse weather conditions: Models A and B have values well below one to two orders of magnitude the dose necessitating evacuation of inhabitants in the vicinity of the power plant. This also applies to model C; the values for model D are much lower still. This new study thus confirms the attractive safety properties known from previous investigations.
The waste situation was also reconsidered: The material activated by fusion neutrons was found to lose its radioactivity relatively fast in all four models. In a hundred years it drops to a ten thousandth of its initial value. In model B, for example, almost half of the material is no longer radioactive a hundred years after shutdown and can be passed for any other use. The other half could – with the advent of appropriate technology – be recycled and re used in new power plants: Permanent storage would then not be necessary. This also applies to the other three models.
Questions of cost
From model A to D there is an increase in the efficiency with which fusion energy can be tapped from the blanket and – with rising coolant temperature – the efficiency of power production. In addition, the plasma states attained from A to D become more and more favourable: The load on the walls decreases and less electric power has to be fed back into the power plant to keep it supplied. From model A to D less and less fusion power is needed to produce about 1500 megawatts of electric power. Furthermore, the plasma volume decreases from model A to D by more than half, i.e. plants can be made more compact.
Accordingly, different electricity prices are expected with the four power plant models: Model A entails the highest cost of electricity, followed by models B and C; avant garde model D costs the least. Even B and C would, however, be competitive with power production costs of 5 to 10 cents per kilowatt hour.
On balance the study indicates that already the first generation of commercial fusion power plants – as represented by the two more contemporary models A and B, whose development does not call for any major advances in plasma physics and materials research – will afford favourable safety and environmental properties and will be an economic proposition. Models C and D show the great potential for further physical and technological improvement.