A fusion power plant will have to produce all its own tritium. Recently, popular science articles voiced doubts about the feasibility of this idea. However, measurements carried out at the Frascati Neutron Generator in Italy now demonstrate that the concept is valid and enough tritium can be produced.

A fusion reactor needs to produce all the tritium that it burns, plus a little bit more. For the design of a fusion power plant, this requirement is translated into a tritium production rate inside blankets that surround the burning plasma. The production depends on the neutron flux inside the blanket’s tritium breeding zone. Determining this neutron flux, however, is a complex task: It varies spatially inside the vessel and the neutrons interact with other wall materials before entering the breeding zones. Also, the blankets are designed differently depending on their location inside the vessel, because they have other functions, too. They absorb the energy of the fusion neutrons to heat a cooling fluid, which drives the turbine, and they shield the magnets from the high-energy neutrons. The achievable tritium breeding ratio for a fusion reactor is derived via simulation calculations. A burning plasma generating neutrons is simulated in the vessel (with a number of holes for heating and diagnostic systems) and with divertor and breeding blankets of a certain design. Based on the neutron flux one calculates the tritium production rate in the blankets and then adjusts the design of all components in order to meet the required value, without impairing all their other functions. The calculations need very accurate nuclear data and cross sections of the neutron reactions with all reactor materials. Most of the available data, however, were accurate for neutrons generated in fission reactors, and these carry much less energy than fusion neutrons. The project to develop a high quality nuclear database for fusion began under EFDA and now continues within the R&D program of Fusion for Energy.

Measurements prove concept

The measurements at the Frascati Neutron Generator (FNG) were carried out to validate the simulations and the neutron data used to calculate the tritium breeding ratio in fusion reactor design. Scientists from several European institutions – ENEA, KIT, Technical University of Dresden, Cracow University of Science and Technology and Jožef Stefan Institute of Ljubliana – were involved in these experiments. The experiments were carried out with mock-ups of the two breeding blanket concepts developed in Europe for a demonstration reactor DEMO. Even though in FNG the rate of neutrons per second and per area is much lower than in a fusion power plant, it produces neutrons with similar energy as the fusion reaction, 14 mega electron volts. The neutron energy spectra inside the mock-up breeding zone was the same as it will be in ITER. Over a large number of days the mock-ups were irradiated and the tritium produced measured. The measurements match the simulation calculations within 5.9% error. These results clearly show that the nuclear mechanisms behind the breeding blankets are well understood and therefore also demonstrate that a tritium breeding ratio above one – as predicted by simulations for DEMO – can be realised.
The next step will be to measure tritium breeding in the blanket test modules in ITER, to validate the calculations in a real tokamak environment. While FNG is a point neutron source with well defined conditions, the ITER plasma is a spatially distributed neutron source and creates a harsh environment (high heat and radiation), which also affects the measurement instruments. Therefore new types of detectors, which are expected to withstand these conditions and provide data with the required accuracy, have to be qualified for ITER experiments. These detectors have already been tested at FNG and will be further tested and qualified. Testing could be done also in a deuterium tritium campaign at JET, which is planned for around 2015.

Tritium and fusion power
A fusion power plant burns nuclei of the hydrogen isotopes tritium and deuterium and produces one helium nucleus and one fast neutron. To generate one gigawatt of electricity for one year, the plant consumes more than 55 kilos of tritium. Tritium is practically notn-existent in nature and the annual production of nuclear fission reactors worldwide is less than two kilos. Therefore the plant has to breed its own tritium using a reaction between the fusion neutron and the light metal lithium, which generates one tritium and one helium nucleus. The challenge lies in the fact that one burnt tritium yields one neutron, which generates exactly one tritium, so one would have to harvest 100 percent of the fusion neutrons for the breeding. Some neutrons, however, are lost in other wall areas. Moreover, the plant has to produce slightly more tritium than it burns to make up for losses and to keep a small buffer. The solution lies in neutron multipliers. These are elements like beryllium and lead, which react with neutrons and produce a second neutron.
The tritium breeding ratio of a fusion plant is the ratio between the amounts of generated and burnt tritium. This ratio must be defined very accurately, as the plant must not build up too much stock of the radioactive tritium (its half life time is 12.3 years). Scientists estimate that a breeding ratio of only very few percent above one is sufficient. The tritium production process takes place in the blankets, which are about one meter large, thick steel elements located at the wall inside the vacuum vessel. Europe is currently investigating two different design concepts for the tritium breeding blankets: One version, called Helium Cooled Pebble Bed, employs helium as a cooling fluid, ceramic lithium pebbles for breeding and beryllium pebbles as neutron multiplier. The other concept, called Helium Cooled Lithium Lead uses a lithium-lead-fluid to breed tritium and multiply neutrons.