Fusion decay heat

The full report is available here: Decay heat validation, FISPACT-II & TENDL-2014, JEFF-3.2, ENDF/B-VII.1 and JENDL-4.0 nuclear data libraries CCFE-R(15)25


Fusion Neutron Source experimental results

14 MeV neutrons are generated by a 2 mA deuteron beam impinging on a stationary tritium-bearing titanium target. The total neutron flux at the sample location, for this experiment, is in the range of 1.0×1010 [n cm−2 s−1], the same order of magnitude as in the first wall of the Joint European Torus (JET) fusion experiment when op- erating with D-T plasma. However, the irradiation time at the FNS were of 5 minutes and 7 hours in comparison with the few seconds flat burn achieved during the DTE1 JET fusion 1996 campaign. As a point of reference the total flux in a power plant is typically expected to be in the region of 1013 or 1015 [n cm−2 s−1], three to five or- ders of magnitude higher than in JET or FNS, and also for much longer irradiation times.

Thin samples, 25×25 mm2 in area, and typically 10 μm thick, have been used, either as metallic foil or powder sandwiched between tape. Use of a thin sample minimises the self-absorption of β rays emitted in the sample itself and allows their measurement. A total of 74 different materials have been used across the different phases of the experiment.

The decay energy in each irradiated sample was mea- sured in the Whole Energy Absorption Spectrometer (WEAS), which comprises two large bismuth-germanate BGO scintillators in a geometric arrangement, provides almost 100 % detection efficiency for both β and γ-rays. Correction factors need to be applied for γ-ray efficiency and for β and electron energy loss in the sample itself (less than 15% generally), and for other effects such as the decay heat due to the plastic tape used for the pow- der samples. The overall experimental uncertainty totals between 6 to 10% in most cases, although it rises to higher levels at particular cooling time for certain samples. The WEAS provides high sensitivity, down to powers less than 1 pW, which is valuable for measurement of some nuclides with long half-lives. It also has a wide dynamic range: measurements of up to a few mW have been achieved in the experiments.

The experimental time-dependent decay-power measurement program at JAEA FNS combined with the FISPACT-II simulations performed provide a unique check of the calculational method and nuclear databases associated with the prediction of decay power for the set of material samples analysed. The results of the comparison give confidence in most of the decay heat values calculated, although the predominantly 14 MeV neutron spectrum in FNS means that the low neutron energy reactions of importance in other devices have not yet been fully considered. This statement limits the scope of validation and possible conclusions reached in this study to the decay power predicted through the identified pathways. However, it covers the decay data of all the isotopes involved irrespective of their production routes.

The experimental uncertainty, calculational uncertainty and E/C values have been systematically produced. Their direct comparison demonstrates that the method chosen to calculate and propagate these calculational uncertainties in the FISPACT-II code system is verified and validated (V&V), and that the TENDL uncertainties file could be further improved along the same lines.

From the overall results, a set of inadequacies, not only in the cross sections but also in the decay libraries, have been identified that will require some corrective actions to be taken. These corrections and/or amendments will benefit the next generation of the TENDL library cross sections, associated variance and covariances, and decays data files. As expected, they impact both the production paths and/or decay data of some specific radionuclides without impairing the overall picture. A large proportion of the decay powers calculated in this validation exercise with TENDL-2015 are in good agreement (within a few %) with the experimental values for cooling times spanning from tens of seconds, and this is a unique insight in the isomers space, up to more than a year.