A series of verification and validation (V&V) reports available on the documentation page have covered the full range of aspects of FISPACT-II, including direct nuclear data validation, verification of processing and benchmarking against respected experimental results.
For FISPACT-II, V&V employs a set of tools involving integral and differential data for:
- Differential and integral validation of input nuclear data
- Verification of technological nuclear data using systematics and statistical checks
- Verification of the processing of the nuclear data files by FISPACT-II and utility codes
- Validation of code simulations against experimental data
This represents a full and consistent approach which covers the full range of simulation process, from the fundamental ingredients to final product. Each V&V exercise uses the most complete and robust data sets for comparison, for example drawing upon all of EXFOR, fusion activation experiments or fission decay heat measurements. Each of these topics is the subject of a report and are described in links below.
Decay heat and inventory calculations for irradiated fission fuels comprise two of the fundamental tasks for time-dependent Bateman solvers in the nuclear industry. Detailed and accurate knowledge of these time-dependent characteristics, as well as trustworthy uncertainty values, are of primary importance for reactor safety cases and the handling of irradiated fuel – issues which cover a great many
activities representing billions of euros in current and future effort.
Development of the FISPACT-II code has resulted in new and unique simulation methods for a variety of nuclear observables, including fission decay heat and inventory calculations. To perform these simulations, massive libraries which contain the complete probability distributions for fission product formation, as well as the complete decay data for all of these products (reaching from the long-lived to those with sub-second half-lives), must be maintained and validated with sophisticated and sturdy simulation software. All of the physics of nuclear interactions, fissions and decays is contained within the nuclear data files, which hide one half of the simulation within the evaluation method behind those files.
While most time-dependent inventory and observables codes rely upon one bespoke nuclear data library, the ability to harness any dataset affords a unique opportunity to cross-check data and provide feedback which ultimately improves the code/data system. By performing a verification and validation on FISPACT-II with all of the major international nuclear data libraries, this exercise goes beyond demonstrating the capabilities of the code/data system in simulating decay heat and inventories, giving precise information on which nuclides should have their fission yield or decay data re-evaluated and in which library.
Little experimental data exists for structural material samples irradiated under all nuclear plants relevant neutron spectra and even when data does exist the measured quantities are either specific activity and/or gamma spectroscopy. In particular, no or very little experimental data on decay power has previously existed for fission plant structural materials and for materials under high energy irradiation conditions (i.e. fusion, fast fission). It was to fill this gap that a series of experiments were performed using the Fusion Neutron Source (FNS) facility at the Japan Atomic Energy Agency.
Material samples were irradiated in a simulated D-T neutron field and the resulting decay power was measured for cooling times of up to thirteen months. Using the highly sensitive Whole Energy Absorption Spectrometer (WEAS) method, both beta and gamma emission decay energies were measured at selected cooling times and, quite impressively, as soon as a few tens of seconds after the end of irradiation.
Validation of decay power predictions by means of direct comparison with integral data measurements of sample structural materials under neutron spectra allow confidence to be given to the decay power values calculated. It also permits an assessment of the adequacy of the methods and nuclear data, and indicates any inaccuracy or omission that may have led to erroneous code predictions. It is clear that certain safety margins can be derived from such a validation exercise, if relevant to plant operation, materials and design, and applied as bounding conditions in operational Safety and Environmental (S&E) analyses.
For a great many target nuclides and reactions, there are few experimental measurements to rely upon in the nuclear data evaluation process. Many energies are too difficult to probe, for example resonance regions or energies between a few MeV and 14 MeV for neutrons. These limitations in the data force us to be more proactive in validation, drawing upon different, complementary sources to draw conclusions where individual sets of differential measurements are lacking.
A series of irradiations of various materials in several complementary neutron fields have been carried out over several decades. Analyses of the results have produced integrated effective cross-sections attributed to various nuclear interactions.
Neutron spectra calculated for each experiment can be convolved over energy with library cross-sections for comparison with experimental results. The measurement techniques vary between experiments, from calorimetric to spectroscopic, fairly mono-energetic to ‘white’ spectra. Each presents its own challenges, but the extraction of useful data on individual reaction channels can be done even in calorimetric measurements such as those performed with total decay heat from FNS.
To best gauge the quality and extent of the conclusions that can be drawn from the available set of integral measurements, differential data from EXFOR is compared against the evaluated cross sections with all isomeric production, if present. The combination of these has great value in highlighting areas for re-evaluation or providing the most robust activation validation.
Systematic verification of nuclear data
As a truly general purpose nuclear data library, TENDL provides complete data for all target nuclides and reaction channels. This is a double-edged quality, offering new and robust information where none would otherwise exist, but also potentially exposing less predictive regimes of the code system, faults in the input reference data or other general errors in the library production. The verification of the global nuclear data poses a complex question due to the inability of testers from manually checking every file, channel and data entry. To address this, a series of checks which probe different aspects of the files were devised. These have detected some anomalous issues in previous TENDL versions which have addressed in the recent releases.
Aside from these global checks, there are established datasets for thermal cross sections, resonance integrals and Maxwellian-averaged cross sections. The former two are contained in the well-known Atlas of Neutron Resonances, while the latter is provided by the KADoNiS database. The feedback from these tests on the TENDL neutron files resulted in indicated that some unresolved issues with input libraries remained. These have been comprehensively dealt with in the most recent versions.