95-Am-241(n,γ)Am-242g,m

Requester (Each request should be "owned" by a single individual)

David BERNARD and Vanessa VALLET, CEA/DES/IRESNE Cadarache, FRANCE

Øystein BREMNES, EDF/DT Lyon, FRANCE

Quantity (Quantity of the reaction process that is of concern to the requester, e.g., cross section, resonance parameter, spectrum, average multiplicity, etc.)

Isomeric ratio to the third level (ENSDF) = metastable (ENDF/LISO=2) state J^π = 5^- and to the ground state J^π = 1^- of ²⁴²Am with half-lives of 141 years and 16 hours respectively.

Impact documentation (Impact of the requested data improvement for the application in terms of safety, reliability, cost, etc.)

The reduction of margins (inducing costs reduction) for LWR/MOx decay heat is crucial for industrials. Less concerned is the UOx fuel for which any new experiment will also marginally improve the prediction of its decay heat. The second justification concerns the ability to transmute ²⁴¹Am in Gen-IV actinide burner reactors.

Requested accuracy on the reaction quantity (in %, split into different energy/angle ranges if necessary)

According to the target of an additional penalty/margin of +0.5%@2σ for MOx decay heat compare to UOx decay heat from one hand, and to the averaged sensitivity of MOx decay heat after 10 days to 1.5 years to ground state isomeric ratio (²⁴²Cm build-up) of approx. 0.3%/% in the other hand, the 2σ-requested relative accuracy for ground state feeding is about 1.7%. If considering that isomeric ratio to ground state actual values are close to 0.87, the 1σ-absolute requested accuracy for ground and metastable states feeding is ~0.007.

Justification documentation (The need for these data, inadequacy of existing information, etc., should be clearly established – Quantitative support, e.g. from sensitivity studies, is required for high priority requests)

The MOx decay heat differs from the UOx one in a sense that higher amount of ²⁴²Cm is produced when starting the production of minor actinides from Pu (Z=94) instead of U (Z=92). Hence, for MOx cycle studies (accidental conditions, margins to account for…especially without integral experimental validation like for UOx fuels [GE, HEDL, MERCI]), the ²⁴²Cm amount becomes a key parameter as its contribution to decay heat (between 10 days and 1.5 years) after reactor shutdown is about 20 to 34% (see Fig. 1). The main route for its build-up is through consecutive neutron radiative capture reactions, and especially the last one: ²⁴¹Am(n,γ)^242gAm followed by radioactive β- decay to ²⁴²Cm with 16h half-life.

Fig. 1: MOx decay heat isotopic-breakdown after 3 irradiation cycles

Quasi-all nuclear data involved in ²⁴²Cm build-up are rather well known:

1. ^242gAm half-life,
2. ^242gAm ε+β⁺/β^- decay branching ratio reported to be 0.171±0.001 to ε+β⁺ in [Bernard] for which the uncertainty is 3 times lower than those reported in current (ENDF/B-VIII.0 = ENSDF/31-Dec-2021) Radioactive Decay Data evaluation files,
3. ²⁴¹Am(n,γ) cross-section as reported in [OECD/NEA/WPEC-41] and included in recent general purposes evaluation.

Except one which is the ²⁴¹Am(n,γ) branching ratio (or isomeric) for which, except for JEFF-3.3 (inherited from the older version JEFF-3.1.1), no covariances are given. As ²⁴¹Am reaction rate are partly thermal and partly resonant in LWR, we need an accurate description of its presumed fluctuations from a resonance to another one (for a given J^π to another one): see Fig. 2. These fluctuations are of importance since the neutron spectrum is going softer (²⁴¹Am reaction rates spectrum as well) as a function of the exposure.

Fig. 2: ²⁴¹Am(n,γ)^242mAm branching ratios as reported in recent General Purposes files.

The value at thermal energy of the branching ratio to ^242mAm is available from several (semi-)integral measurements, e.g., 0.10 ± 0.09 [Shinohara], 0.086 ± 0.007 [Fioni], 0.105 ± 0.004 [Bringer]. As far as the authors know, the only measurements beyond the thermal energy for this data are reported in [Bernard] where the three integral values for quasi thermal to quasi epithermal spectra are 0.109±0.002, 0.133±0.003 and 0.164±0.004, respectively. The difference between the two extreme values is leading to a modification of the MOx decay heat by about ~1.5%@T_shutdown+15 days and 2.2%@T_shutdown+115 days (almost 3 times the target accuracy of decay heat for this parameter alone). This is not strictly speaking a sensitivity study but it serves as a global uncertainty for this unique parameter. In addition, the difference (on isomeric ratios only) between ENDF/B-VIII.0 and JEFF-3.1.1 leads to a difference of +1% on MOx decay heat (and -18% on ^242mAm content at high burnup).

The second justification concerns Gen-IV reactor (and actinide burner especially). At higher energy, the branching ratio is composed by individual s-, p- and d- neutrons waves starting to be opened after an incident energy of about 30keV. Thus, we may expect different values in the URR than in RRR. An averaged value of 0.15 for the branching ratio to ^242mAm was consistent with the results of the PROFIL experiments [PX]. This value should be valid for a standard Na-cooled fast reactor, but it has to be confirmed from additional integral measurements in different neutron spectra and/or differential measurements. This is important to estimate the ability of burner to close the fuel cycle.