Requester: at ,
Email:

Project (context):

Impact:
There is currently large uncertainty in reactivity for chloride fast reactors, due mainly to the uncertainty in the Cl-35 (n,p) cross section at high energies. This is a difficulty for chloride fast reactor design because the required fissile loading has large uncertainty, and this uncertainty also propagates to other parameters important to safety, such as reactivity coefficients. Production of S-35 via this reaction is also important for corrosion. Additionally, above 1.2 MeV there is no uncertainty available for Cl-35 (n,p) in any evaluated library, except TENDL (as far as I am aware). This makes it difficult to justify a reasonable total uncertainty. A number of organisations are designing or studying chloride fast reactors and are therefore sensitive to this nuclear data. In addition to Moltex Energy and Terrapower (MCRE and MCFR) and has also seen evidence that the CEA are performing chloride fast reactor R&D. Previous work led by EDF R&D on the REBUS-3700 fast chloride reactor also concluded that more accurate nuclear data for Cl was needed (https://doi.org/10.1016/j.enconman.2006.02.013). It is also worth noting that other Cl nuclear data is important for chloride fast reactors, particularly Cl-35 (n,gamma) and Cl-36 capture, for Cl-36 waste considerations.

Accuracy:
To achieve a target k-eff uncertainty < 300 pcm, table (i) below suggests uncertainties < 2% would be required above ~100 keV. However, such a low uncertainty is unlikely to be achievable using differential measurements. k-eff uncertainty > 300 pcm would be tolerable, given the uncertainty is currently estimated to be at least ~1000 pcm. On this basis a target of 5 – 8 % is suggested. It is realised that it may be challenging to achieve even this uncertainty. Any reduction in uncertainty, and/or improved covariances, would be valuable. Integral experiments would likely be needed to reduce uncertainties to a level of 2 – 3 %.

Justification document:
33 group sensitivity coefficients have been generated for the main output parameters, as part of a collaboration with ANL (using the PERSENT code and ENDF/B-7.0 data). Those for Cl-35 (n,p) are large, as expected. Uncertainties and target uncertainties have then been derived by UPM through WPEC SG46, using this sensitivity data. Using the TENDL-2021 evaluation for Cl-35, and ENDF/B-7.1 otherwise, gives a total uncertainty of 836 pcm, with 631 pcm from Cl-35 alone, and dominated by the (n,p) cross section (595 pcm, next largest 173 pcm from (n,alpha)). Below is a table showing target accuracy requirements for the top 10 most important reactions for the Moltex SSR-W. Recent measurements in the US ((Batchelder, 2018) https://doi.org/10.1103/PhysRevC.99.044612 and (Kuvin & Lee, 2020) https://doi.org/10.1103/PhysRevC.102.024623) also indicate that the cross section in all libraries could be too high above ~1.2 MeV. Direct perturbation of the Cl-35 (n,p) reaction in this energy range shows large sensitivity (increase of ~1000 pcm for reduction by 50% between 2.23 and 3.68 MeV). TerraPower indicates also that very recent measurements ((Warren, 2021) http://rave.ohiolink.edu/etdc/view?acc_num=ohiou1606843490983827) of the 35Cl(n,p) cross section have significantly disagree with the evaluated nuclear data in ENDF/B VIII.0 that was not updated in the last update of chlorine nuclear data – ENDF/B VII.1 (Chadwick, 2011). The attached Figure 1 presents the discrepancy between ENDF/B-VII.1 and recent LANL measured cross sections. Other measurements made by UC Berkeley and Ohio University show similar resonance behavior. The resonance at 2.639 MeV Berkeley measured aligns with a similarly sized resonance in LANL’s cross section measurements (Batchelder, 2018) (Kuvin & Lee, 2020).

Comment from requester:
-------------------------Upfile: Upfile:

Review comment:

Entry Status:

Main references:

Additional file attached:
Additional file attached: