Benchmark Calculations in Support for FENDL-3 Generation

Benchmark Calculations in Support for FENDL-3

Generation

Mohamed SawanTim Bohm

Fusion Technology InstituteThe University of Wisconsin-Madison

2nd RCM on FENDL-323-26 March 2010

IAEA, Vienna

Outline

• Application of ENDF/B-VII.0 to FNG integral experiments

• Proposed material addition to FENDL-3• Benchmark analysis for initial FENDL-3

startup library (release 2)• Problems with processed data• Benchmark analysis for corrected

FENDL-3 startup library

M. Sawan 2nd RCM on FENDL-3, 23-26 March 2010, IAEA, Vienna 2


Data Source for FENDL-2.1

Total 71 elements/isotopes


Majority (40) of materials in FENDL-2.1 taken from ENDF/B-VI.8

We investigated the effect of the recently released ENDF/B-VII.0 data (December 2006) on results for the FNG ITER relevant integral experiments

Examined tungsten, bulk shielding, streaming and HCPB breeder mock-ups

Compared calculated to experimental results (C/E) for 3 cases

⁃ FENDL-2.1⁃ All FENDL-2.1 data replaced by ENDF/B-VII.0⁃ FENDL-2.1 with data for the 40 materials only

replaced by ENDF/B-VII.0

ENDF/B-VII.0 Applied to ITER Relevant FNG Experiments

FNG Experiments Analyzed


Tungsten

Bulk shielding

Streaming

HCPB breeder

Tungsten experiment activation foils


Reaction Ethreshold (MeV) FENDL-2.1 ENDF/B-VII Partial ENDF/B-VIIMn-55(n,)Mn-56 0 1.454/1.785 1.450/1.748 1.453/1.771Au-197(n,)Au-198 0 0.981/0.905 0.984/1.097 0.982/1.093Ni-58(n,p)Co-58 0.8 1.061/1.094 1.066/1.119 1.049/1.075In-115(n,n')In-115m* 0.8 0.929/0.869 0.923/0.864 0.924/0.862Al-27(n,)Na-24 3 1.030/1.055 1.041/1.058 1.031/1.047Fe-56(n,p)Mn-56 3 0.961/0.945 0.962/0.941 0.958/0.930Ni-58(n,2n)Ni-57 10 1.020/1.034 1.040/1.069 1.036/1.049Zr-90(n,2n)Zr-89 10 1.034/1.049 1.036/1.055 1.043/1.057Nb-93(n,2n)-Nb92 10 1.024/1.080 1.013/1.051 1.011/1.056

Average/Maximum of Calc/Exp

Bulk Shielding experiment activation foils


Reaction Ethreshold (MeV) FENDL-2.1 ENDF/B-VII Partial ENDF/B-VIIMn-55(n,)Mn-56 0 0.902/0.763 0.890/0.722 0.903/0.743Au-197(n,)Au-198 0 0.910/0.789 0.908/0.790 0.922/0.802Ni-58(n,p)Co-58 0.8 0.944/0.728 0.949/0.634 0.940/0.701 In-115(n,n')In-115m* 0.8 0.751/0.501 0.748/0.450 0.751/0.500 Al-27(n,)Na-24 3 0.912/0.848 0.917/0.741 0.903/0.787 Fe-56(n,p)Mn-56 3 0.897/0.827 0.921/0.805 0.890/0.786Ni-58(n,2n)Ni-57 10 0.923/0.873 0.933/0.889 0.924/0.876Zr-90(n,2n)Zr-89 10 - - -Nb-93(n,2n)-Nb92 10 0.916/0.782 0.899/0.710 0.901/0.776


Streaming experiment activation foils


Reaction Ethreshold (MeV) FENDL-2.1 ENDF/B-VII Partial ENDF/B-VIINi-58(n,p)Co-58 0.8 0.991/1.099 0.990/1.104 0.987/1.100Al-27(n,)Na-24 3 0.869/0.703 0.888/0.712 0.877/0.713Nb-93(n,2n)-Nb92 10 0.911/0.841 0.902/0.825 0.902/0.827


HCPB Breeder activation foils


Reaction Ethreshold (MeV) FENDL-2.1 ENDF/B-VII Partial ENDF/B-VIIAu-197(n,)Au-198 0 0.904/0.829 0.921/0.875 0.928/0.853Ni-58(n,p)Co-58 0.8 0.982/0.928 0.998/0.942 0.978/0.935Al-27(n,)Na-24 3 1.001/1.034 1.008/1.037 0.998/1.034Nb-93(n,2n)-Nb92 10 0.977/0.930 0.979/0.939 0.977/0.929


HCPB Breeder tritium production


Depth (mm) FENDL-2.1 ENDF/B-VII Partial ENDF/B-VII42 0.855/0.817 0.862/0.830 0.853/0.815

105 0.929/0.865 0.934/0.876 0.928/0.870168 0.939/0.869 0.933/0.856 0.938/0.863231 0.931/0.852 0.925/0.859 0.919/0.838

Average/Maximum of C/E values at each 12 pellet ENEA group

Finding


Calculations of foil activation and tritium breeding for ITER relevant FNG integral experiments yield nearly similar results for FENDL-2.1 and ENDF/B-VII.0


FENDL-3/SLIB Starter Library

• A starter library (FENDL-3/SLIB) was generated based on several agreed upon rules of creation– Replace present evaluations with updates– Adopt evaluations from libraries that contain standards (H-1 from ENDF)– Use isotopic evaluations where available

• The library includes 88 isotopes with updated evaluations from ENDF/B-VII.0, JENDL-HE, JEFF-3.1, and BROND

• Only evaluation switch between libraries occurred for H-1 and He-3 (JENDL-3.3 ENDF/B-VII.0) that could impact analysis of systems with water-cooling like ITER

• Sn is the only material with elemental rather than isotopic evaluation• FENDL-3/SLIB materials

– 47 from ENDF/B-VII.0– 35 from JENDL-HE– 4 from JEFF-3.1– 2 from BROND-2


Proposed Added Materials in FENDL-3

• Following the 1st RCM meeting in December 2008, we reviewed the list of material in the FENDL-2.1 general purpose data library regarding its completeness to satisfy the needs for fusion neutronics analysis

• We solicited input from the fusion neutronics community for possible addition of materials that are needed and are missing from the FENDL-2.1 library

• 17 elements were identified for consideration to be added in the new library to enhance its usefulness for the different fusion applications

• We identified the application needed for each proposed material and gave a priority to help us with decision if we want to limit the number of materials added

• Availability of evaluations for these elements was determined• In addition, 6 isotopes are missing from FENDL-2.1 and need to be

considered for inclusion in FENDL-3

17 Elements Proposed for Addition in FENDL-3


17 Elements Proposed for Addition in FENDL-3 (continued)



17 Elements Proposed for Addition in FENDL-3 (continued)

• These cover all materials requested by ITER (M. Loughlin) except Cs, U, Th, Np• Additional materials needed for IFMIF analysis were provided by U. Fischer

Six Isotopes Missing for Elements in FENDL-2.1

17

Abundances for C-13 and Pb-204 are significant and inclusion in FENDL-3 should be considered

O-18 and V-50 have low abundances and no evaluations and could be ignored in FENDL-3

Although O-17 and W-180 have very low abundances, evaluations exist for them and could be considered for FENDL-3

M. Sawan 2nd RCM on FENDL-3, 23-26 March 2010, IAEA, Vienna

M. Sawan 18

ITER Calculational Benchmark

To quantify impact of data changes, we performed MCNP calculations for a 1-D ITER calculational benchmark that was utilized during FENDL development process

Calculations carried out using FENDL-2.1 library, FENDL-2.1 with data for the 40 materials replaced by ENDF/B-VII.0, and the FENDL-3/SLIB library (release 2 posted 6/24/2009 designated SLIB2)

Results for flux, heating, dpa, and gas production were compared

2nd RCM on FENDL-3, 23-26 March 2010, IAEA, Vienna


Findings of Data ComparisonBetween ENDF/B-VI.8 and ENDF/B-VII.0

Minor impact on ITER nuclear analysis is expected except for ITER-TBM nuclear analysis due to changes in data for Li-6, Pb-208, and F-19

Effects of changes could be large in other fusion systems

• Power plants with breeding blankets• Inertial fusion systems (e.g., H-3 and Au-197

data are important for ICF target neutronics)1. M.E. Sawan, “Benchmarking FENDL-2.1 with Impact of ENDF/B-VII.0 Release on Nuclear

Heating and Damage,” Proceedings of the 23rd IEEE/NPSS Symposium on Fusion Engineering (SOFE), May 31-June 5, 2009, San Diego, CA, ISBN No. 978-1-4244-2636-2, IEEE Cat. No. CFP09SOF-CDR. Also University of Wisconsin Fusion Technology Institute Report, UWFDM-1361 (May 2009)

2. M. Sawan, “Impact of ENDF/B-VII.0 Release on Fusion Evaluated Nuclear Data Library FENDL-2.1,” Fusion Science & Technology, vol. 56, pp 766-770 (August 2009)


Peak Neutron and Gamma Flux Results

• Using ENDF/B-VII.0 data results in slightly higher flux values. However, the change is <1% with much smaller differences at the front FW zones facing the plasma

• Using FENDL-3/SLIB2 results in much higher flux values particularly at the magnet that is heavily shielded by water-cooled SS shield and vacuum vessel

M. Sawan 21

Neutron Spectra

• Using FENDL-3/SLIB2 results in significantly softer neutron energy spectrum at the heavily shielded outboard magnet that yields enhanced gamma generation

• At OB magnet E>0.1 MeV flux increases by only 3% while the E<0.1 MeV flux increases by 82%



Peak Nuclear Heating Results

• While using ENDF/B-VII.0 data results in slightly higher heating values (<1%) using FENDL-3/SLIB2 results in much higher heating values particularly at the magnet

• Most of increase in outboard magnet heating is due to gamma heating increase (21% increase vs. 5% increase in neutron heating)


Peak Magnet Nuclear Parameters

• While using ENDF/B-VII.0 data results in slightly higher magnet parameters using FENDL-3/SLIB2 results in much higher parameters with the largest increase in magnet heating due to enhanced gamma production

FENDL-3/SLIB2 Calculations for FNG Bulk Shield Experiment


Small difference compared to FENDL-2.1 Bulk shield mock-up thickness is much smaller than magnet

shielding in ITER


To further investigate the increase in low energy (E<0.1 MeV) neutron flux observed in the ITER 1-D calculational benchmark with FENDL-3/SLIB2 we performed simple calculations for slabs with all water, all SS316, and SS316/H2O mixture

Compared neutron fluxes in the various slabs calculated with MCNPX using FENDL-2.1 versus FENDL-3/SLIB

Used isotropic 14.1 MeV neutron source

5%increase

FENDL-3/SLIB2 in Slabs of H2O, SS316 and SS/H2O (80/20)

FENDL-3/SLIB2 in Slabs of H2O, SS316 and SS/H2O (80/20)

M. Sawan 26

30%increase

750%increase

Observed large differences are not related to change in H data but are related to SS data


FENDL-3/SLIB2 in Mo Slab


500%increase

• Performed calculations for slabs made of each constituent element of SS316

• Largest difference obtained in Mo (JENDL-3.3JENDL-HE)

• Mo resonances are below 0.1 MeV compared to higher energy resonances for Fe

• Resonance data might have changed or data processing with NJOY might have an error

Revised FENDL-3 Starter Library


Independent assessment by Andrej Trkov confirmed the large differences

An anomaly was noticed in plots of neutron emission spectra

The symptoms suggested that it is likely an NJOY processing problem which is very specific for the combination of data representation in the JENDL-HE files

A temporary patch to NJOY was proposed that solves processing problems for all affected materials

New processed ACE files were posted on 2/16/2010

We performed calculations with the corrected FENDL-3 starter library designated FENDL-3/SLIB2a and compared results to FENDL-2.1 and the previous FENDL-3/SLIB2a


Peak Neutron and Gamma Flux Results with FENDL-3/SLIB2a

• Using the corrected FENDL-3 starter library (FENDL-3/SLIB2a) removed the large overestimate at deep penetration locations

• Neutron and gamma fluxes at VV and magnet that are heavily shielded by water-cooled SS are higher by 1-3% than predicted by FENDL-2.1

M. Sawan 30

Neutron Spectrum at OB Magnet with FENDL-3/SLIB2a

• Using FENDL-3/SLIB2a corrected the anomaly in low energy spectrum caused by wrong secondary energy spectra

• Energy spectra from FENDL-3/SLIB2a and FENDL-2.1 are close with differences up to ~10% at E>2 MeV



Notable change in FENDL-3 is switch for H-1 from JENDL-3.3 to ENDF/B-VII.0

Lower (n,) cross sections at high energy could be reason for higher neutron flux (particularly at high energy) in regions heavily shielded by water-cooled SS

H-1 Data in FENDL-3 vs. FENDL-2.1


Peak Nuclear Heating Results with FENDL-3/SLIB2a

• Using the corrected FENDL-3 starter library (FENDL-3/SLIB2a) removed the large overestimate at deep penetration locations

• Nuclear heating at magnet is higher by ~2% than predicted by FENDL-2.1


Peak Magnet Nuclear Parameters with FENDL-3/SLIB2a

• Nuclear parameters at magnet are higher by ~1.5-3.5% than predicted by FENDL-2.1

• Largest differences are in fast neutron fluence and dpa that are determined by high energy flux

FENDL-3/SLIB2a in Slabs of SS/H2O (80/20)

M. Sawan 34

No large differences observed with corrected library (FENDL-3/SLIB2a)


FENDL-3/SLIB2a in Mo Slab

No large differences observed with corrected library (FENDL-3/SLIB2a)

M. Sawan 352nd RCM on FENDL-3, 23-26 March 2010, IAEA, Vienna


Conclusions

Replacing the ENDF/B-VI.8 data in FENDL-2.1 by ENDF/B-VII.0 while keeping data from other evaluations unchanged resulted in minor change (<1%) in nuclear parameters for ITER relevant calculations

However, larger changes in calculated ICF target neutronics parameters and tritium breeding were observed in previous analysis

Calculations of foil activation and tritium breeding for ITER relevant integral experiments yield nearly similar results for FENDL-2.1 and ENDF/B-VII.0

Using FENDL-3/SLIB2 instead of FENDL-2.1 resulted in large increase in nuclear parameters behind the thick water-cooled shield. This was identified as processing error and was corrected in FENDL-3/SLIB2a

Using FENDL-3/SLIB2a instead of FENDL-2.1 in ITER relevant calculations gives 2-4% higher nuclear parameters in regions that are heavily shielded with water-cooled SS (vacuum vessel, magnets)

Documents

Benchmark Calculations in Support for FENDL-3 Generation