Investigation of Nuclear Data Accuracy for the Accelerator- Driven System with Minor Actinide Fuel

Transcription

1 Investigation of Nuclear Data Accuracy for the Accelerator- Driven System with Minor Actinide Fuel Kenji Nishihara, Takanori Sugawara, Hiroki Iwamoto JAEA, Japan Francisco Alvarez Velarde CIEMAT, Spain Andrei Rineiski KIT, Germany 11th IEM, San Francisco 1-5, Nov,

2 Contents IAEA CRP benchmark activity o Depletion analysis for 800 MWt ADS (commercial grade) o Result by several nuclear libraries Uncertainty analysis using covariance matrix o Estimation of uncertainty by JENDL-3.3 and JENDL-4.0 o Reduction of uncertainty by the Transmutation Physics Experimental Facility, TEF-P. 11th IEM, San Francisco 1-5, Nov,

3 IAEA-CRP benchmark Heat exchanger Proton beam Pb-Bi Pump LBE cooled tank type reactor 800 MW (thermal power) 600 day/cycle Proton beam: 1.5GeV, 10-20mA LBE target with window Nitride fuel, (Pu+MA)N+ZrN o Pu-N/MA-N/Zr-N=19/31/50 (weight percent) o Total actinide amounts 4.2 ton Transmutation: 500kg/cycle (12%/cycle) MA fuel Beam window A. Stanculescu, The IAEA Coordinated Research Project (CRP) on Analytical and experimental benchmark analyses of accelerator driven systems, OECD/NEA 11 th Information Exchange Meeting, San Francisco, 1-4 Nov (2010) 11th IEM, San Francisco 1-5, Nov,

4 Feature of this benchmark Commercial grade ADS o High energy proton (1.5GeV) o MA-rich fuel o Burn-up up to 120 GWd/HMt Simplified R-Z geometry Fundamental knowledge of calculation accuracy at the present time for the commercial grade ADS Main issues for the calculation accuracy are, o Simulation of high energy reactions o Nuclear data for MA 11th IEM, San Francisco 1-5, Nov,

5 LBE target LBE buffer Inner core Outer core LBE reflecctor SS reflector B 4 C shielding Beam duct (void) Nuber density of actinide (10 24 /cm 3 ) (cm) Calculation model SS reflector Gas plenum SS reflector (cm) R-Z model 3.0E-3 2.5E-3 2.0E-3 1.5E-3 1.0E-3 5.0E-4 0.0E+0 Inner core Outer core Fuel composition K.Tsujimoto, et.al, J. Nucl. Sci. Tech., 41(1), 21 (2004). Cm-246 Cm-245 Cm-244 Cm-243 Am-243 Am-242m Am-241 Pu-242 Pu-241 Pu-240 Pu-239 Pu-238 Np-237 U-236 U th IEM, San Francisco 1-5, Nov,

6 Participants Participant Code Library JAEA (Japan) CIEMAT (Spain) KIT (Germany) PHITS, NMTC or MCNPX (MC code for proton and neutron >20MeV) SLAROM (Cross section code) TWODANT (Deterministic neutron transport code) ORIGEN (Burn-up code) EVOLCODE2 (MCNPX-based burnup code) High energy particles are not analyzed. C4P, ZMIX (Cross section code) DANTSYS (Deterministic neutron tansport code) TRAIN (Burn-up code) JENDL-3.3 JENDL-4.0 JENDL-3.2 ENDF/B-VI JEFF-3.0 JEFF-3.0 (JEFF-3.1) a) (ENDF/B-VI) a) ENDF/B-VII JEFF-3.1 a) Library in parenthesis is only for the beginning of cycle (BOC). Deterministic vs. Monte-Carlo JENDL vs. ENDF vs. JEFF PHITS vs. MCNPX for high energy 11th IEM, San Francisco 1-5, Nov,

7 k-eff, Endf/B-VI k-eff, JEFF3.0 k-eff, JEFF 3.1 Deterministic vs. Monte-Carlo Monte Carlo 0.4%dk Monte Carlo %dk Determin istic %dk Monte Carlo Deterministic Determin istic BOC JAEA (Det.) EB6 CIEMAT (MC) EB6 EOC JAEA (Det.) JEFF3.0 CIEMAT (MC) JEFF3.0 BOC EOC KIT (Det.) JEFF3.1 CIEMAT (MC) JEFF3.1 BOC EOC ENDF/B-VI JEFF 3.0 JEFF 3.1 Good agreement is observed ( %dk). Influence of calculation methods is small. 11th IEM, San Francisco 1-5, Nov,

8 k-eff, JENDL k-eff, ENDF k-eff, JEFF JENDL vs. ENDF vs. JEFF JAEA J3.3 JAEA J3.2 JAEA J4.0 J EB6 JAEA EB6 CIEMAT EB6 KIT EB JF JAEA JEFF3.0 CIEMAT JEFF3.0 CIEMAT JEFF3.1 KIT JEFF EB JF J3.2, J BOC EOC 0.93 BOC EOC 0.93 BOC EOC JENDL ENDF JEFF k-effective disperse from 0.98 to 1.0 at BOC and 0.93 to 0.96 at EOC. Version-up changes k-effective by 2 to 3 %dk for each library. 11th IEM, San Francisco 1-5, Nov,

9 Δkeff due to library exchange (pcm) All Np-237 Pu-238 Pu-239 Pu-240 Pu-241 Pu-242 Am-241 Am-243 Cm-244 N-15 Pb-204 Pb-206 Pb-207 Pb-208 Bi-209 Δkeff due to library exchange (pcm) Cr-50 Cr-52 Cr-53 Cr-54 Mn-55 Fe-54 Fe-56 Fe-57 Fe-58 Ni-58 Ni-60 Ni-61 Ni-62 Ni-64 Mo-92 Mo-94 Mo-95 Mo-96 Mo-97 Mo-98 Mo-100 Zr-90 Zr-91 Zr-92 Zr-94 Zr-96 Which nuclide is responsible at BOC? J4 to EB7 J4 to JF J4 to EB7 J4 to JF Statistics error=20pcm Statistics error=20pcm Δkeff due to library exchange of single nuclide from JENDL-4 to ENDF/B-VII or JEFF-3.1 estimated by MCNPX. Large changes are observed at Np-237, Pu-240, Am-241, Am-243, N-15 and Fe-56. Sensitivity analysis is needed in the future. 11th IEM, San Francisco 1-5, Nov,

10 Effect of isomer-state branching ratio (IR) of Depletion of keff, %dk Am-241 capture reaction 6% JAEA J3.3 JAEA EB6 JAEA JEFF3.0 KIT JEFF3.1 JAEA J4 KIT EB7 CIEMAT JEFF3.0 5% 4% 3% 2% 1% ENDF JENDL JEFF 0% 75% 80% 85% 90% 95% IR, isomer-state branching ratio, Am242 / (Am242+Am242m) The depletion of keff becomes larger by 0.8%dk if the IR changes from 80% to 90%. JEFF-3.1 library gives smaller depletions than JENDL-4 and ENDF/B- VII. 11th IEM, San Francisco 1-5, Nov,

11 Proton current (ma) Proton beam current JENDL-3.2, 3.3 JEFF3.0 J3.3 J3.2 J4.0 EB6 JEFF3.0 5 JENDL-4 ENDF/B-VI 0 BOC EOC Criticality affects too much on proton beam current to make comparison. 11th IEM, San Francisco 1-5, Nov,

12 Effect of high-energy particle calculation on current PHITS, NMTC or MCNPX (LA, model) Storage of neutron generated <20MeV TWODANT with JENDL-3.3 Comparison between high-energy codes was done with common low-energy code and library. 11th IEM, San Francisco 1-5, Nov,

13 Spallation neutrons per proton Spallation neutron per proton Neutron produced at energy below MCNPX MCNPX PHITS NMTC (model) (LA150) N/P MeV Number of neutron and average energy (MeV) 20 MeV Energy (MeV) Spectrum of spallation neutron Number of neutron: PHITS > MCNPX Average energy of neutron: PHITS > MCNPX PHITS results in 8% larger efficiency of source proton. PHITS NMTC MCNPX (model) MCNPX (LA150) 11th IEM, San Francisco 1-5, Nov,

14 Proton current (ma) Result of beam current MCNPX PHITS, NMTC PHITS NMTC MCNPX (model) 12 MCNPX (LA150) 10 0 day 600 day Proton beam current based on JENDL-3.3 The difference of current is 8% between MCNPX and others. 11th IEM, San Francisco 1-5, Nov,

15 Summary of IAEA-CRP benchmark Small difference between participants (codes) with same library. k-effective disperse from 0.98 to 1.0 at BOC and 0.93 to 0.96 at EOC. o At BOC, Np-237, Pu-240, Am-241, Am-243, N-15 and Fe-56 affect. Beam current is not comparable due to different k- effective. o Based on JENDL-3.3, current differs by 8% between MCNPX and PHITS. To design an ADS is very difficult, if estimated k- effective changes 2 to 3 %dk. Uncertainty and its reduction by ctritical experiments should be evaluated. 11th IEM, San Francisco 1-5, Nov,

16 Uncertainty evaluation The error caused by the nuclear data is calculated as square root of GMG t. Sensitivity coefficient G o SAGEP code gives sensitivity of cross sections on: criticality, coolant void reactivity and Doppler coefficient Covariance data M o Covariance data are presented in JENDL-3.3 and JENDL o Provisional values were employed for nuclides and reactions which were not prepared in JENDL. 11th IEM, San Francisco 1-5, Nov,

17 Covariance data in JENDL-3.3 Nuclide Capture Fission ν Elastic Inelastic χ μ-bar Nuclide Capture Fission ν Elastic Inelastic χ μ-bar Pu-238 N Pu-239 Fe Pu-240 Cr Pu-241 Ni Pu-242 Zr Np-237 Zr Am-241 Zr Am-242m Zr Am-243 Zr Cm-242 Pb Cm-243 Pb Cm-244 Pb Cm-245 Pb Cm-246 Bi Most of reactions for MA, Zr and Pb are missing in the JENDL-3.3. They are supplemented with provisional estimations. : Formally included :Provisional estimation, Nakagawa (2007) :Provisional estimation, Shibata (2007) 11th IEM, San Francisco 1-5, Nov,

18 Covariance data in JENDL-4.0 Nuclide Capture Fission ν Elastic Inelastic χ μ-bar Nuclide Capture Fission ν Elastic Inelastic χ μ-bar Pu-238 N Pu-239 Fe Pu-240 Cr Pu-241 Ni Pu-242 Zr Np-237 Zr Am-241 Zr Am-242m Zr Am-243 Zr Cm-242 Pb Cm-243 Pb Cm-244 Pb Cm-245 Pb Cm-246 Bi All of the actinide reactions are covered in the JENDL-4.0! But, data for N-15, Zr and Pb are needed. (Provisional data in the JENDL-3.3 are used in the present study.) 11th IEM, San Francisco 1-5, Nov, Lack

19 Uncertainty of JENDL-3.3 and 4.0 JENDL-3.3 JENDL-4.0 Value (GMG t ) 0.5 Value (GMG t ) 0.5 Criticality (k-eff) pcm (1.4%) pcm (1.1%) Coolant void reactivity [Δk/k] 5300 pcm 350 pcm (6.6%) 3500 pcm 280 pcm (8.1%) Doppler reactivity coefficient [TΔk/dT] -3.5e-4 7.0% -3.3e-4 6.3% Uncertainties evaluated by JENDL-4.0 are equal or less than those by JENDL-3.3. The uncertainty of criticality, 1090 pcm, is smaller than the result of IAEA benchmark activity ( pcm). 11th IEM, San Francisco 1-5, Nov,

20 MU Uncertainty in criticality (pcm) NU NU NU CHI NU NU NU NU NU Uncertainty in criticality (pcm) Breakdown of uncertainty (criticality) J-3.3 J-4.0 Total uncertainty= 1320(J-3.3), 1090 (J-4.0) Np-237 capture Pu-239 fission and chi Am-241 capture Am-243 capture Np237 Pu238 Pu239 Pu240 Pu241 Pu242 Am241 Am243 Cm244Cm J-3.3 J-4.0 Total uncertainty= 1320(J-3.3), 1090 (J-4.0) N-15 elastic Fe-56 inelastic Bi-209 elastic 0 11th IEM, San Francisco 1-5, Nov, N15 Fe56 Ni58 Zr90 Zr91 Zr92 Zr94Pb204 Pb206 Pb207 Pb208 Bi209

21 MU Uncertainty in Doppler coeficient (%) Uncertainty in criticality (pcm) Breakdown of uncertainty (void reactivity) J-3.3 J-4.0 Void reactivity = 5300pcm(J-3.3), 3500pcm(J-4.0) Total uncertainty= 350pcm(J-3.3), 280pcm (J-4.0) NU CHI NU NU NU Np237 Pu239 Pu240 Am241 Am243 Cm244 Np-237 capture Pu-239 chi Am-241 capture J-3.3 J-4.0 Void reactivity = 5300pcm(J-3.3), 3500pcm(J-4.0) Total uncertainty= 350pcm(J-3.3), 280pcm (J-4.0) N-15 elastic Fe-56 inelastic, mu Ni-58 inelastic Bi-209 elastic and inelastic N15Cr52 Fe56 Ni58 Zr91 Zr92 Zr94 Zr96 Pb204Pb206 Pb207Pb208 Pb208 Bi209 11th IEM, San Francisco 1-5, Nov,

22 Uncertainty in Doppler coeficient (%) NU NU CHI NU NU Uncertainty in Doppler coeficient (%) Breakdown of uncertainty (Doppler) J-3.3 J-4.0 Total uncertainty= 7.0%(J-3.3), 6.3% (J4.0) Np-237 capture Pu-239 chi Am-241 capture Np237 Pu238 Pu239 Pu240 Pu241 Am241 Am243 Cm J-3.3 J-4.0 Total uncertainty= 7.0%(J-3.3), 6.3% (J4.0) N15 Fe56 Zr90 Zr91 Zr92 Zr94 Pb206 Pb207 Pb208 Bi209 Zr-90 capture Zr-92 capture Zr-94 capture Pb-206 capture Bi-209 capture 11th IEM, San Francisco 1-5, Nov,

23 Reduction of uncertainty by TEF-P Movable shield Fixed half assembly Fuel loading machine Stainless steel matrix Coolant simulator (Pb, Na, etc.) Core Plate-type fuel Spallation target Beam duct Proton beam Fuel drawer Proton beam Pin-type fuel Safety/control rod drive mechanism Movable half assembly 5.5cm 5.5cm Three calculation cases were performed to confirm effect of MA amount on the uncertainty reduction : CASE-mg: Measurement of the reaction rate ratio by using mg-order MA sample CASE-g: Measurement of the MA sample worth by using g-order MA sample CASE-kg: Various experiments by using kg-order MA to simulate spectrum dependences. 11th IEM, San Francisco 1-5, Nov,

24 Experiments in CASE-kg Experiment descriptions of case-kg o Criticality: 1 case o Coolant void reactivity:3 cases(3 different void fractions) o Doppler reactivity:3 cases (300, 550, 800 ) o Measurement of reaction rate ratio:8 cases (Denominator: Pu-239 fission, Numerator: Fission and Capture for 4 MA nuclides, Np-237, Am-241, Am-243 and Cm-244) o Total 15 cases 11th IEM, San Francisco 1-5, Nov,

25 J33 R233 CASE-mg CASE-g + Criticality + Void +Doppler +RR Error caused by nuclear data [%] Calculation Results (1/2) 1.4 ADS: Criticality J33: Result by using JENDL % CASE-kg R233: Adjusted by existing 233 integral data *1 ( It is able to consider R233 result as the current best value) +Criticality : Adjusted by CASE-g results and criticality experiment in CASE-kg. +RR : Adjusted by +Doppler result and measurement of reaction rate ratio in CASE-kg. Case The error was reduced significantly in CASE-g although the result in CASE-mg was hardly reduced. Experiments in CASE-kg was also effective to improve the error. The value was about 35% smaller than that in R th IEM, San Francisco *1: T. Hazama, et al., JNC TN , 1-5, (2002) Nov,

26 J33 R233 CASE-mg CASE-g + Criticality + Void +Doppler +RR J33 R233 CASE-mg CASE-g + Criticality + Void +Doppler +RR Error caused by nuclear data [%] Error caused by nuclear data [%] Calculation Results (2/2) 7 ADS: Void reactivity 7 ADS: Dopper reactivity % 45% CASE-kg 1 0 CASE-kg Case For both coolant void and Doppler reactivity, the errors were hardly improved in CASE-mg and CASE-g. The measurement of coolant void reactivity or Doppler reactivity with kg-order MA was effective for the error reduction. Case 11th IEM, San Francisco 1-5, Nov,

27 Summary for uncertainty analysis To estimate the uncertainty caused by the nuclear data o The uncertainty was 1090 pcm for the criticality, 280 pcm for the void reactivity and 6.3% for the Doppler reactivity based on the JENDL-4.0 covariance data. o However, the uncertainty of criticality, 1090 pcm, looks smaller than the result of IAEA benchmark activity ( pcm). To estimate the impact of the TEF-P o The uncertainty would be reduced up to 45% by the TEF-P. Thank you for your attention! 11th IEM, San Francisco 1-5, Nov,