MOx Benchmark Calculations by Deterministic and Monte Carlo Codes

MOx Benchmark Calculations by Deterministic and Monte Carlo Codes G.Kotev, M. Pecchia C. Parisi, F. D Auria San Piero a Grado Nuclear Research Group (GRNSPG), University of Pisa via Diotisalvi 2, 56122 Pisa g.kotev@ing.unipi.it, m.pecchia@ing.unipi.it, c.parisi@ing.unipi.it, dauria@ing.unipi.it ABSTRACT A depletion calculation benchmark devoted to MOx fuel is an ongoing objective of the OECD/NEA WPRS following the study of depletion calculation concerning UOx fuels. The objective of the proposed benchmark is to compare existing depletion calculations obtained with various codes and data libraries applied to fuel and back-end cycle configurations. In the present work the deterministic code NEWT/ORIGEN-S of the SCALE6 codes package was used to calculate the masses of inventory isotopes. Then a methodology to apply the Monte Carlo based code MONTEBURNS2.0 to this benchmark is presented. 1 INTRODUCTION A depletion calculation benchmark devoted to MOx fuel is an ongoing objective of the OECD/NEA Working Party on Reactor Systems (WPRS), following the study of depletion calculation concerning UOx fuels [1]. The benchmark is developed and coordinated by the French CEA. The objective is to compare existing depletion calculations obtained with various codes and data libraries applied to fuel and back-end cycle configurations. Calculated quantities are the nuclide densities of the most important nuclides implied in the fuel cycle. The specification reported here is devoted to the second phase of the Benchmark and focuses on MOx fuels, more precisely on the typical plutonium vector for material derived from the reprocessing of thermal reactor UO 2 fuels. In this work the final mass inventory calculated by the deterministic code NEWT/ORIGEN-S of the SCALE6.0 codes package is presented. Then a methodology to apply the Monte Carlo based burnup code MONTEBURNS2.0 to this benchmark is also proposed. 2 BENCHMARK SPECIFICATION The benchmark model is made of 4 fuel assemblies with a MOx composition that is representative of realistic MOx fuels irradiated in a mixed UO2-MOx PWR core and 3 adjacent UO2 fuel assemblies with an initial enrichment of 3.25%, already burned to 24 GWd/tHM. The requested calculations were performed to attain a constant target burnup of 38 GWd/tHM for two selected MOx fuel pin (Q17 and L14, referring to Fig. 1). Results reported in this paper refer to the Q17 fuel pin calculations. Several fuel impurities are added to MOx fuel composition in order to calculate the activation products masses coming from such impurities. All the data regarding material initial compositions, geometrical data and material temperatures are reported in [1]. Fuel 209.1

209.2 irradiation history is reported in Table 1. A sketch of the benchmark model with the position of the reference fuel pins is reported in Fig 1. Fig 1: Sketch for benchmark model. The mass inventory presented here is calculated for the Q17 MOx fuel pin Table 1: Power history Q17 MOx fuel pin Operating cycles Time Units EOC burnup [GWd/tHM] Cycle 1 285 [d] 12 Downtime 60 [d] Cycle 2 300 [d] 25 Downtime 40 [d] Cycle 3 280 [d] 38 Cooling 5, 50, 100, 300 [y] 3 THE NEWT/ORIGEN-S MODEL NEWT is a multigroup discrete-ordinates radiation transport code with flexible meshing capabilities that allow two-dimensional neutron transport calculations using complex geometric models [2]. Automated grid generation capabilities provide a simplified user input specification in which elementary bodies can be defined and placed within a problem domain. The Cross Section Libraries used by NEWT are a pre-definend 238 energy group cross section library, prepared at ORNL by the AMPX code and based on the ENDF-B/VII data. Self-shielding calculations by the CENTRM module are performed using a point-wise cross section library, also processed by the AMPX code and based on the ENDF-B/VII data. The MCDANCOFF module was used to calculate Dancoff factors in complicated geometries using Monte Carlo integrations. The Dancoff factors are used in SCALE to correctly self-shield multigroup cross sections for a given problem, determining an equivalent cell for CENTRM. The benchmark model is reported in Fig. 2.

209.3 Fig. 2: The NEWT/ORIGEN-S benchmark model 4 RESULTS BY NEWT/ORIGEN-S CALCULATION Results from NEWT/ORIGEN-S calculations are reported in Tables 2-5. They are divided in actinides, fission products, activation products and fission/activation products, respectively. Only actinides and fission products with an higher concentration are reported. Table 2: Masses of actinides Nuclide Discharge 5 years 50 years 100 years 300 years U234 1.873E+01 4.458E+01 2.406E+02 3.906E+02 6.376E+02 U235 7.085E+02 7.099E+02 7.221E+02 7.356E+02 7.898E+02 U236 3.050E+02 3.084E+02 3.400E+02 3.768E+02 5.186E+02 Np237 1.686E+02 1.774E+02 3.844E+02 6.860E+02 1.713E+03 Pu238 5.856E+02 6.658E+02 4.673E+02 3.154E+02 6.567E+01 Pu239 9.489E+03 9.614E+03 9.608E+03 9.599E+03 9.558E+03 Pu240 6.498E+03 6.577E+03 6.868E+03 6.892E+03 6.757E+03 Pu242 2.499E+03 2.499E+03 2.499E+03 2.499E+03 2.498E+03 Am241 2.104E+02 1.091E+03 3.751E+03 3.780E+03 2.769E+03 Am242m 5.364E+00 5.234E+00 4.195E+00 3.282E+00 1.228E+00 Am243 7.884E+02 7.886E+02 7.852E+02 7.815E+02 7.670E+02 Cm245 5.071E+01 5.069E+01 5.051E+01 5.030E+01 4.948E+01 Cm246 6.674E+00 6.669E+00 6.625E+00 6.576E+00 6.387E+00 Cm245 5.071E+01 5.069E+01 5.051E+01 5.030E+01 4.948E+01 Table 3: Masses of fission products Nuclide Discharge 5 years 50 years 100 years 300 years Rb87 1.482E+02 1.482E+02 1.482E+02 1.482E+02 1.482E+02 Sr88 1.978E+02 1.978E+02 1.978E+02 1.978E+02 1.978E+02 Tc99 8.678E+02 8.727E+02 8.726E+02 8.724E+02 8.719E+02 Ru101 9.381E+02 9.381E+02 9.381E+02 9.381E+02 9.381E+02 Rh103 7.014E+02 7.785E+02 7.785E+02 7.785E+02 7.785E+02 Pd107 5.623E+02 5.623E+02 5.623E+02 5.623E+02 5.622E+02

209.4 Ag109 1.907E+02 1.910E+02 1.910E+02 1.910E+02 1.910E+02 Xe131 5.067E+02 5.177E+02 5.177E+02 5.177E+02 5.177E+02 Xe132 1.351E+03 1.357E+03 1.357E+03 1.357E+03 1.357E+03 Xe134 1.685E+03 1.685E+03 1.685E+03 1.685E+03 1.685E+03 Xe136 2.666E+03 2.666E+03 2.666E+03 2.666E+03 2.666E+03 Cs133 1.227E+03 1.243E+03 1.243E+03 1.243E+03 1.243E+03 Cs135 5.038E+02 5.048E+02 5.048E+02 5.047E+02 5.047E+02 Ce140 1.233E+03 1.265E+03 1.265E+03 1.265E+03 1.265E+03 Nd143 7.879E+02 8.149E+02 8.149E+02 8.149E+02 8.149E+02 Nd144 7.809E+02 1.149E+03 1.153E+03 1.153E+03 1.153E+03 Nd145 6.465E+02 6.467E+02 6.467E+02 6.467E+02 6.467E+02 Nd146 7.172E+02 7.172E+02 7.172E+02 7.172E+02 7.172E+02 Nd148 4.227E+02 4.227E+02 4.227E+02 4.227E+02 4.227E+02 Nd150 2.496E+02 2.496E+02 2.496E+02 2.496E+02 2.496E+02 Sm147 6.122E+01 2.035E+02 2.552E+02 2.552E+02 2.552E+02 Sm150 3.563E+02 3.563E+02 3.563E+02 3.563E+02 3.563E+02 Sm152 1.358E+02 1.358E+02 1.359E+02 1.359E+02 1.359E+02 Eu153 1.796E+02 1.812E+02 1.812E+02 1.812E+02 1.812E+02 Table 4: Masses of activation products Nuclide Discharge 5 years 50 years 100 years 300 years Cl36 1.565E+00 1.565E+00 1.565E+00 1.565E+00 1.564E+00 Ca41 2.642E-01 2.642E-01 2.641E-01 2.640E-01 2.636E-01 Mn53* N/A N/A N/A N/A N/A Mn54 1.543E-02 2.676E-04 3.774E-20 0.000E+00 0.000E+00 Fe55 1.276E-01 3.583E-02 3.908E-07 1.198E-12 0.000E+00 Fe60* N/A N/A N/A N/A N/A Co60 1.605E+00 8.315E-01 2.235E-03 3.112E-06 1.170E-17 Ni59 1.737E+00 1.737E+00 1.736E+00 1.736E+00 1.732E+00 Ni63 3.378E-01 3.263E-01 2.389E-01 1.691E-01 4.231E-02 Mo93 0.000E+00 0.000E+00 0.000E+00 0.000E+00 0.000E+00 * Not available Table 5: Masses of fission and activation products Nuclide Discharge 5 years 50 years 100 years 300 years H3 FP (fission) 7.441E-02 5.618E-02 4.476E-03 2.693E-04 3.524E-09 AP(activation) 8.361E-07 6.312E-07 5.030E-08 3.026E-09 3.961E-14 Be10 FP 1.137E-02 1.137E-02 1.137E-02 1.137E-02 1.137E-02 AP 1.061E-04 1.061E-04 1.061E-04 1.061E-04 1.061E-04 C14 FP 4.836E-02 4.833E-02 4.807E-02 4.778E-02 4.664E-02 AP 9.084E-02 9.079E-02 9.029E-02 8.974E-02 8.760E-02 Zr93 FP 5.612E+02 5.615E+02 5.615E+02 5.615E+02 5.615E+02 Nb94 FP 1.685E+00 1.685E+00 1.683E+00 1.680E+00 1.668E+00 Sn119m FP 1.699E+01 1.714E+01 1.714E+01 1.714E+01 1.714E+01 Sn121m FP 1.168E-02 1.770E-05 1.004E-05 5.346E-06 4.299E-07 Sn126* FP 3.168E+01 3.168E+01 3.167E+01 3.165E+01 3.161E+01 AP N/A Sb125 FP 1.307E+01 3.710E+00 4.049E-05 1.241E-10 0.000E+00 * AP not available

209.5 5 METHODOLOGY TO MONTEBURNS2.0 CODE It is planned to apply also the Monte Carlo based code MONTEBURNS2.0 to this benchmark and to compare its results with the NEWT/ORIGEN-S results. In this chapter the methodology developed to update MONTEBURNS2.0 code in order to fulfill benchmark specification is presented. 5.1 The MONTEBURNS2.0 code MONTEBURNS2.0 is a fully automated tool that links the Monte Carlo transport code MCNP5 with the radioactive decay and burnup code ORIGEN2.2 [3]. The program processes input from the user that specifies the system geometry, initial material compositions, and other code-specific parameters. MONTEBURNS2.0 transfers the one-group cross section (Xsec) and flux values from MCNP5 output to ORIGEN2.2 input, then the resulting material compositions (after irradiation and/or decay) are collected from ORIGEN2.2 output and send back to MCNP5 input in a repeated, cyclic fashion. The ORIGEN2.2 Xsec are updated for each material composition listed in the MONTEBURNS2.0 input deck. 5.2 The Flux Normalization A computer patch was developed to allow MONTEBURNS2.0 to normalize the MCNP5 fluxes using the given MOx fuel specific pin power. Such patch was necessary because MONTEBURNS2.0 generally uses the total power of the simulated system to normalize the fluxes calculated by MCNP5. The developed patch was verified against another Monte Carlo based depletion code, the KENO-VI/ORIGEN-S codes of the SCALE6.0 codes package [4]. KENO-VI/ORIGEN-S codes allow to define a specific power in a selected material region, thus re-normalizing all the calculated fluxes consistently with such specific power. In this verification test a simple system made of 2 fuel regions was modeled for both codes (see Fig 2), then a specific power was assigned to the inner region and several burnup steps were run (reflective boundary were selected). Results from both codes about final mass inventory and specific powers were compared. The resulting deviation in actinides masses was below 5% (see Fig. 4). The calculated fractional power matched exactly in the initial step, while in the final step the deviation was below 1.5%. Fig 2: The verification model. Different fuel materials are the yellow and red regions

209.6 5.3 The MCNP5 Model Fig 4: Relative actinides mass differences in the test case The MCNP5 developed model for the benchmark is equivalent to the NEWT model reported in the chapter 3. Both input decks were set up using the same input data, i.e material composition, densities and geometrical data. The MCNP5 card KCODE is used, setting 30000 particles per cycle for 1060 cycles, thus resulting in a total of 3 x 10 7 histories simulated. The first 60 cycles are discarded in order to let the source distribution to converge. The initial neutron source distribution is setup using the KSRC card. Reflective boundary conditions are selected, consistently with the NEWT/ORIGEN-S calculation. The nuclear data library used is the continuous energy ENDF/B-VII nuclear data file, processed at required temperatures by the NJOY99 code [6]. A computer routine, developed in mixed languages C/PERL, was used to fully automatize the Xsec libraries processing by NJOY. This routine enables also consistency/error checks and at the end of the procedure a simple test case is run using MCNP5 to check if the new Xsec libraries work properly. 5.4 Minor Fission Product Lumping Irradiation of nuclear fuel produces an accumulation of fission products (FP). The most relevant isotopes were directly considered in the simulations but the cumulative effect of the other FP could not be neglected. To take into account the effects of such FP, a procedure to lump these isotopes together in a pseudo fission product Xsec (p-xsec) was developed, since MONTEBURNS2.0 does not have such capability. A special MCNP5 Xsec library including the contribution of all FP not directly taken into account was set up. The nuclear data are based on the ENDF/B-VII continuous energy library. The formula used to calculate an averaged cross section is the following, i σ ( E ) = n σ ( E) (1) Pseudo i { Nuclide} i

209.7 where σ Pseudo (E) is the p-xsec of fission products, σ i (E) the nuclide Xsec and n i the isotope atom fraction. The processing is executed for continuous energies Xsec. The p-xsec is calculated for each MT file that contributes to total Xsec (MT = 1 in ENDF6 terminology [7]). This datum is taken from ENDF/B-VII library. The atom fractions are calculated from ORIGEN2.2 output collected by MONTEBURNS2.0. A computer routine written in mixed programming language C/PERL was developed to automatize the process of creation and for checking the σ Pseudo (E). The NJOY99 code [6] was also used within such procedure, to consider temperature effects in the initial step and to convert the library to the ACE format used by MCNP5 in the last step. An example of produced p-xsec is reported in Fig. 6. Fig 6: Principal cross section behaviour for p-xsec at 20 GWd/tHM The p-xsec were calculated and included into the depletion process after the first MONTEBURNS run. 6 CONCLUSION In this work the results of the application of the deterministic code NEWT/ORIGEN-S to a CEA - OECD/NEA MOx fuel benchmark are presented. A methodology to update and apply the MONTEBURNS2.0 code to this benchmark is also proposed. This methodology will allow in future works a comparison of the deterministic NEWT/ORIGEN-S calculations with the MONTEBURNS Monte Carlo based calculations. REFERENCES [1] B. Roque, P. Marimbeau, J.P.Grouiller, L. San-Felice, Specification for the Phase 2 of a Depletion Calculation Benchmark devoted to MOx Fuel Cycles, NEA/NSC/DOC(2007)9 [2] M. D. DeHart, NEWT: A new transport algorithm for two-dimensional discrete ordinates analysis in non-orthogonal geometries, ORNL/TM-2005/39, Version 6, Vol. II, Sect. F21, January 2009

209.8 [3] D. I. Poston, H. R. Trellue, User s Manual, Version 2.0 for MONTEBURNS, version 1.0, LA-UR-99-4999, September, 1999, Los Alamos National Laboratory, USA [4] D. F. Hollenbach, L. M. Petrie, S. Goluoglu, N. F. Landers, M. E. Dunn, KENO-VI: A General Quadratic Version of the KENO Program, ORNL/TM-2005/39, Version 6, Vol. II, Sect. F17, January 2009 [5] M. D. DeHart, TRITON: a two-dimensional transport and depletion module for characterization of spent nuclear fuel, ORNL/TM-2005/39, Version 6, Vol. I, Sect. T1, January 2009 [6] R. E. MacFarlane, D. W. Muir, The NJOY Nuclear Data Processing System Version 91, LA-12740-M UC-413, October, 1994, Los Alamos National Laboratory, USA [7] National Nuclear Data Center, ENDF-6 Formats Manual, Report BNL-NCS-44945-05- Rev., April, 2001, Brookhaven National Laboratory, USA