Nuclear data impact on the core neutron design

NUCLEAR – 2009 ████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████ NUCLEAR DATA IMPACT ON THE CORE NEUTRON DESIGN G. GLINATSIS, D. GUGIU*, G. GRASSO** and C. PETROVICH ENEA, Bologna, Italy, [email protected] & [email protected] * INR, Institute for Nuclear Research, Piteşti, Romania, [email protected] ** Nuclear Engineering Laboratory, University of Bologna, Bologna, Italy, giacomo.grasso@ mail.ing.unibo.it ABSTRACT The impact of the different sets of nuclear data on the criticality level and reactivity coefficients of critical and sub-critical LMFR cores is discussed. A sensitivity analysis, by direct substitution of the nuclear data of different origin (JEFF, ENDF/B, JENDL families), was performed. Large discrepancies on the neutron parameters, among the different nuclear data sets, are observed. Such an investigation, in stochastic approach by single isotope or group of isotopes substitution, highlights the isotopes and/or the reaction events that are mainly responsible for the most significant discrepancies. Both radiotoxicity and residual risk reduction require high accuracy of the nuclear data, mainly for the Minor Actinides. Finally, the impact of these uncertainties could be an important issue also for the economic aspects of dedicated cores. Key words: (Core Design, Minor Actinides, Transmutation, Monte Carlo, Nuclear Data) Background The use of the nuclear energy, able to substitute fossil fuels at acceptable costs, is strongly related to two critical aspects of nuclear power production: ∑ potential catastrophic risks, and ∑ the amount of radioactive wastes. The latter solution can be offered by engaging nuclear reactors, preferably of fast neutrons spectrum. The minimization of the potential catastrophic risks requires enhanced safety performances, while the reduction of the high-level radioactive wastes (HLW) imply Partitioning & Transmutation (P&T) techniques for both minor actinides (MA) and long-lived fission products (LLFP). To meet both enhanced safety behaviour and improved environmental impact, the worldwide scientific community efforts are addressed to innovative concepts and materials. The Generation-IV initiative aims mainly to 8████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████ 01 ████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████ NUCLEAR – 2009 ████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████ ∑ generate sustainable energy and promote long-term availability of nuclear fuel; ∑ minimize the nuclear waste and reduce its long term management. These goals are integrated by some elements of the EU energy policy, such as: ∑ Generation-IV systems with closed fuel cycles to substantially minimize the volume, the radiotoxic content and the thermal load of the residual high-level waste requiring geological disposal; ∑ the development of new applications of nuclear energy in Europe, and preliminary roadmaps for nuclear energy related technologies. All these objectives constitute the tentative to provide a global solution of the energy demand in a sustainable manner, meeting also the public acceptability. The Generation-IV innovative nuclear systems present new characteristics and requirements concerning both the reactor cores (in terms of: “exotic” and/or U-free fuels, ceramic materials for cladding, LHM for coolants, core architectures to optimise neutron performances, etc.) and the associated fuel cycles (in terms of: new fuels fabrication, reprocessing, waste minimization, etc.). This introduces the general issue of the accuracy and reliability of the “results”, related to both data uncertainties and applied methods [1], [2]. Nuclear data uncertainties and their impact must be assessed in order to validate the core design studies of some innovative solutions. The observed large discrepancies among the different neutron parameters, mainly in “dedicated” core solutions, are attributed to the nuclear data (and their uncertainties). Boundary conditions of the core configurations Two innovative core solutions have been investigated: the sub-critical ADS core of the EFIT reactor (within the 5th and 6th EU Framework Programs, [3]) and the critical core of the ELSY reactor (within the 6th EU Framework Programs, [4]). The first one is fuelled by (Pu, MA)O2-x + MgO innovative Ufree Cer-Cer type fuel with high content of MA and sized at about 380 MWth, while the second one is fuelled by “classical” MOX of type (U, Pu)O2-x and sized at 1500 MWth. The structural materials ferritic-martensitic steel T91 and the coolant material Pb are common materials for both the subcritical and critical cores. The fuel assembly (FA) pin arrangement of the first one follows the classical hexagonal lattice layout inside an hexagonal wrapper tube, while the wrapperless FA pins of the second one are disposed following a square lattice layout. Both the investigated solutions are optimised to flatten the radial power distribution, with specific design requirements, beyond the energy production, that is: ∑ maximum MA transmutation rate (theoretical value: ~ 42 kg MA/TWthhr) with zero Pu net balance for the equilibrium EFIT core, and ∑ equilibrium cycles loading only natural (or depleted) Uranium and discharging only fission products at (about) constant MA content in the core for the ELSY reactor. From the neutron design point of view, any design constraint must be fulfilled (as far as possible with adequate margin) under any plant foreseeable operating condition pertaining either Design Base Conditions (DBC) or Design Extension Conditions (DEC). To avoid undesirable feedbacks on the criticality, the reactivity level and (mainly) the reactivity coefficients should be known with high accuracy. It is worthwhile to mention that, in case of an erroneous evaluation of the critical mass or of the sub criticality level, a “correction” is possible during the “loading” of the core, while a similar “correction” is not as much feasible for the differential parameters, such as the reactivity coefficients. Since the simulation results are heavily influenced by the nuclear data, their importance appears evident: the nuclear data constitute a crucial point for the “decisions” of every neutron design. Since extensive studies on the uncertainty evaluations and on the systematic of existing and required uncertainties have already been performed in [1], [5], this investigation (which involves both ENDF/B 7.0 and JENDL 3.3 evaluated nuclear data files), deals mainly with their impact on the reactivity parameters, with respect to the reference results obtained using the JEFF 3.1 data. MCNP/MCNPX 8████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████ 01 ████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████ NUCLEAR – 2009 ████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████ codes [6] were used for the neutronics, while the NJOY (release 99, update 259) code [7] system was used to process the evaluated data. Results The results to be discussed come from a sensitivity analysis through the “direct substitution” approach: by substituting the cross section data of one or more isotopes. This approach is the natural consequence of the neutronic investigation tool choice, because the stochastic transport codes do not allow to perform perturbation and uncertainty evaluations in the sense of the classical deterministic transport concept. The impact of the different sets of nuclear data on the multiplication factor, delayed neutron effective fraction, reactivity burn up swing, materials reactivity worth, coolant void effect and on the other neutron parameters was investigated in [8], [9]. The results obtained for some relevant neutron parameters are presented as % variation with respect to the reference configuration results. EFIT Core Results Because of the sub-critical nature of the EFIT core, the following neutron parameters, which are related to the proton beam or the external neutron source (coming from the spallation reactions on the Pb target), are defined as below: ∑ n-source: the spallation neutrons appearing below 20 MeV per impinging proton; ∑ M: fission neutrons born per n-source; ∑ ks: n-source multiplication factor: ks = M/(M + 1); * ∑ j : n-source importance: j *= [(1-keff) / keff] / [(1-ks) / ks]. Table 1 collects the obtained results: Table 1 BoC EFIT core neutron source parameters. JEFF 3.1 n-source M ks ϕ* ENDF/B 7.0 (%) 23.09 − 0.91 13.71 −32.7 0.93201 − 3.2 0.544 − 0.92 JENDL 3.3 (%) / −56.20 − 8.00 − 2.57 Concerning some of the main neutron parameters of the EFIT core configuration, the following results have been carried out: Table 2 collects the average micro cross-sections variation for some fuel isotopes, while Table 3 collects some of the main neutron parameters of the configuration at BoC. Table 2 BoL EFIT core average micro cross sections percentage variation among the ENDF files. 237 Np 239 Pu 241 Pu sf sc n tot sf sc n tot sf sc n tot ENDF/B 7.0 (%) -0.382 4.734 1.389 0.294 -1.760 0.035 -0.844 -9.237 -0.320 JENDL 3.3 (%) -0.484 241 Am 4.748 1.387 0.287 243 Am -1.870 0.034 -0.877 244 Cm -9.313 -0.321 sf sc n tot sf sc n tot sf sc n tot ENDF/B 7.0 (%) -0.981 2.441 -3.466 -0.907 -6.378 -0.930 -0.184 -16.172 -0.215 JENDL 3.3 (%) -1.128 2.435 -3.465 -1.063 -6.420 -0.931 -0.285 -16.133 -0.217 8████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████ 01 ████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████ NUCLEAR – 2009 ████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████ The discrepancies of the data from the Tables 1, 2 and 3 are really relevant for both cases. Table 3 BoC EFIT core integral and differential neutron parameters variation among the ENDF files. ENDF/B 7.0 JENDL 3.3 (%) (%) k eff − 1.765 − 4.458 < En > − 0.635 − 5.853 <n> − 0.911 − 1.336 11.76 beff − 2.703 (*) BUswing −0.00294± 37E-5 −0.00087± 37E-5 - 0.724 - 2.985 Fission events 1.485 4.834 Capture events: 2.091 -13.62 (n, xn) events: W = (∆k/k) /(∆N/N) ; AC = Active Core ENDF/B 7.0 (%) W_fuel Pucontent W_fuel vol_fract W_clad W_cool AC (∆k/k) AC Void 0.634 − 1.822 − 2.632 − 13.79 − 7.801 0.049 / −2.798 2.067 / 1.543 Pu/MA Fis evs. Pu/MA Cpt evs. Pu/MA n,xn evs − 15.57 / 30.08 JENDL 3.3 (%) 2.325 − 6.275 − 13.16 − 6.897 29.33 − 1.721 / − 6.366 7.366 / 4.467 − 25.22 / 5.624 Both integral and differential parameters are afflicted by significant discrepancies and Fig.1 confirms the neutron average energy variation among the different ENDF. To extend the panorama concerning the nuclear data impact on the reactivity parameters, several other simulations have been performed by substituting the whole material cross sections. By substituting the whole fuel isotope cross sections (including MgO matrix), with the corresponding ENDF/B 7.0 and JENDL 3.3 ones, the following results have been obtained for the BoL core configuration: ∑ whole fuel cross sections substitution: ENDF/B 7.0: (∆k/k)eff = − 0.01196 ± 0.00037 JENDL 3.3: (∆k/k)eff = − 0.01965 ± 0.00036. ∑ Pb isotopes cross sections substitution: ENDF/B 7.0: (∆k/k)eff = − 0.01065 ± 0.00038 JENDL 3.3: (∆k/k)eff = − 0.02616 ± 0.00036. More significant information about the different neutron parameter discrepancies, using various ENDF, can be obtained from the following results: Fig. 1. BoC neutron flux spectra in the fuel zones. ∑ .∑ 239 Pu cross sections substitution: 241 Am cross sections substitution: ENDF/B 7.0: JENDL 3.3: ENDF/B 7.0: JENDL 3.3: (∆k/k)eff = + 0.00068 ± 0.00034 (∆k/k)eff = + 0.00160 ± 0.00035; (∆k/k)eff = − 0.01169 ± 0.00037 (∆k/k)eff = − 0.01570 ± 0.00038. These results confirm previous evaluation results using cross sections from ENDF/B 6 and from JEF 2.2 evaluated nuclear data [10]. While the 239Pu contribution remains always very limited, in agreement with the current knowledge on the 239Pu data, strong discrepancies are observed in the case of the substitution of the 241Am and of Pb data. Of course, because of the U-free nature of the CERCER fuel of EFIT, any information concerning U-isotopes micro cross sections or Doppler coefficient evaluations is missing. 8████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████ 01 ████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████ NUCLEAR – 2009 ████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████ ELSY Core Results Although the neutron design activities are still in progress, some evaluations concerning the ENDF impact are performed. An extensive investigation of the flux-weighted average micro cross sections is included in Table 4 (limited to JEFF 3.1 and ENDF/B 7.0 evaluated data), while for the reactivity level Table 4 ELSY core fuel average micro cross sections and their % variation. U235 U238 Np237 Np238 Pu238 Pu239 Pu240 Pu241 Pu242 Am241 Am243 Cm242 Cm243 Cm244 Pb207 Pb208 σfiss 1.896 0.035 0.317 3.525 1.214 1.766 0.360 2.529 0.252 0.245 0.176 0.613 3.267 0.400 σelas 8.794 8.872 JEFF 3.1 σcapt 0.546 0.283 1.563 0.175 0.537 0.497 0.489 0.482 0.492 1.814 1.633 0.474 0.296 0.861 σn,2n 5.59E-04 3.93E-04 ν 2.455 2.757 2.805 2.818 2.982 2.938 3.037 2.970 3.114 3.473 3.498 3.598 3.481 3.487 σtot 8.944 8.899 ENDF/B 7.0 ∆(%) σfis σcapt 0.369 0.733 2.857 0.707 / 5.054 -37.418 142.857 - 9.390 38.734 0.680 - 0.201 / 5.112 - 0.237 - 7.469 - 3.571 - 7.317 / 3.252 1.136 - 4.715 -76.835 -29.114 -10.162 87.838 1.000 -15.796 σelas σn,2n 0.284 4.293 0.259 -4.835 ν 0.163 -0.725 1.462 -9.368 0.268 0.034 -1.482 -0.202 -2.216 -3.455 0.943 6.337 -0.460 -0.201 σtot 0.291 0.258 and other integral parameter variations, the following results have been obtained: ∑ (∆k/k)eff = (keff, ENDF/B7.0 – keff, JEFF3.1)/ keff, JEFF3.1 = – 0.00531 ± 0.00019; ∑ ∆< En > = 0.107 % & ∆< n > = – 0.196 %; ∑ ∆(Fis Events) = – 0.262 %; ∆(Cpt Events) = + 0.194 % & ∆(n,xn Events) = + 7.660 %. For the ELSY critical core, fuelled with “classical” MOX fuel (enriched around the value of 18% in Pu), the integral parameters are similar for the two different xs-libraries. The differences in the crosssections between the files JEFF 3.1 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 and ENDF/B-VII.0 range from less than 1% to about 140%, with limited discrepancies for 235U, 238U 8 and 239Pu isotopes. The impact of ENDF/B 7.0 7 the different ENDF files on the JEFF 3.1 BISERM-2 DPA parameter have been evaluated 6 in [9] for some critical zones of the 5 ELSY reactor. Figure 2 shows the 4 Inner Vessel DPA evaluations; it can be observed that both JEFF 3.1 3 and ENDF/B 7.0 data underestimate 2 from 5.5% to 6.7% the reference results (BISERM-2), but with DPA 1 limited discrepancies between Fig. 2. Inner Vessel DPA (20 y) axial distribution (bottom to top). themselves. 8████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████ 01 ████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████ NUCLEAR – 2009 ████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████ Neutron Parameter Importance For a better understanding of the importance of the various neutron events and in order to “translate” the evaluated percentage variations in “neutron importance”, we start from the neutron balance equation: Total Loss = Total Gain (1) where: Total Loss = Capt_Loss + Fiss_Loss + (n, xn)_Loss + Leakage (2. 1) Total Gain = n*Fiss_Loss / keff + (n, xn)_Gain (2. 2) (n, xn)_Gain = 1*(n, 1n) + 2*(n, 2n) + 3*(n,3n) + … (2. 3) Assuming that keff is a continuous function of the above independent variables and applying: δk eff = ∑ j δk eff it follows that: k eff where: ∂k eff ∂ξ j = ∑( j δξ j ; δk eff k eff ξ j = ν , Cpt _ Loss , Fis _ Loss , ... ) j = ∑Cj j δξ j ; ξj Cξ =ν = 1; Cξ = Fis _ Loss = 1 − Cj = ∂k eff / k eff ∂ξ j / ξ j C ξ =Cpt _ Loss = −k eff k eff ν Cξ =( n , xn ) _ Gain = + k eff (3. 1) (3. 2) Cpt _ Loss ν ∗ Fis _ Loss C ξ =( n , xn ) _ Loss = −k eff ; (n, xn) _ Gain ; Cξ = Leak ν ∗ Fis _ Loss (n, xn) _ Loss ν ∗ Fis _ Loss Leak = −k eff ν ∗ Fis _ Loss (3. 3) The set of the equations (3.1 - 3.3) allows the evaluation of the neutron importance of each neutron event. At this point, it should be underlined that the neutron multiplicity associated factor is equal to 6 (Dk/k) cpt (Dk/k) fis (Dk/k) nou 1=EFIT-BoC JEFF 3.1 - ENDF/B 7.0 2=EFIT-BoC JEFF 3.1 - JENDL 3.3 3=EFIT-BoC JEFF 3.1 - A.C Void Eff. 4=ELSY-BoC JEFF 3.1 - ENDF/B 7.0 5=EFIT-BoC 239Pu from ENDF/B 7.0 6=EFIT-BoC 241Am from JENDL 3.3 5 4 3 2 Neutron Events Importance 1 -0,03 -0,02 -0,01 0,01 0,02 0,03 0,04 Fig. 3. Main neutron event importance Dk/k for some ENDF or physics “ transitions”. 0,05 8████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████ 01 ████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████ NUCLEAR – 2009 ████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████ unity, while all the other neutron event associated factors (Cj) are lower than one. At the same time the fission and capture event associated factors are close (in absolute value) one to the other. Applying the above formulation, Figure 3 summarises the main neutron event importance for some of the discussed cases. In detail, for example: ∑ the substitution of the 239Pu cross sections (JEFF3.1 to ENDF/B-7.0, see case 5 of Fig. 3) implies a neutron multiplicity variation of +0.074%, i.e. neglecting any other neutron event, the associated reactivity variation of 74 pcm (eq. 3.1-3.3), is very close to the simulation result of 68 pcm; when all the neutron events are included, the final result (eq. 3.1-3.3) is equal to 67 pcm. ∑ in case of 241Am cross sections substitution (JEFF3.1 to JENDL 3.3, see case 6 of Fig. 3) a reactivity variation of –1570 pcm is observed. Also the following results are observed: a neutron multiplicity variation of –0.9694% corresponding to a reactivity variation of –969 pcm, a fission events variation of –0.6061% corresponding to a reactivity variation of –418 pcm, a capture events variation of +0.2647% corresponding to a reactivity variation of –183 pcm, etc. Neglecting all the other neutron events the associated reactivity variation, (eq. 3.1 – 3.3) is equal to –1570 pcm, while including all the neutron events a value of –1576 pcm is obtained. The same methodology could be applied to all the other investigated cases. Of course, in the case of the void effect, cross sections data substitution does not matter and the reactivity variation is due only to the physics. Remarks The current investigation shows that the integral and the differential neutron parameters are afflicted by “uncertainties” up to few percent, equivalent to few “thousands” pcm, challenging in this way the neutron design. It was found that both the ENDF/B 7.0 and JENDL 3.3 evaluated nuclear data underestimate the JEFF 3.1 results up to about 4.5% in the keff integral parameter. At the same time the ENDF/B 7.0 data set underestimates the differential parameters, for example up to about 8% for the void effect reactivity, while the JENDL 3.3 data set overestimates the same parameter up to 29%. In addition, using the ENDF/B 7.0 and JENDL 3.3 data, both Pb and fuel reactivity contributions are overestimated, between ~1000 to ~2500 pcm. In terms of single isotope impact, acceptable differences have been observed for 239Pu, while very strong overestimation has been observed for 241 Am in the JEFF 3.1 data. Finally, the high discrepancies carried out by the present sensitivity analysis confirm the results of other previous studies. It is to be noted that the above uncertainties impact on the FAs number required to assure the criticality level and also on the Decay Heath Removal component, due to the 241Am (242Cm via neutron captures) high uncertainties, especially in the case of a large amount of MA loaded in the core. Acknowledgements The authors thank the partners of the IP-EUROTRANS project for their fruitful contribution to the project. Special thanks to the European Commission for the financial support through the FP5 and FP6 programs. 8████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████ 01 ████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████ NUCLEAR – 2009 ████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████ References [1] G. Aliberti, G. Palmiotti, M. Salvatores, C.G. Stenberg, Impact of nuclear data uncertainties on transmutation of actinides in accelerator–driven assemblies, Nucl. Sci. Eng., 46, 13-50, 2004. [2] G. Paliotti, G. Aliberti, M. Salvatores et al., A global approach to the Physics validation of simulation codes for future nuclear systems, International Conference on Reactor Physics, PHYSOR’08, Interlaken, Switzerland, September 14-19, 2008. [3] Artioli, C., et al., Minor Actinides transmutation in ADS: The EFIT core design, International Conference on Reactor Physics, PHYSOR’08, Interlaken, Switzerland, September 14-19, 2008. [4] M. Sarotto, C. Artioli, G. Grasso, D. Gugiu, ELSY core design static, dynamic and safety parameters with the open square FA, ENEA, FPN-P9IX-006, 2009 (WP2 / Deliverable D8). [5] Korovin, Yu.A., et al., Transmutation of radioactive nuclear waste – present status and requirement for the problem-oriented nuclear data base, PRAMANA, Vol.68, N° 2, February 2007, pp. 181-191. [6] Hendricks, J.S., et al., MCNPX, Version 2.6D, Los Alamos National Laboratory report LA-UR07-4137, June 20, 2007. [7] ORNL, 2000. NJOY99.0 Code system for producing point wise and multigroup neutron and photon cross sections from ENDF/B data, PSR 480, March 2000. [8] G. Glinatsis, C. Artioli, C. Petrovich, M. Sarotto, Reactivity coefficients and uncertainty evaluations on the EFIT core neutron design, International Conference on Reactor Physics, PHYSOR’08, Interlaken, Switzerland, September 14-19, 2008. [9] D. Gugiu, G. Glinatsis, G. Grasso, C. Petrovich and C. Artioli, Radiation damage and activation evaluations for the ELSY reactor, 2nd International Conference on Sustainable Development through Nuclear Research and Education, NUCLEAR2009, Piteşti, România, May 27-29, 2009. [10] G. Glinatsis, Stochastic Approach Studies on the 3 Zones EFIT-MgO/Pb-Coolant Core, ENEA Technical Report, FPN-P9EH-005 rev.0, 25 Jun. 2007, Bologna (Italy). 8████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████ 01 ████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████