Activities are essentially aimed at supplementing the evaluated nuclear data libraries and improving the reaction models for materials in transmutation fuels, ...
NUDATRA: NUCLEAR DATA FOR TRANSMUTATION IN IP-EUROTRANS
Enrique M. González1, A. Koning2, S. Leray3, A. Plompen4, J. Sanz5 On behalf of the NUDATRA project 1 CIEMAT (Spain), 2NRG (The Netherlands), 3 CEA (France), 4IRMM/Geel (JRC/EU), 5UNED (Spain)
Abstract The objective of NUDATRA, Domain 5 of the EU Integrated Project EUROTRANS (FI6W-CT-2004516520), is to improve and validate the nuclear data and simulation tools required for the development and optimisation of nuclear waste transmutation, ADS dedicated transmutation systems and the associated fuel cycle. Activities are essentially aimed at supplementing the evaluated nuclear data libraries and improving the reaction models for materials in transmutation fuels, coolants, spallation targets, internal structures, and reactor and accelerator shielding, relevant for the design and optimisation of the ETD and XT-ADS. These activities are distributed over four Work Packages: Sensitivity Analysis and Validation of Nuclear Data and Simulation Tools; Low- and Intermediate-energy Nuclear Data Measurements; Nuclear Data Libraries Evaluation and Low-intermediate Energy Models; and High-energy Experiments and Modelling. The main accomplishments expected from NUDATRA are: 1) new measurements and evaluations of Pb-Bi cross-sections, i.e. inelastic, (n,xn) and isomer branching ratios (Po production); 2) new measurements and evaluations for minor actinides particularly the capture in fission on 244Cm;
243
Am and
3) improvement of TALYS as an evaluation tool and as an a priori model for the estimation of low- and intermediate-energy reaction cross-section; 4) high-energy model improvement based on measurements, particularly for the prediction of the spallation products, and gas (H, He) production cross-sections; 5) sensitivity and uncertainty analysis of ETD fuel cycle and related covariance issues. One year after the project start, substantial advances had been achieved with many new measurements being analysed, progress on the capabilities of the TALYS code to perform data evaluation and low-energy reaction modelling, with new versions of the high-energy codes INCL4 and ABLA and new versions of simulation codes like MCB and EVOLCODE being tested, and a systematic study of the possible methodologies and data for uncertainty evaluations in fuel cycle simulations and its inclusion in the ACAB inventory code (soon to become part of EVOLCODE2). 397
Introduction Computer simulation remains – and will continue to remain for years – the basis of the evaluation and extrapolations of performance, viability, cost and safety of proposed transmutation devices. Improvements in the simulation tools, programs and evaluated nuclear data libraries are key elements to analyse our cost/benefit evaluations and to prepare a solid base for the definition of DEMO reactors that can provide the required validation before industrial deployment of transmuters. IP-EUROTRANS, the FP6 EU project on transmutation (FI6W-CT-2004-516520), has dedicated a Domain to address the improvements of these tools. The Domain is designated NUDATRA; it includes improvement of simulation programs and underlying models, data measurement and evaluation, uncertainty propagation and sensitivity analysis. Taking into account that EUROTRANS is concentrated on the transmutation of actinides in subcritical systems, mainly those cooled by lead-bismuth, special priorities had been assigned to the data requirements for this type of devices. In addition the scope of the projects has been extended to include the data needs for the associated fuel cycles. The selection of activities within this framework was initially based on the relevance of proposed activities for the viability, performance, safety and cost of the transmutation or its fuel cycle, according to basic principles. The correctness of the choices had been later confirmed by independent sensitivity analysis [1]. In addition, the viability in the FP6 time frame and the complementarities with other EU programmes (in particular FP5), were taken into account when selecting experimental measurements and validations. From the fuel cycle point of view, the nuclear data for transmutation are defined according to the needs to compute the isotopic composition of the equilibrium/transitory fuel in all the steps of the fuel cycle. This includes the characterisation at fabrication, loading and reprocessing of the different fuels involved in the selected strategy, as well as the corresponding losses finally going to the storage. This composition will determine the radioactivity, neutron emission and heat and the associated radioprotection and cooling needs. The composition is determined by the isotopic composition of the LWR wastes fed into the transmutation reactor, the isotope decay constants, the neutron flux intensity (reactor power) and the effective cross-sections of the activation reactions. In consequence, the activation reaction cross-section [(n,), (n,)*, (n,2n), …] of actinides with medium and long half-lives (> 100 d) will be very important. In addition, the effective cross-section depends on the neutron flux spectrum. For this reason the elastic, inelastic, (n,2n) reactions cross-section of fuel matrix, structural materials and coolant need to be known with high precision. On the other hand, the transmutation takes place in a reactor, critical or subcritical (ADS), but in all cases with new features. In all cases new fuels, with high content on minor actinide and high mass Pu isotopes, and very high burn-up per irradiation cycle are proposed. Furthermore, most frequently the devices present a dominant fast neutron flux spectrum. In the case of subcritical configurations, the spallation source brings the very high-energy proton and neutrons with the corresponding needs on high-energy models and data. Finally, new technologies are proposed for these transmutation devices, like molten lead or Pb/Bi for coolant, and inert matrix fuels. These technologies introduce new elements and isotopes in the reactor with the corresponding new data needs (mainly activation, elastic, inelastic and (n,xn) reactions). The more systematic sensitivity analysis of [1] shows that from the impact on Keff, eff, source importance, and many other parameters, the most relevant data improvement needs on minor actinides are on 241Am (capture and fission), 243Am capture and 244Cm fission. After these isotopes the next most
398
important reactions are the Pb and Bi inelastic and (n,2n) cross-section that are especially relevant for He, H production, reactivity loss during irradiation and void coefficient. In the final table of most-needed data, these reactions and a few other fission cross-sections already measured at nTOF appear. Among the selection of measurements were also considered the measurements and expected results from the recently finished nTOF [2] and HINDAS FP5 projects, and the huge difficulty to perform direct measurements with the very short-lived isotopes 238Pu, 241Am and 242mAm. The NUDATRA EU project The NUDATRA project general objective is to improve nuclear data evaluated files and models including sensitivity analysis and validation of simulation tools, low- and intermediate-energy nuclear data measurements, nuclear data libraries’ evaluation at low and medium energies, and high-energy experiments and modelling. It includes the contribution of 22 participants (13 research centres and 9 universities): CEA (France), CIEMAT (Spain), CNRS (France), CSIC (Spain), FZJ (Germany), FZK (Germany), GSI (Germany), INFN (Italy), INRNE (Bulgaria), NRG (Netherlands), PSI (Switzerland), SCK CEN (Belgium), JRC-Geel (EC) and the universities of: AGH (Poland),TUW (Austria), KTH (Sweden), ULG (Belgium), UNED (Spain), USDC (Spain), USE (Spain), UU (Sweden), ZSR (Germany). The project is organised into four Work Packages (WP): WP5.1 – Sensitivity analysis and validation of nuclear data and simulation tools; WP5.2 – Low- and intermediate-energy nuclear data measurements; WP5.3 – Nuclear data library evaluation and low- and intermediate-energy models; WP5.4 – High-energy experiments and modelling. The NUDATRA activities of these Work Packages concentrate in very few topics related to the issues previously identified as highly needed for the transmutation development:
Pb-Bi cross-sections: inelastic, (n,xn), Po production (B.R. in the Bi capture reaction);
minor actinides: capture in 243Am and fission on 244Cm;
TALYS improvements for minor actinides evaluation and test on new Pb data;
high-energy measurements: gas (He) and light-charged particles production and absolute spallation product cross-section;
high-energy models improvement (INCL & ABLA);
sensitivity analysis of experimental transmutation device (ETD), fuel cycle;
new versions of transmutation simulation systems.
Low- and intermediate-energy nuclear data measurements The first group of measurements are on the Pb and Bi cross-sections and branching ratios. A large amount effort has been put toward achieving high resolution excitation functions for the inelastic scattering cross-sections of Pb and Bi. As indicated before and evaluated numerically in [1], these cross-sections are critical for modelling the ADS neutron spectra and thus all the effective one-group cross-sections. 206Pb, 207Pb, 208Pb and 209Bi (n,n ) have already been measured by NUDATRA from 399
threshold to 20 MeV at Gelina, Figure 1. The gamma-ray production cross-sections are measured at several angles for several gamma-rays (up to 39 in the case of 209Bi), and then the total gamma-ray production, the level and the total inelastic cross-sections are deduced from the data. In fact, a fraction of the data is already in the form of EXFOR files. The characteristics of Gelina and an optimised measurement set-up have allowed obtaining a high-energy resolution, largely improving the energy range and the number of points of previous measurements. Measurement precision has also been improved and work is ongoing to minimise the theoretical uncertainties involved in the cross-section determination from the measurements. Figure 1. 206Pb, 207Pb, 208Pb and 209Bi (n,n) cross-sections measured at Gelina within NUDATRA The points on the new measurements are so close that results are difficult to distinguish in the figure and may be taken as a continuous red line
209
206
Bi
Pb
Total inelastic
Total inelastic
207
208
Pb
Pb
Total inelastic
Total inelastic
Measurements of the Pb and Bi (n,xn) cross-section, which contribute to the neutron multiplication, the source importance and neutron spectra of ADS cooled with Pb-Bi or using Pb-Bi spallation target, are also very well advanced. Measurements at Gelina with on-line HPGe detectors are being performed and analysis is ongoing for: 207Pb(n,2n)206Pb, 208Pb(n,2n)207Pb, 208Pb(n,3n)206Pb and 209Bi(n,2n)208Bi. The preliminary results show an improvement in the energy resolution and the measurement statistics. Two independent measurements using different techniques will provide further validation. The third type of ongoing measurement is the Bi capture branching ratio [209Bi(n,)210m,gBi]. The relevance of this branching ratio is that the production of 210gBi is the mechanism leading to 210Po generation, expected to be one of the largest hazards from the target and coolant during and shortly after irradiation. On the other hand, 210mBi decay to 206Tl at a much slower rate. The time-of-flight 400
technique will be used at Gelina, with two or three HPGe detectors in coincidence to distinguish between capture events leading to the ground state and the meta-stable state. Set-up has been prepared and optimised and the corrections to compensate for angular dependence has been tested. Preliminary data was presented at PHYSOR 2006 [3]. Finally, the last coolant related cross-section measured is the inclusive Pb(n,n+X) at 100 MeV. The data is fundamental to make a precise estimation of shielding needs on Pb or Pb-Bi transmutation plants, particularly on ADS systems, but up to now no double-differential data was available. This cross-section also contributes to the multiplication of the spallation neutrons. The determination will be based on measurements done at the Scandal facility at Uppsala. Set-up and analysis techniques are being optimised on the basis of the preliminary analysis of 2003-2004 data. The low-energy measurements for minor actinide concentrate within NUDATRA only in two reactions. First the capture cross-section 243Am(n,) at nTOF-Ph2 (at CERN). As seen at [1], 243Am capture cross-section uncertainties is one of the main contributions to the production of the 244,245,246,247 Cm family. The measurement will use the very high instantaneous beam intensity of the nTOF CERN time-of-flight line with a 4 BaF2 Total Absorption Calorimeter, the last generation of electronics and FADC-based data acquisition systems. The methodology was developed and set up in 2004 at the FP5 nTOF-ADS project. From the beginning of NUDATRA we have performed a preliminary analysis of 2004 data [4], that is allowing to optimise the new measurement campaign set-up and the analysis chain. A proposal including 243Am(n,) measurement has been submitted by the nTOF_Ph2 collaboration to the Research Board. The second MA measurement will provide the 244Cm neutron-induced fission cross-section. The direct measurement is extremely difficult because of the short half-life and high spontaneous fission probability of 244Cm(n,f). At NUDATRA the 244Cm(n,f) will be obtained from the measurement of the transfer reaction 243Am(3He,pf) at Orsay. Both reactions generate the same composite nucleus and, with the help of theoretical models for the different probability of formation of the composite nucleus, the cross-section can be evaluated. After set-up optimisation the first data have been taken. The precision and viability of the measurement have been demonstrated in the concurrent channels 243Am(3He,f) and 243Am(3He,3Hf). The first channel allows to evaluate the 241Am(n,f) obtaining excellent agreement with the direct measurements. The second channel provides the 242Cm(n,f). Nuclear data library evaluation and low- and intermediate-energy models Low- and intermediate-energy models have a double potential of application for transmutation. On one hand, each measurement must be evaluated to become useful for simulations, and these models are required in the evaluation. On the other hand, a priori models can help to complete libraries providing estimation of many channels and isotopes without experimental data, in many cases, with precisions better than 30%, above the resolved resonance region. In NUDATRA both aspects are developed in the environment of the TALYS code. For the first, the nuclear reaction models of low and intermediate energy included in the code TALYS are being improved in several areas. One relevant example is the generalised superfluid model for level densities implemented, at present being tested with U isotopes. In addition there is an important effort in the development of methods to generate covariance data to be implemented within the code. Two such methods are being studied and are making good progress. A second line of work is the preparation of TALYS for the evaluation of minor actinide data. The optical model, pre-equilibrium, compound nucleus and fission model parameters are being fine-tuned 401
for this purpose. Once available, the priority will be to re-evaluate the americium isotopes, beginning with 241Am, in the fast neutron range but also including the resolved resonance regions. Indeed, new isospin-dependent dispersive optical model potentials for actinides have already been developed and sent for publication. Finally TALYS is already being used in the re-evaluation of data libraries for Pb and Bi, using the data from Gelina described earlier to complement the existing and FP5 data (nTOF,…). The first test made shows good agreement with the experiment and very promising results. High-energy experiments and modelling Within NUDATRA we define the high-energy range as above 200 MeV. This range is specific of the ADS, where the high-energy protons are used in the spallation target to generate the neutrons required to maintain the fission chains in the subcritical system. This energy range is characterised by the lack of cross-section libraries and so the simulations are based on models validated by experimental measurements. The objective of NUDATRA in this field is to complete the experimental database of the HINDAS FP5 project, and to improve the existing high-energy models, particularly the INCL and ABLA models. High-energy experiments for radioactivity, chemical modification and damage assessment are included in the project, together with the improvement of the absolute value of the total fission cross-section for Pb and W in the 200 MeV-1 GeV energy range. The present progress on the high-energy nuclear data measurement includes the measurement of the total fission cross-sections in reactions induced by 208Pb on 1H and 2H at 500 A MeV at GSI. On the other hand new results are being measured for the production of long-lived intermediate mass fragments (IMF) as 7Be and 10Be from Bi, W, Ni targets in the range 100-1 000 MeV. Further, He production in Pb, Ta and Fe, between 100 and 800 MeV, are being measured in the NESSI and PISA experiments at FZJ and at Hannover University (Figure 2). The new results are already providing better compatibility between the results of different facilities and measurement techniques. Finally, by performing several measurements at the two experiments NESSI and PISA of the same facility, our confidence regarding the absolute determination of the corresponding cross-sections is increasing (Figure 3). Several improvements are being introduced in the high-energy nuclear models. For example, the implementation of a dynamical coalescence model for cluster production in INCL4 seems to correctly reproduce the production of high-energy light clusters (except for the 4He/3He ratio). In addition, Figure 2. He production cross-section in lead at ZSR and in iron at NESSI and ZSR (Michel)
Helium production cross−section (mb)
p + Fe
1000
10
402
NESSI Michel Bertini−Dresner INCL4−ABLA INCL4−GEMINI INCL4−Clus−GEMINI NESSI new SPALADIN preliminary
100
0
500
1000 1500 2000 Incident energy (MeV)
2500
3000
Figure 3. He and Li production cross-section in Au at NESSI and PISA 4
o
Nessi1 Nessi1 23 Entries Entries 23 Mean 23.5 Mean 23.5 11.72 RMS RMS 11.72
Au(p, He)X 2.5GeV 100 Au(p,4He)X 2.5GeV 100º 2
10
Au(p,4He)X
at 2.5 GeV,
100o
Au(p,7Li)X at 2.5 GeV, 35o
10 o
Nessi (105 ) Nessi (105º) Pisa 2005 Pisa 2005
1
10-1
-2
10
0
20
40
60
80
100
isospin- and energy-dependent average nucleon potentials and improvements of pion dynamics are being introduced in this code. On the other hand, ABLA is being improved in its description of the fission, light-charged particles and intermediate mass fragments production. A detailed description of the progress in these models and the preliminary comparison of their prediction with the available data have recently been presented at [5]. After implementing the new versions of these models in standard high-energy transport codes (particularly MCNPX), a quality assessment, validation and impact of the new models in ADS (ETD) simulations will be preformed with calculations of radiotoxicity, radioactivity due to residue production in the MEGAPIE experiment and performing calculations of dpa, chemical composition modifications, and activities in ETD with the new codes. Sensitivity analysis and validation of nuclear data and simulation tools Most of the present neutronic computer simulations of nuclear systems have an uncertainty on their results that, very often, is dominated by the propagation of the uncertainties of the nuclear data used in the simulation. Tracking back from the uncertainties on the simulation of critical parameters (safety, viability, performance, cost,…) it should be possible to identify the largest contribution to these uncertainties and so the most relevant and useful new nuclear data measurements. This is the concept of the sensitivity analysis. Some previous work has explored this question from the point of view of a generic ADS devoted to transmutation. NUDATRA will address the problem from the transmutation (ETD) fuel cycle point of view (specific reactor related sensitivity analysis will still be performed by other groups of IP_EUROTRANS), and in this sense a list of topics for sensitivity evaluation of the transmutation ADS fuel cycles had been selected. Several difficulties are encountered as regards the sensitivity analysis process. On one hand the present databases are sorely lacking in uncertainty information for most nuclear data involved in the transmutation simulation. Within NUDATRA a wide review, compilation and comparison of uncertainties from the most recent activation data files, evaluated nuclear data files and bibliography proposals has been performed. Although it is possible to find uncertainties for approximately 400 cross-sections, less than 50 have the full covariance matrixes. To complete the uncertainty propagation the database must be completed, so several proposals of methodologies for the definition of covariance matrixes on nuclear data, when not available, had been tested and compared. In particular a new proposal based on a parametric covariance matrix has been used to estimate the possible effects on the 403
uncertainty propagation and the sensitivity analysis derived form the lack of full covariance matrix. As an example the relative uncertainties in the fraction of isotopes as 238Pu, 242Cm or 245Cm in the discharged fuel of a typical ADS for transmutation can change by more than a factor of 2 using different covariant matrixes. The second difficulty is that few codes are able to use the covariance information and even the uncertainty on the cross-section. In NUDATRA a special task is dedicated to design and prototype a new version of the simulation codes, in particular EVOLCODE2, which should be able to fully exploit all the available covariance information [6]. At the same time EVOLCODE2 will be adapted to increase the number of available activation reactions, to better represent the fuel evolution in the presence of large fast (and high-energy) neutron fluences. A third difficulty emerges from the large number of parameters that can affect the uncertainty of a particular result. This might generate very asymmetric confidence margins for the estimated result. More generally, the probability distribution of the estimated result could present peculiarities. This possibility will be analysed in detail at NUDATRA for the ETD fuel cycle applying a Monte Carlo technique that allows obtaining an estimation of the estimation probability distribution. The result best estimate and uncertainty obtained in this way will be compared with the linear combination of variances. First comparisons had already been obtained with simplifying hypothesis of the neutron flux spectrum dependence on the reactor position, and they show good agreement between both methods but some deviation from the Gaussian behaviour on isotope contents after irradiation (e.g. 242Cm). The sensitivity analysis is complemented with the development and validation of simulation programs for transmutation plant. This effort concentrates on the new version of codes like MCB (from KTH), EVOCODE2 (from CIEMAT) and KAPROS/KARBUS (from FZK). For EVOLCODE2, significant upgrades had been introduced to improve the approximation of the energy dependence of fission yields, the isomer production, the cross-section convolution method, and the management of the isotopes without transport cross-sections. New options and functionalities have also been added. For example, now it is possible to use different cross-sections (temperature, library,…) for the same isotope at different cells. We also have largely increased the number of isotopes for the burn-up steps (particularly short-lived spallation products). Finally we have introduced improvements on the coding and portability and all known bugs were fixed. In the case of MCB, new options, like the possibility to normalise to proton intensity or total power, the possibility to include the spallation products or like an improved fuel lattice handling, have been introduced. Most known bugs were also fixed. Both codes are being used in several international projects and intercomparisons to improve the validation of their data, models and programs. In addition, NUDATRA has foreseen two activities on nuclear data and models validation, one for the spallation target and one for low-energy data. The first is based on the measurements of residual nuclei production in SINQ targets and on determinations of absolute radioactivities of the residues (e.g. 194Hg, 207Bi) produced in the irradiation of spallation target models at Dubna, within the preparation of the SAD project. Indeed, an international IAEA benchmark has been predefined from the data of these Dubna tests. The second activity line addresses the minor actinide and Pb nuclear data validation in integral experiments. The experiments performed in MASURCA (Cadarache) will allow validating the fission cross-section of 240,241,242Pu, 237Np and 241,243Am. The first analysis done from MASURCA 1A’ and 1B has provided information on the main isotopes of the MOX fuel. Other minor actinide and Pb nuclear data will be validated with the results from ISTC projects (BFS, SAD, Yalina facilities). 404
Conclusions and outlook Very substantial progress has been obtained in the first year of the NUDATRA project: In WP5.1 it is possible to highlight the development and validation of sensitivity analysis methodologies and collection of existing covariance information, the development of new versions of transmutation simulation codes (EVOLCODE2, MCB) and the test of old and new codes like KAPROS and ALEPH, respectively. In WP5.2 the most significant achievement is the measurement of 206,207,208Pb and 209Bi, from threshold to 20 MeV by (n,n), of (n,xn) for the same isotopes and with two techniques, at Gelina, the preparatory set-ups and analyses for the future 244Cm(n,f), 243Am(n,), Pb (n,nX) and the ongoing 209 Bi(n,)210m,gBi branching ratio determination. In WP5.3 the largest achievement is the improvement of TALYS (low-intermediate energy reactions modelling code used for evaluation) to be able to handle actinides and on the update of Pb (inelastic) cross-sections evaluation including the most recent data. In WP5.4, the measurements of He and Be production in high-energy proton reactions at NESSI/PISA, the preparations for the measurements of total spallation cross-sections at GSI, and the improvements on high-energy reaction models INCL and ABLA provide a remarkable step forward in the field of high energies. NUDATRA will run until 2008 and in the coming years we plan to complete the evaluation of uncertainties for transmutation fuel cycles most relevant parameters and to identify any additional data needs. We will complete a set of reliable databases for Pb and Bi data at low energies and will improve the precision and reliability of most relevant Am and Cm data. A large upgrade is foreseen for TALYS, which will allow its use to prepare complete cross-section and covariance libraries generated by a priori models and to improve the evaluations of Pb, Bi and Am isotopes. Finally the new high-energy data and the progress of theoretical models should provide validated reliable high-energy codes within MCNPX.
REFERENCES
[1]
Aliberti, G., et al., “Impact of Nuclear Data Uncertainties on Transmutation of Actinides in Accelerator-driven Assemblies”, Nuclear Science and Engineering, 146, 13-50 (2004).
[2]
Paradela, C., et al., “n_TOF Fission Data of Interest to GEN-IV and ADS”, PHYSOR-2006, ANS Topical Meeting on Reactor Physics, Vancouver, BC, Canada, September 2006.
[3]
Borella, A., et al., “Determination of the Branching Ratio for the 209Bi(n,)210Bi Reaction from 500 eV to 20 keV”, PHYSOR-2006, ANS Topical Meeting on Reactor Physics, Vancouver, BC, Canada, September 2006.
405
[4]
Guerrero, C., et al., “Measurement at n_TOF of the 237Np(n,) and 240Pu(n,) Cross Sections for the Transmutation of Nuclear Waste”, PHYSOR-2006, ANS Topical Meeting on Reactor Physics, Vancouver, BC, Canada, September 2006.
[5]
Leray, S., et al., “Achievements and Deficiencies of Nuclear Models Used for the Design of Spallation Sources”, PHYSOR-2006, ANS Topical Meeting on Reactor Physics, Vancouver, BC, Canada, September 2006.
[6]
García-Herranz, N., et al., “Applicability of the MCNP-ACAB System to Inventory Prediction in High-Burn-up Fuels: Sensitivity/Uncertainty Estimates”, International Conference on Mathematics and Computation, Supercomputing, Reactor Physics, and Biological Applications (M&C’2005), Avignon, France, September 2005.
406
