Nuclear Data Needs for Generation IV Nuclear Energy Systems

PREFACE.

CONTENTS.

No abstract received.

A workshop on the data needs for Generation IV Nuclear Energy Systems was held in the U.S. in April 2003. A summary of this workshop is provided in this paper. Discussions during the workshop evolved along the traditional nuclear data topical areas of data needs, measurements, evaluations, processing and validation. Recommendations were made on how the Generation IV needs could be better defined and on approaches for resolving the needs.

Four of the six nuclear systems identified by the Gen-IV international forum are relying in fast reactors. The high performance required from these future FR's calls for very innovative core characteristics compared with conventional fast reactor designs, which in turns give rise to new challenges for the available neutronics methods and data. The ERANOS “formulaire” was developed for reliable, precise and efficient calculations of sodium-cooled fast neutron reactor cores. Such a “formulaire” enables the prediction of all the neutronic quantities of interest for reactor design, operation and safety studies, along with their corresponding uncertainties. Given its achievement in terms of accuracy for existing sodium cooled reactors, the way the ERANOS “formulaire” has been designed should be looked as an example for developing similar tools for GENIV fast cores. The methodology for defining nuclear data needs is briefly covered. Target accuracies for GEN-IV neutronic characteristics is an important point to start with. The covariance data is of significant importance for quantifying the needs and evaluators should provide them with their nuclear data. Hence, nuclear data requests should be associated to uncertainty values. Integral experiments have a complementary role to differential measurements for meeting some nuclear data needs (for instance Pu239 fission). High Priority Level Requests should include those nuclides accessible with current differential measurement technology and should include less important nuclides A list of potential requests is existing, quantifying their uncertainties remain a significant effort particularly when looking at fuel cycle quantities.

One of the topics of the International Workshop on Nuclear Data Needs for Generation IV Nuclear Energy Systems was the assessment of the overall quality and validation status of current nuclear data files and processing capabilities. In this paper we report on such an assessment for the latest preliminary releases of ENDF/B-VII and JEFF-3.1.

The Super Critical water Fast Reactor is a Generation IV reactor concept, which presents new and challenging design issues. A correct estimation of the void effect for this water-cooled pressurized system is of fundamental importance to assess its theoretical feasibility. Hence, in this work an overview of the void effect analysis is shown together with the resulting core design issues. The effect of the application of different cross section libraries and models on the core design is also treated.

The Gas cooled Fast Reactor (GFR) is a high priority in the CEA R&D program on Future Nuclear Energy Systems. After preliminary neutronics and thermo-aerolic studies, a first He-cooled 2400MWth core design based on a series of carbide CERCER plates arranged in an hexagonal wrapper were selected. Although GFR subassembly and core design studies are still at an early stage of development, it is nonetheless possible to identify a number of nuclear data needs that could have some impact on the actual design: new materials, decay heat contributors….

A GCR (Gas Cooled Reactor) employs graphite as a neutron moderator and a reflector. At thermal energies, the scattering of the neutrons is affected by the binding characteristics of the scattering nucleus in the moderator. Thus, these effects should be carefully described by well defined scattering laws. The calculations for the scattering laws require an exact shape of the phonon frequency distribution of a material as an input parameter, as well as its lattice structure. Currently several variations of the phonon frequency spectra are available. We have generated different sets of temperature dependent scattering laws for graphite with the module LEAPR of the NJOY using the available phonon frequency spectra. The temperature range of the generated data sets was from 300 to 2000 °K. To observe the effect of these different scattering laws on the criticality of a GCR core, MCNP calculations were carried out and their results were compared with each other. As the basis of a comparison, the k_eff and the temperature coefficients for the moderator and reflector were used.

Transmutation of long-lived radio-nuclides is an option for reducing the hazards linked to the back-end of the nuclear fuel cycle. Previous studies have demonstrated that the main contributor to spent fuel radio-toxicity is by far Pu, followed by Am and Cm. Prerequisite for any efficient transmutation strategy is therefore Pu multiple recycling, whereas Am and Cm could be treated in different ways, including multiple recycling or once-through burning in dedicated targets. In all cases, however, the transmutation efficiency must be maximised, a condition best achieved if, firstly, uranium-free fuels are considered, and secondly, if multiple reprocessing and recycling is considered. In Europe, and in particular at the Institute for Transuranium Elements (ITU), extensive experimental work is being performed to develop fabrication processes for these innovative compounds, and to characterise their properties under irradiation. This work is mostly done within European collaborations, and is partially funded under the European Framework Programmes.

Three of the six Generation IV nuclear energy systems are fast spectrum reactors, and all intend to utilize high concentrations of minor actinides in their fuel either in initial core loads, or as part of the subsequent loadings during recycle. The fuel for each system varies, as do the selected materials for in-core use. These advanced fuels and materials will require nuclear data for analyzing the behavior of the reactor during operation, but there are gaps in the data as it pertains to these designs. It appears that new, more complete nuclear data evaluations are needed to capture the details required in analyzing these systems.

On February 16, 2005 Mexico signed the Kyoto protocol. This official act is very important and hopefully will trigger in the near future the initiation of a nuclear plan for generation of electricity in Mexico. The research related with the fourth-generation of nuclear reactors should be a key step of that plan. Actually the nuclear energy accounts near six percent of the total electricity production with just one plant with two units, Laguna Verde Nuclear Power Plant (LVNPP). At the National Institute of Nuclear Energy (ININ) organization responsible of the research in the nuclear field, there is a group of researches working in projects to assist the operation of the LVNPP and the analysis of new reactors. Therefore the Generation IV is a key point in research and development.

As a contribution to the feasibility assessment of Gen IV and AFCI relevant systems, a sensitivity and uncertainty study has been performed to evaluate the impact of neutron cross section uncertainty on the most significant integral parameters related to the core and fuel cycle. Results of an extensive analysis indicate only a limited number of relevant parameters and do not show any potential major problem due to nuclear data in the assessment of the systems considered. However, the results obtained depend on the uncertainty data used, and it is suggested to focus some future evaluation work on the production of consistent, as far as possible complete and user oriented covariance data.

The Thermal Molten Salt Reactor (TMSR) using the thorium cycle can achieve the GEN IV objectives of economy, safety, non-proliferation and durability. Its low production of higher actinides, coupled with its breeding capabilities - even with a thermal spectrum - are very valuable characteristics for an innovative reactor. Furthermore, the thorium cycle is more flexible than the uranium cycle since only a small fissile inventory (<2 tons by GWe) is required to start one reactor. The potential of these reactors is currently being extensively studied at the CNRS and EdF /1,2/. A simplified chemical reprocessing is envisaged compared to that used for the former Molten Salt Breeder Reactor (MSBR). The MSBR concept was developed at Oak Ridge National Laboratory (ORNL) in the 1970's based on the Molten Salt Reactor Experiment (MSRE). The main goals of our current studies are to achieve a reactor concept that enables breeding, improved safety and having chemical reprocessing needs reduced and simplified as much as reasonably possible. The neutronic properties of the new TMSR concept are presented in this paper. As the temperature coefficient is close to zero, we will see that the moderation ratio cannot be chosen to simultaneously achieve a high breeding ratio, long graphite lifetime and low uranium inventory. It is clear that any safety margin taken due to uncertainty in the nuclear data will significantly reduce the capability of this concept, thus a sensitivity analysis is vital to propose measurements which would allow to reduce at present high uncertainties in the design parameters of this reactor. Two methodologies, one based on OECD/NEA deterministic codes and one on IPPE (Obninsk) stochastic code, are compared for k_eff sensitivity analysis. The uncertainty analysis of k_eff using covariance matrices available in evaluated files has been performed. Furthermore, a comparison of temperature coefficient sensitivity profiles is presented for the most important reactions. These results are used to review the nuclear data needs for the TMSR thorium fuelled reactor.

Since the beginning of the nuclear industry, thousands of integral experiments related to reactor physics and criticality safety have been performed. Many of these experiments can be used as benchmarks for validation of calculational techniques and improvements to nuclear data. However, many were performed in direct support of operations and thus were not performed with a high degree of quality assurance and were not well documented. For years, common validation practice included the tedious process of researching integral experiment data scattered throughout journals, transactions, reports, and logbooks. Two projects have been established to help streamline the validation process and preserve valuable integral data: the International Criticality Safety Benchmark Evaluation Project (ICSBEP) and the International Reactor Physics Experiment Evaluation Project (IRPhEP). The two projects are closely coordinated to avoid duplication of effort and to leverage limited resources to achieve a common goal. A short history of these two projects and their common purpose are discussed in this paper. Accomplishments of the ICSBEP are highlighted and the future of the two projects outlined.

Russian conceptual design development of Gas Cooled Reactor with Fast Neutron spectra (FGR) positions it in kind of source of energy which has capabilities to be used in artificial energy carrier (hydrogen) production and in fissile material breeding. Uncertainty analysis for it is a part of the conceptual design development. Uncertainties of nuclear reactor parameters conventionally can be divided onto two parts where one of them is connected with criticality modeling and the next one – with neutron field and nuclides concentration evolution. It is very conventional division but can be useful for comparisons of individual sources of uncertainty. The evolution of the field of nuclide concentrations in materials of the studied system is described by a dynamic model in lumped parameters. The coefficients of the matrix of the nuclide transmutations are determined from calculation of the steady-state neutron and photon fields and are normalized taking into account the complete capacity of the plant.

We present an approach to uncertainty quantification for nuclear applications, which combines the covariance evaluation of differential cross-sections data and the error propagation from matching a criticality experiment using a neutron transport calculation. We have studied the effect on Pu-239 fission cross sections of using a one-dimensional neutron transport calculation with the PARTISN code. The evaluation of Pu-239 differential cross-section data is combined with a criticality measurement (Jezebel) using a Bayesian method. To perform the uncertainty quantification for such calculations, we generate a set of random samples of cross sections, which is representative of the covariance matrix, and estimate the distribution of calculated quantities, such as criticality. We show that inclusion of the Jezebel data reduces uncertainties in estimating neutron multiplicity.

No abstract received.

A reliable assessment of the uncertainties in calculated integral reactor parameters depends directly on the uncertainties of the underlying nuclear data. Unfortunately, covariance nuclear data are scarce, not only because a significant experimental database for the isotope under consideration must be available, but also because the covariance evaluation process can be rather complex and time-consuming. We attack this problem with a systematical approach and developed, following the initial ideas of D. Smith (ANL), a method to produce a complete covariance matrix for evaluated data files on the basis of uncertainties of nuclear model parameters. This is accomplished by subjecting the nuclear model code TALYS to a Monte Carlo method for perturbing input parameters, an approach that is now possible with the available computer power. After establishing uncertainties for parameters of the optical model, level densities, gamma-ray strength functions, fission barriers etc., we produce random input files for the TALYS code. These deliver, provided enough calculations (samples) are performed, uncertainties + all off-diagonal elements for all open reaction channels. The uncertainties of the nuclear model parameter are tuned such that the calculated cross section uncertainties coincide, to a reasonable extent, with uncertainties obtained from covariance evaluations based on experimental data. If this method proves to be successful, and we will show here that we are not too far off, it will enable mass production of credible covariance data for isotopes for which no covariance data exists……and this constitutes a very significant part of the periodic table of elements.

A simplified MCNP model of a CANDU-SCWR lattice was used to test the sensitivity of the calculated reactivity to various nuclear data files involving issues of current interest. These tests were performed for cooled and voided conditions, with and without axial neutron leakage, for a uniform lattice of mid-life fuel and a mixed lattice with high-burnup and low-burnup fuel in alternate channels. Results were compared using different room-temperature data files for deuterium, various thermal-scattering-law data files for hydrogen bound in light water and deuterium bound in heavy water, and for pre-ENDF/B-VII and ENDF/B-VI.8 data for uranium. The reactivity differences observed were small (typically <1 mk) and increased with axial neutron leakage.

Sustainable development of atomic energy will require development of new types of reactors able to exceed the limits of the existing reactor types in terms of optimum use of natural fuel resources, reduction in the production of long-lived radioactive waste, economic and safety competitiveness. Lead cooled fast neutrons reactor is one of the most interesting candidates with a potential to address these needs. BREST-300 is a 300 MWe lead cooled fast reactor developed by the NIKIET (Russia) with a deterministic safety approach which aims to exclude reactivity margins greater than the delayed neutron fraction. The development of innovative reactors (lead coolant, nitride fuel…) and fuel cycles with new constraints such as cycle closure or actinide burning, requires new technologies and new data from various disciplines: fuel types, fuel designs and fuel reprocessing. In this connection, the tool and neutron data used for the calculational analysis of reactor characteristics requires thorough validation, even if computational codes in Russia and France relies to the calculation of fast reactors' parameters and “fast” experiments. NIKIET developed a reactor benchmark fitting of design type calculational tools (including neutron data). In the frame of technical exchanges between the NIKIET and the EDF (Electricité De France), results of this benchmark calculation concerning the principal parameters of fuel evolution and safety parameters has been intercompared, in order to estimate the uncertainties and validate the codes for calculations of these new kind of reactors. Different codes and cross-sections data have been used, and sensitivity studies have been performed to understand and quantify the uncertainties sources.

We evaluated the sensitivity of several design and safety parameters with regard to five different nuclear data libraries, JEF2.2, JEFF3.0, ENDF/B-VI.8, JENDL3.2, and JENDL3.3. More specifically, the effective multiplication factor, burn-up reactivity swing and decay heat generation in available LFR and SFR designs were estimated. Monte Carlo codes MCNP and MCB were used in the analyses of the neutronic and burn-up performance of the systems. Thermo-hydraulic safety calculations were performed by the STAR-CD CFD code. For the LFR, ENDF/B-VI.8 and JEF2.2 showed to give a harder neutron spectrum than JEFF3.0, JENDL3.2, and JENDL3.3 data due to the lower inelastic scattering cross-section of lead in these libraries. Hence, the neutron economy of the system becomes more favourable and k_eff is higher when calculated with ENDF/B-VI.8 and JEF2.2 data. As for actinide cross-section data, the uncertainties in the k_eff values appeared to be mainly due to ²³⁹Pu, ²⁴⁰Pu and ²⁴¹Am. Differences in the estimated burn-up reactivity swings proved to be significant, for an SFR as large as a factor of three (when comparing ENDF/B-VI.8 results to those of JENDL3.2). Uncertainties in the evaluation of short-term decay heat generation showed to be of the order of several per cent. Significant differences were, understandably, observed between decay heat generation data quoted in literature for LWR-UOX and those calculated for an LFR (U,TRU)O₂ spent fuel. A corresponding difference in calculated core parameters (outlet coolant temperature) during protected total Loss-of-Power was evaluated.

No abstract received.

Generation IV focuses research for advanced nuclear reactors on six concepts. Three of these concepts, the lead, gas and sodium fast reactors (LFR, GFR and SFR) have fast neutron spectra, whereas a fourth, the super-critical water reactor (SCWR), can be configured to have a fast spectrum. Such fast neutron spectra are essential to meet the sustainability objective of GenIV. Nuclear data requirements for GenIV concepts will therefore emphasize the energy region from about 1 keV to 10 MeV. Here, the potential is illustrated of the GELINA neutron time-of-flight facility and the Van de Graaff laboratory at IRMM to measure the relevant nuclear data in this energy range: the total, capture, fission and inelastic-scattering cross sections. In particular, measurement results will be shown for lead and bismuth inelastic scattering for which the need was recently expressed in a quantitative way by Aliberti et al. for Accelerator Driven Systems. Even without completion of the quantitative assessment of the data needs for GenIV concepts at ANL it is clear that this particular effort is of relevance to LFR system studies.

Recent activities on the measurement of neutron capture cross sections by a Japan Nuclear Cycle Development Institute (JNC) and Kyoto university group is reviewed focusing on ²³⁷Np and ²³⁸Np. Firstly, an experimental issue on the measurement of thermal-neutron capture cross section and resonance integral of ²³⁷Np is discussed. In second, the effectiveness of utilizing a double-neutron capture reaction for the measurement of the neutron capture cross section for short-lived nuclei ²³⁸Np is discussed. In third, the measurements of the energy dependence of the neutron capture cross section of ²³⁷Np are reviewed focusing on the experimental progress, including a total-energy detector and a flash-ADC based data-taking system. The experimental results were compared each other and also with some nuclear data libraries to make the problems clear.

An inventive method that allows to determine neutron-induced cross sections of very short-lived minor actinides is presented. We have successfully applied this method, based on the use of transfer reactions, to ²³³Pa, a key nucleus in the ²³²Th-²³³U fuel cycle. A recent experiment using this technique has also been performed in order to obtain the neutron-induced fission cross sections of ^{242, 243, 244}Cm and ²⁴¹Am which are present in the nuclear waste of the current U-Pu fuel cycle. These cross sections are highly relevant for the design of reactors capable to incinerate minor actinides. Preliminary experimental results will be presented.

The OECD Nuclear Energy Agency (NEA) Data Bank is part of an international network of data centres in charge of the compilation and dissemination of basic nuclear data. Through its activities in the nuclear data field, the NEA participates in the production of data and their distribution to nuclear data users. The high priority request list is an example of such a project. The NEA thus provides an essential link between producers and users of nuclear data. The NEA Data Bank distributes the main computer codes and nuclear databases with bibliographical information, evaluated libraries, e.g. JEFF, and experimental data in the data base EXFOR comprising published neutron induced as well as charged particle induced nuclear reaction data. The new data library JEFF-3.1 will be presented here, as well as the data display tool JANIS. The NEA is also involved in the work in the Generation IV International Forum (GIF) technical working groups that are developing research programs for advanced reactor concepts.

The development and production of various types of nuclear data represents an important activity of the IAEA Nuclear Data Section, as demanded for a wide range of energy-based applications. Recent worldwide initiatives to study innovative designs of nuclear power systems require extensive and accurate nuclear data for materials to which little attention has been devoted in the past. Four highly relevant IAEA projects are described in this paper: two are on going, and two are in the planning stage.

Evaluation activities are crucial steps to combine experimental data from microscopic experiments into coherent and comprehensive nuclear data sets for nuclear energy system applications. The nuclear data group of the CEA/DEN is involved in the elaboration of the European library JEFF since the beginning of the project. New evaluated files for fission products and actinides rather than successive revisions of previous evaluations are planned to satisfy GEN-IV issues.

We report on our current efforts to re-evaluate neutron-induced reactions on americium isotopes (241, 242g,m and 243) in the fast energy region up to 20-MeV. Modern theoretical modeling was used extensively in all those evaluations, due to the rather limited number of experimental data available. This paper is mostly concerned with Am fission cross sections evaluations.

No abstract received.

The benchmark calculations for several graphite-moderated reactors have been performed by the Monte Carlo code MCNP4C using the libraries based on the ENDF/B-VI.8, JENDL-3.3 and JEFF-3.0. The calculation results for each library have been compared with the experimental values. The isotopic contributions to the k_eff calculation have been estimated to clarify the causes of the differences among the three libraries.

AUTHOR INDEX.