GENERATION AND INITIAL VALIDATION OF A NEW CASMO5 ENDF/B-VIII.0 NUCLEAR DATA LIBRARY

A new nuclear data library for the CASMO5 advanced lattice physics code has been generated based on the recently-released ENDF/B-VIII.0 evaluation. The ENDF/B-VIII.0 evaluation represents the state-of-the-art in nuclear data and features new evaluations from the CIELO project for 1 H, 16 O, 56 Fe, 235 U, 238 U and 239 Pu. A description of the library generation procedure used to process these data into the CASMO5 586 energy group structure is provided. Initial validation of the new ENDF/B-VIII.0-based library, referred to as the E8R0 library, is also presented and involves the comparison of predicted k–eff and fission rate distributions to measurements from various critical experiments. The critical experiments used in the initial validation of the E8R0 library consist of the B&W 1810 series, B&W 1484 series, DIMPLE S06A/B, and TCA reflector experiment with iron plates. The results from the initial validation indicate that the new E8R0 library provides a satisfactory performance in terms of CASMO5 predicted k–eff and fission distributions.


INTRODUCTION
The fundamental nuclear data used in the generation of multi-group, incident-neutron cross sections is the starting point and foundation of all nuclear reactor simulation and analysis. Advanced in-core fuel management tools, such as Studsvik's Core Management System 5 (CMS5), rely on the two-step scheme for production-level efficient design, analysis and optimization of Light Water Reactors (LWRs). The CMS5 two-step scheme involves the use of the CASMO5 [1] advanced lattice physics code for the generation of homogenized data for downstream use by the SIMULATE5 [2] advanced three-dimensional nodal simulator, which supports the analysis of Boiling Water Reactors (BWRs), Pressurized Water Reactors (PWRs) [3] and VVERs [4,5].
The release of a new nuclear data evaluation affords the opportunity to use the best-available data to possibly reduce biases and improve predictions in LWR analysis. However, experience indicates that careful testing and validation are necessary to determine whether the new and updated data may be regarded as an improvement. Furthermore, the processing of the basic nuclear data evaluation may require changes to the library generation procedure.

UPDATED CASMO5 LIBRARY GENERATION PROCESS
A depiction of the CASMO5 library generation procedure is shown in Fig. 1, which has been adapted from reference [10]. The ENDF/B-VIII.0 neutron cross section, thermal scattering (TSL), radioactive decay and fission yield sub-libraries are depicted near the top of Fig. 1. The neutron and TSL sub-libraries are processed through a sequence of NJOY2016 [11] modules: MODER, RECONR, BROADR, HEATR, PURR, THERMR, and GROUPR. The NJOY2016 sequence is repeated for each individual nuclide present in the CASMO5 library. Once neutron data for a nuclide has been processed with NJOY2016, a modified version of the NJOY94 POWR module is used to read the NJOY2016 output tape and generate an ASCII text file in the expected format. After all nuclides have been processed with NJOY2016 and NJOY94, a utility program (NJXMRG) is used to generate pre-mixed materials available in the CASMO5 library. The resulting ASCII files are concatenated by the NJXLIB utility into a single file, labeled as the ASCII Library in Fig. 1. The ASCII Library is structured into seven sub-files, each of which corresponds to certain type of data for all nuclides, e.g., all resonance data for resonant nuclides is grouped together into a single sub-file for efficiency. Finally, the ASCII Library is converted into a binary file using the CASLIB utility code. Since Transport-Corrected P0 (TCP0) cross sections are written by the CASLIB code into the final library, a dedicated infinite-medium ܲ ே calculation must be performed by CASMO5 using a preliminary library. The output TCP0 cross sections from this calculation are subsequently embedded into CASLIB and the final CASMO5 library is generated. The radioactive decay and fission yield sub-libraries are processed by the RDDFYD utility code. Either cumulative or independent fission yields can be used, which in turn depends on the depletion chains implemented in the CASMO5 code. The explicit representation (or omission) of metastable states of a certain nuclide in the CASMO5 depletion chains is also handled by RDDFYD. Finally, the determination of a representative incident-neutron energy is determined by RDDFYD when processing the fission yield data.
The generation of Intermediate Resonance (IR) λ factors and Resonance Upscatter (RUP) corrections are performed by the RABBLE [12] and MCSD [13] codes, respectively. Delayed neutron data and fission spectra χ are extracted from the NJOY94 output via the CHICALC utility code. These data are not placed in the CASMO5 library and instead are embedded into the source code, as depicted in Fig. 1, which allows for greater versatility when testing options, such as RUP [7].

NEW CASMO5 ENDF/B-VIII.0-BASED E8R0 LIBRARY
Various advanced numerical schemes and features have been implemented into CASMO5 since the original release [1]. The CASMO5 neutron data library has also been updated from the original ENDF/B-VII.1-based E7R1 200-series library [10] to the current E7R1 202-series library. The new E8R0 300series library shares the following features with the E7R1 202-series library: x 586 energy groups with 128 fast groups (20 MeV to 9.118 keV), 41 resonance groups (9.118 keV to 10 eV), 375 fine groups (10 eV to 0.625 eV), and 42 thermal groups (below 0.625 eV). x 1095 nuclides and materials with cross section data (full or absorption-only).
x A total of 119 heavy nuclides (from 221 Rn to 255 Fm) and 491 fission products available in the library.
x Shielded resonance data tabulated at 19 background cross sections and up to 10 temperatures ranging from 239 K to 2700 K. x High-order scattering matrices supporting 2D transport calculation with anisotropic sources.
x No ad hoc adjustment to 238 U resonance absorption as done for ENDF/B-VI data [10].
The new E8R0 300-series library features the following updated data from the ENDF/B-VIII.0 release: x Absorption, fission, transport, and scattering (including ܲ ே ) cross sections.
x Radioactive decay and fission yield.
x Prompt and delayed neutron fission spectra, along with IR λ factors and RUP correction.
x Energy release per fission and capture data.

INITIAL VALIDATION OF NEW CASMO5 E8R0 LIBRARY
A summary of the initial validation of the new CASMO5 E8R0 library is presented in this section. All CASMO5 calculations were performed using a 95 energy-group structure. Energy condensation from the 586-group library structure to the 95 groups is performed through a set of one-dimensional pin cell calculations. The 2D transport solution uses the default Linear Source (LS) Method of Characteristics (MOC) scheme and angular quadrature (64 azimuthal angles, 3 polar angles and a 0.05 cm ray spacing). All numerical results use the TCP0 unless otherwise noted. The problem-specific axial bucklings, used to model the axial leakage effects, have been gleaned from the various critical experiment reports.

B&W 1810 Critical Experiment Series
The    The average CASMO5 calculated k-eff using the E8R0 library is 133 pcm higher than unity over a wide range of burnable absorber types and loadings. The FR statistics indicate a Root-Mean-Square (RMS) difference of less than 2 % relative to the Central Assembly and less than 6 % relative to diagonal pins. The CASMO5 E8R0 predictions are comparable to those obtained with the E7R1 library.

B&W 1484 Critical Experiment Series
The Babcock & Wilcox 1484 critical experiments [15] consist of twenty-one configurations involving low and high leakage cores, as well as PWR fuel storage configurations. A depiction of the Core I through VI configuration is shown in Fig. 3 and Core X and XI in Fig. 4. k-eff results are given in Table III.   Cores I and II do not involve any heterogeneity and only differ in size and shape. Core I consists of 2.459 % enriched (by mass) 235 U fuel pins arranged in a high leakage circular core. Core II consists of the same fuel enrichment as Core I, but the core is arranged in a low-leakage configuration. Cores III through IX represent various hypothetical fuel storage configurations where 15×15 fuel assemblies are ranged in 3×3 space lattices. The spacing between the assemblies and interstitial absorber pin configuration is varied. Cores X through XXI also represent fuel storage configurations. Unlike the previous configurations, isolating/absorbing plates are introduced between the fuel assemblies. The CASMO5 average predicted k-eff for Cores I through IX is only 6 pcm higher than unity and -147 pcm lower for Cores X through XXI. The second set of cores exhibits greater spatial non-uniformity with the presence of the absorber plates, where boron/aluminum isolation plates are present in Cores XIII through XXI. The CASMO5 E8R0 300 library predictions for the B&W 1484 critical experiments are comparable to results obtained with the E7R1 202 library.

DIMPLE S06A/B Critical Experiments
The AEA Winfrith DIMPLE experimental program [16] conducted critical experiments involving a cruciform core configuration, which resembles a rectangular corner of a PWR core, and consists of five 16×16 PWR assemblies with 3 % 235 U enriched UO2 fuel pins. The DIMPLE S06A configuration is surrounded by a water reflector region, whereas the DIMPLE S06B configuration involves a 2.67 cm stainless steel baffle region between the fuel and water reflector regions. The DIMPLE critical experiments indicate the CASMO5 performance for multi-assembly calculations routinely performed to generate PWR reflector data. The geometry of critical experiments is depicted in Fig. 5. The k-eff for each core configuration are given in Table IV. Fission rates for 235 U and 238 U were also measured in select fuel pins. Table IV also provides descriptive statistics for the relative error in the CASMO5 fission rates compared to the measured values.

TCA Iron Reflector Critical Experiments
The Tokai Research Establishment of JAERI conducted critical experiments using a Tank-type Critical Assembly (TCA) while varying the thickness of the steel or steel-water reflector slabs [17]. The fuel assembly consists of a 15×15 PWR design and 2.6 % 235 U enriched fuel. The objective of the experiment was to measure the reactivity effect of two reflector types: steel-only and steel-water reflector containing about 90 % steel and 10 % water. The TCA geometry for a steel-only reflector case is shown in Fig. 6 and a comparison of the CASMO5 predicted and measured reactivity effect [17] is shown in Fig. 7.  The reactivity effect is the change in reactivity from the reference critical water height, which is taken from the assembly with the steel replaced by water, to the critical height with the reflector present. The ratio of the effective delayed neutron yield and mean neutron lifetime, ߚ /Λ, is also provided for the bare TCA configuration. The CASMO5 E8R0 computed ߚ /Λ is 162.8 ‫ݏ‬ ିଵ , as compared to a measured value of 161.5 ‫ݏ‬ ିଵ ± 5.0 ‫ݏ‬ ିଵ . The CASMO5 results using the E8R0 library are comparable to those obtained using the E7R1 202 library.

CONCLUSIONS
A new commercially available CASMO5 nuclear data library has been generated based on the recentlyreleased ENDF/B-VIII.0 evaluation. A summary of the library generation procedure and main features of the new CASMO5 586-group library (referred to as E8R0) is provided in this work. Initial validation of the E8R0 library versus the B&W 1810, B&W 1484, DIMPLE and TCA critical experiments indicate comparable accuracy relative to the previous E7R1 202 library. Although changes were made to Fe cross sections in ENDF/B-VIII.0, they do not have a significant impact in terms of core k-eff, such as in the TCA reflector experiments. Future work involves further validation against MOX critical experiments and comparisons to Post Irradiation Examination (PIE) isotopic data.