Accepted Manuscript
Processing of JEFF-3.3 and ENDF/B-VIII.0 and testing with critical benchmark experiments and TRIGA Mark II research reactor using MCNPX
Kabach Ouadie, Chetaine Abdelouahed, Benchrif Abdelfettah
PII: S0969-8043(19)30148-4
DOI: https://doi.org/10.1016/j.apradiso.2019.05.015 Reference: ARI 8729
To appear in: Applied Radiation and Isotopes Received Date: 18 February 2019
Revised Date: 17 April 2019 Accepted Date: 13 May 2019
Please cite this article as: Ouadie, K., Abdelouahed, C., Abdelfettah, B., Processing of JEFF-3.3 and ENDF/B-VIII.0 and testing with critical benchmark experiments and TRIGA Mark II research reactor using MCNPX, Applied Radiation and Isotopes (2019), doi: https://doi.org/10.1016/
j.apradiso.2019.05.015.
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
M AN US CR IP T
AC CE PT ED
Processing of JEFF-3.3 and ENDF/B-VIII.0 and Testing with Critical Benchmark Experiments and TRIGA Mark II Research Reactor
using MCNPX
KABACH Ouadie1, CHETAINE Abdelouahed1 and BENCHRIF Abdelfettah2
1) Mohammed V University, Faculty of Science, Nuclear Reactor and Nuclear Security Group Energy Centre, Physics Department, 4 Avenue Ibn Battouta B.P. 1014 RP, Rabat 10000, Morocco.
2) National Centre for Nuclear Energy, Sciences and Technology (CNESTEN), Morocco.
Abstract
A comparative study has been performed with the latest evaluated nuclear data libraries JEFF-3.3 and ENDF/B-VIII.0. The study has been conducted through the benchmark calculations for 120 criticality problems and the TRIGA Mark II research reactor with the libraries processed using NJOY21 for MCNPX Monte Carlo transport code.
The criticality benchmarksassemblies, taken from the ICSBEP benchmark, cover Uranium (highly enriched uranium, intermediate enriched uranium, low enriched uranium, and 233U) and Plutonium fuel systems in a various metal forms, and under a various spectral conditions. The Moroccan TRIGA Mark II research reactor calculation is used to look into the predictive capability of those nuclear data libraries and then to compare the accuracy of the predicted results with the experimental data published elsewhere.
Actually, the purpose of this study is to investigate some neutronic and kinetic parameters of those benchmarks for both libraries. The former consist of effective multiplication factor, heat distribution, neutron flux distribution, effective delayed neutron fraction (βeff), prompt removal lifetime (τr) and the mean neutron generation time (Ʌ). The results show that the calculated effective multiplication factor, heat distribution, neutron flux distribution, and the kinetic parameters are in good agreement with references. However, it is found that the computed values are strongly depending on the nuclear data set used in calculations.
Keywords: ENDF/B-VIII0, JEFF-3.3,Criticality calculation, TRIGA Mark II,NJOY21, MCNPX.
1. Introduction
The nuclear data evaluations have been separately released from different countries. ENDF/B-VI of Cross-Section Evaluation Working Group (CSEWG) and JEFF of OECD/NEA have been widely utilized in the nuclear community in light of the fact that these libraries include reliable, evaluated data for all neutron reactions. In the last years, new versions of the two nuclear reaction data libraries were released: JEFF-3.3 in 2017 (O. Cabellos et al., 2017) and ENDF/B-VIII.0 in 2018 (Brown et al., 2018) with a significant upgrade in data for a number of nuclides(Carlson et al.
2018). The ENDF-6formats (Trkov, Herman, and Brown, 2018) are accessible at the IAEA website (International Atomic Energy Agency, 2018).
This work was intended to generate the continuous energy neutron data libraries for the MCNPX (Pelowitz, 2008) based on the latest evaluated nuclear data and to analyze them through benchmarking with the 2 MW TRIGA Mark-II research reactor and the criticality benchmarks calculations. Application libraries have been independently created with the NJOY21(1.0.0.json) Nuclear Data Processing System (Macfarlane et al., 2017) available at LANL website (Los Alamos National Laboratory, 2018).
M AN US CR IP T
AC CE PT ED
A broad outline of this paper begins with: i) testing the processed libraries of critical benchmark models defined in the International Criticality Safety Benchmark Evaluation Project (ICSBEP) handbook (Nuclear Energy Agency, 2010). In this part of this work the effective multiplication factors (keff) are calculated and compared to the published benchmark values (Nuclear Energy Agency, 2010); ii) benchmarking the processed libraries with TRIGA Mark II research reactor for the well-known both neutronic and kinetic parameters. The complete 3D geometry of the TRIGA reactor core, using MCNPX, is implemented with high accuracy and details, exploiting all the available data about the geometry and materials. Finally, as part of the evaluation process, comparisons with the JEFF-3.2 (Oscar Cabellos, 2014) and ENDF/B-VII.1 (Chadwick et al., 2011) libraries were indeed carried out.
2. Calculation methods 2.1. Library generation
The nuclear data evaluations are physical representations of the data encoded in a unified computer- readable format called ENDF-6. They need to be converted into suitable forms for applications, such as MCNPX applications. The NJOY21 Nuclear Data Processing System ( Macfarlane et al., 2017) is perfectly able to handle this unified format to create an ACE format needed by MCNPX.
For the self-shielding effects, the probability tables in the unresolved resonance range were generated using the PURR module. The s(α, β) thermal scattering data treatments were also generated, these thermal scattering data are fundamental for finer modeling of the neutron interactions at energies below ̴ 4 eV. The Fig1 illustrates the NJOY sequence to process the latest libraries in ACE format. Further description of the nuclear data processing can be found in (MacFarlane and Kahler, 2010; Ouadie et al., 2017).
Fig.1. NJOY sequence to process the latest libraries in ACE format(MacFarlane and Kahler, 2010).
2.2. Simulation code
In this study, the MCNPX v2.6.0 (Pelowitz, 2008) computer code was used on benchmarking of both processed data libraries. MCNPX is a general-purpose, continuous-energy, generalized- geometry and Monte Carlo transport code. It can be used for neutron, photon, electron, or coupled neutron/photon/electron transport, including the capacity to figure eigenvalues for critical systems.
It helps to generate the 3D model of either simple or complex geometry in details. In order to achieve an appropriate precision with the MCNPX code of the studied benchmarks, the
M AN US CR IP T
AC CE PT ED
computations were carried out with an HP Elite-Book 8560 workstation Intel® i7 CPU 2.20 GHz, 8 cores, 8 Gb RAM and 6 Mb Cache Memory under Win 10 system and utilizing MPICH2 (Liang and Liu, 2008). The goal of MPICH is to provide an MPI implementation that efficiently supports multi-core architectures.
3. Benchmark calculation 3.1. Criticality benchmark
A brief description of the categorized criticality benchmarks
In this investigation, all criticality benchmarks were taken from OECD-NEA project ICSBEP (Nuclear Energy Agency, 2010). They are subdivided into a few fundamental classes as demonstrated in the Table1.
Table 1. ICSBEP abbreviations used (Marshall and Rearden, 2008).
Abbreviation Meaning
Fissile material
HEU High enriched uranium ( U ≥ 60 wt %)
IEU Intermediate or mixed enrichment uranium (60 wt % > U >
10 wt %)
LEU Low enriched, natural, or depleted uranium ( U ≤ 10 wt %)
PU Plutonium
MIX U233
Mixed uranium and plutonium Uranium U systems Physical form of fissile material
MET Metal
SOL Solution
COMP Compound system, e.g. lattice in water Spectrum
FAST Fast system (≥50% of fissions above 100 keV) THERM Thermal system (≥50% of fissions below 0.625 eV)
Results of criticality benchmark calculations
In this section, we report the keff results of criticality benchmark using the MCNPX code. All results are given in graphical and in tabular form (see Appendix A). The following section contains:
i) The benchmark calculation values for keff and its uncertainty in pcm.
ii) The C/E values for keff (for the sake of clarity we use the term C/E value for keff, where C/E stands for the ratio of calculated-to-expected values) and their uncertainties using the following libraries: JEFF-3.3, ENDF/B-VIII.0, JEFF-3.2 and ENDF/B-VII.1.
iii) The χ and 〈|∆|〉 metrics, as described in (Cornock, 2014), that are used for getting a better read on data library performance.
The results of the calculations using JEFF-3.3 and ENDF/B-VIII.0 libraries have been compared with the published results (Nuclear Energy Agency, 2010), and with those obtained with JEFF-3.2 and ENDF/B-VII.1. Each calculation has used a total average of 3000 active generations of 10000 histories per generation. The results from the first 50 generations were excluded from the statistics.
M AN US CR IP T
AC CE PT ED
Subsequently, the reported results for each case is based on 29500000 active neutron histories. A statistical uncertainty between 0.0001 and 0.0003 is obtained.
Fig.2. results and their uncertainties for the HEU benchmarks.
Fig.3. results and their uncertainties for the IEU benchmarks.
For the large majority of the cases studied, the criticality benchmarks calculations illustrate a good performance for the chosen latest libraries. As shown in Figs 2, 3, 4, 5 and 6, the keff accuracy is quite dramatic. Seventy-eight (accounted for 65%) and sixty (50%) of the one-hundred-twenty values for keff have been improved for ENDF/B-VIII.0 and JEFF-3.3 libraries against ENDF/B- VII.1 and JEFF-3.2. For the rest of the cases, the two studied libraries produce values for keff, which are not significantly different from those produced by JEFF-3.2 and ENDF/B-VII.1 and the following remarks can be underlined:
For the HEU category, the computations have been made for 40 critical assemblies and the results are qualitatively good as demonstrated in Fig.2. With ENDF/B-VIII.0 cross-sections, the worst case result is obtained for “heu-met-fast-003-case11”, the calculated eigenvalue is too large by almost
M AN US CR IP T
AC CE PT ED
860pcm, for the JEFF-3.3 library a deviation of 980pcm was obtained. Nonetheless, these results are still better than those calculated with ENDF/B-VII.1 and JEFF-3.2.
Fig.4. results and their uncertainties for the LEU benchmarks.
Fig.5. results and their uncertainties for the Pu benchmarks.
For both ENDF/B-VIII.0 and JEFF-3.3 IEU system category (15 calculated cases), it is worthy to note that the overall keff C/E results are very good which the average deviations are 55pcm and 36 pcm for ENDF/B-VIII.0 and JEFF-3.3 cross-sections, respectively (Fig.3).
The C/E keff results for LEU case, which is considered as an important benchmark category (10 studied cases); indicate that seven of the 10 assemblies are reduced with ENDF/B-VIII.0 and five of the 10 assemblies for the JEFF-3.3 library (Fig.4). The worst result was reported for “leu-sol-therm- 001” both libraries.
In the Pu category (37 cases), the performance of the new ENDF/B-VIII.0 and JEFF-3.3 libraries versus the performance of the ENDF/B-VII.1 and JEFF-3.2 demonstrate that 19 calculated C/E keff’s values for these experiments are improved for the two reviewed libraries. The worst values
M AN US CR IP T
AC CE PT ED
were observed for “mix-met-fast-008” and “pu-sol-therm-009-case3a” cases, by 2212pcm and 1280pcm, respectively (Fig.5).
The results for 233U benchmarks (18 cases) show that keff values obtained with JEFF-3.3 nuclear data are higher than those obtained with ENDF/B-VIII.0. However, the keff for 6 assemblies among 18 are reduced from JEFF-3.3 to JEFF-3.2. It is worth mentioning that peculiar result has been deduced for benchmarks containing Beryllium (Be) such as for “U233-sol-inter-001-case1”. The calculated eigenvalue, as shown in Fig.6, is too small by almost 1414pcm and 1779pcm for ENDF/B-VIII.0 and JEFF-3.3, respectively, but they still better than ENDF/B-VII.1 and JEFF-3.2.
(Ouadie et al., 2017; Mosteller, 2014) have reported that the beryllium is performing worse when it is used as a reflector. Therefore, it is strongly recommended that their cross-sections be reviewed.
As demonstrated in tables 2 and 3, it should be noted that the ENDF/B-VIII.0 produces excellent agreement with the experimental for the majority of cases than JEFF-3.3 with the exception of a few benchmark cases, whereas for the rest of the cases a similar behavior between both libraries is highlighted.
Fig.6. results and their uncertainties for the 233U benchmarks.
Table 2.The χ between calculated and benchmark keff.
ENDF/B-VII.1 JEFF-3.2 JEFF-3.3 ENDF/B-VIII.0
All benchmarks 3.424 3.863 2.692 2.491
HEU benchmarks 4.535 6.140 2.344 2.533
IEU benchmarks 2.590 2.088 1.732 1.340
LEU benchmarks 2.549 2.826 2.787 2.586
Pu benchmarks 3.166 3.641 3.974 3.310
233U benchmarks 2.665 1.316 1.575 1.623
Table 3.The average difference (〈|∆|〉) in pcm between calculated and benchmark keff. ENDF/B-VII.1 JEFF-3.2 JEFF-3.3 ENDF/B-VIII.0 All benchmarks 280.968 294.209 283.524 245.591 HEU benchmarks 306.925 335.850 296.500 253.050 IEU benchmarks 172.000 170.067 167.867 144.133 LEU benchmarks 284.600 299.500 305.100 273.700
M AN US CR IP T
AC CE PT ED
Pu benchmarks 325.976 335.759 327.024 298.997
233U benchmarks 219.556 216.778 249.667 188.167
The χ and 〈|∆|〉 metrics can be calculated using the following formulates(Cornock, 2014):
χ = ((k − k )/δk )²
n (1)
〈|∆|〉 = "#$%&$− #'()"
* (2)
3.2. TRIGA Mark II research reactor calculation
3.2.1. The MCNPX model description of the TRIGA research reactor and simulation scenario The TRIGA Mark II research reactor installed at Maamora Center of Nuclear Studies (CENM), in Morocco, is a research reactor operated at 2000 kW. It is a pool type reactor cooled and moderated by light water. The fuel is composed of a mixture of uranium (8.5% wt., enriched at 19.7% with
235U), zirconium hydride and encapsulated in a stainless steel cladding. The reactor is controlled by five control rods containing boron carbide. The TRIGA core consists of 101 fuel elements and 17 graphite elements (Fig.7). Further description of TRIGA can be found in (IAEA, 2016;Boulaich et al., 2011). The model for these components were realized trying to preserve as far as possible all the characteristics related to the geometry, dimensions, and compositions. The MCNPX input was prepared (Fig.8) in such a way that a very quick setup of any desired core configuration with an adequate position of all control rods is possible. The MCNPX calculations were run with 50000000 active histories. A total of 50000 histories per generation were used and 1050 generations of neutrons. The first 50 generations were skipped to obtain a well-distributed neutron source. To account for thermal neutron scattering and low-energy neutron interactions, the appropriate s(α, β) treatments were used for water, zirconium hydride and the crystalline graphite for all the studied libraries.
M AN US CR IP T
AC CE PT ED
Fig.7.Technical characteristics (a) and the present core configuration (b) of the TRIGA Mark II research reactor (Boulaich et al., 2011).
M AN US CR IP T
AC CE PT ED
Fig.8. Side (a) and Top (b) view of the MCNPX computational model of the TRIGA Mark II research reactor.
3.2.2. Neutronic parameters calculation results
The calculation results of all the neutronic parameters including the effective multiplication factor, heat distribution, neutron flux distribution, of the TRIGA Mark II reactor core using MCNPX code and both data libraries JEFF-3.3 and ENDF/B-VIII.0 are summarized and investigated in the following sections.
Effective multiplication factor (keff)
The calculations of the keff eigenvalue were performed in the core for three cases; the critical position, core in the operating position and control rods that are totally withdrawn. The comparison between the MCNPX and the experimental results are shown in Table 4. The calculated ( − 1) value of keff for the control rods at critical positions (Huda et al., 2004) were found to be overestimated by 474pcm and 120pcm for JEFF-3.3 and ENDF/B-VIII.0, respectively. For the operating position (El Bakkari et al., 2013), the ( − 1) value of keff were found also to be overestimated by 578pcm for JEFF-3.3 and 254pcm for ENDF/B-VIII.0.However, underestimated values of 51pcm and 341pcm for all control rods in withdrawn positions (Chham et al., 2016) have deduced for JEFF-3.3 and ENDF/B-VIII.0, respectively. It can also be observed that the calculated keff using JEFF-3.3 are systematically higher than this using ENDF/B-VIII.0 except for rods were completely withdrawn position.
Table 4. Comparison between the experiment and calculated keff at different control rod positions using different cross- section libraries.
Control rods position
Core multiplication factor. keff
Reference Calculation
ENDF/B-VII.1 JEFF-3.2 JEFF-3.3 ENDF/B-VIII.0
Critical 1.00 0.99964 ± 0.00010 1.00395 ± 0.00011 1.00474 ± 0.00010 1.00120 ± 0.00010 Operating
position 1.04957 1.05062 ± 0.00010 1.05475 ± 0.00011 1.05564 ± 0.00010 1.05224 ± 0.00011 Withdrawn
(full-out- condition)
1.07902 1.07382 ± 0.00010 1.07763 ± 0.00010 1.07847 ± 0.00010 1.07533 ± 0.00011 Control rod
worth (|ρ|) in - 6910,5 ± 13,7 6810,3 ± 14,4 6804,3 ± 13,6 6885,4 ± 14,3
M AN US CR IP T
AC CE PT ED
pcm
Effect of graphite isotope density over TRIGA multiplication factors
The density of the graphite elements is not directly reported on the reactor documentation (the composition of the reflector is poorly documented). The value we used, equal to 1.65g cm⁄ , was found in Huda et al., 2004. However, other sources report different values, with the density ranging from 1.72g cm⁄ (Matsumoto, 1998) to 1.75g cm⁄ (Engle, 1977).In order to investigate the effects of nuclear graphite density on the core effective multiplication factor at different control rod positions, the PERT card was used (X-5 Monte Carlo Team, 2003).This card allows perturbations in cell material density. This latter estimated without actually changing the input material specifications. The calculation findings are presented in Table 5.
Table 5. Effect of Graphite density over the calculated keff at different control rod positions using different cross-section libraries.
Control rods position
Graphite density (g cm⁄ 3)
Core multiplication factor Calculation
ENDF/B-VII.1 JEFF-3.2 JEFF-3.3 ENDF/B-VIII.0
Critical 1.75 1.00427 ± 0.00016 1.00829 ± 0.00017 1.00921 ± 0.00016 1.00566 ± 0.00017 1.65 0.99964 ± 0.00010 1.00395 ± 0.00011 1.00474 ± 0.00010 1.00120 ± 0.00010 Operating
position
1.75 1.05496 ± 0.00016 1.05909 ± 0.00017 1.06000 ± 0.00010 1.05676 ± 0.00016 1.65 1.05062 ± 0.00010 1.05475 ± 0.00011 1.05564 ± 0.00010 1.05224 ± 0.00011 Withdrawn
(full-out- condition)
1.75 1.07802 ± 0.00017 1.08189 ± 0.00017 1.08303 ± 0.00016 1.07954 ± 0.00017 1.65 1.07382 ± 0.00010 1.07763 ± 0.00010 1.07847 ± 0.00010 1.07533 ± 0.00011
M AN US CR IP T
AC CE PT ED
Fig.9. Comparison of the neutron flux spectra obtained from different graphite densities using ENDF/B-VIII.0 and JEFF-3.3 libraries.
The initial calculated keff at different control rod positions predicted by the model were compared well with the reference values (Table 4). For the perturbation case as demonstrated in Table 5, the analysis of the simulation results with the control rods in different positions highlights that the core multiplication factors increase appreciably about (0.5%) for all studied libraries. This increase can be explained by the fact that an increase in graphite density may allow more energetic neutrons to be better scattered which resulting a rise in keff values. Nevertheless, as shown in Fig.9, no clear changes were exhibited by the comparison of the neutron flux spectra obtained with both graphite densities. The maximum neutron flux differences were found to be 0.13% in the thermal region, 0.3% in the epi-thermal region and 0.43% in the fast region for both ENDF/B-VIII.0 and JEFF-3.3
M AN US CR IP T
AC CE PT ED
libraries. It can be concluded, therefore, that the graphite density change (from 1.65g cm⁄ to 1.75g cm⁄ ) does not significantly affect the neutron spectrum.
Neutron flux and power distributions analysis
For the neutron flux calculations, a comparison was made between the JEFF-3.3 and ENDF/B- VIII.0 (Fig. 10). It is found that the MCNPX predicted values for both libraries have almost the same values as those in the reference core (Chham et al., 2016), and we observe that the flux evaluations by JEFF-3.3 are systematically slightly higher (about 1% on average) than those by ENDF/B-VIII.0. The profile of neutron flux was calculated using the tally F4: N (X-5 Monte Carlo Team, 2003). This tally enables the calculation of the flux average over a cell (particles/cm2). The MCNPX results are given by particle source, the Monte Carlo source normalisation factor F4 was estimated as follows:
Φ 3neutron cm . s : =
ν< =neutronfission @ ∗ pCWE
ϵ = MeVfission@ ∗ 1.602. 10LM = JMeV@ ∗ k OO
∗ ΦPQ3 1
cm : (3)
Fig.10. Comparison between the measured radial flux using different cross-section libraries.
The plots in Figs. 11 and 12 show the power distributions in the core of TRIGA Mark-II by comparing the calculated results to the two studied libraries. In this study, the heat generated in each of the fuel elements is most likely due to two sources: non-fission neutron reactions (mainly elastic and inelastic scattering) and neutron-induced fission reactions. For tallying these results, F6: n tally was used (track length estimator of fission energy deposition) and normalized to the full-power steady state operating condition.
M AN US CR IP T
AC CE PT ED
It is worth mentioning that the fuel elements in the middle rings generate more power than those in the external rings of the core. The maximum power produced in the hottest fuel element at full- power steady state was found to be 29.31kW, 29.35kW, 29.20kW and 29.42kW in B3 fuel element for ENDF/B-VIII.0, JEFF-3.3, ENDF/B-VII.1, and JEFF-3.2, correspondingly. In general, there is good agreement between the power distribution values evaluated by both ENDF/B-VIII.0 and JEFF-3.3. We may see that the differences between them were observed to be 2% in the core center and around 3% in the boundaries of the core.
Fig.11.Comparison between the measured radial power distribution using different cross-section libraries.
M AN US CR IP T
AC CE PT ED
Fig.12.Radial power distribution within the fuel and fuel follower produced using ENDF/B-VIII.0 and JEFF-3.3 libraries.
3.2.3. Kinetics parameters results
The principal analyzed kinetics parameters are effective delayed neutron fraction (βeff), prompt removal lifetime (τr), and mean neutron generation time (Ʌ). They can determine the time- dependent behavior of reactor power after reactivity insertion. In research reactors, these parameters are usually provided by the manufacturer (Boulaich et al., 2011).
M AN US CR IP T
AC CE PT ED
Effective delayed neutron fraction (βeff)
The effective delayed neutron fraction βeff is may be calculated by both MCNP and MCNPX codes using the prompt method, which requires two calculations with and without the TOTNU NO card (X-5 Monte Carlo Team, 2003). The TOTNU NO card was used in order to calculate the effective multiplication factor taking into account just the prompt neutron contribution.
β OO≅ 1 −SST
UVV (4)
The effect of nuclear data library on the calculated value of βeff was investigated for TRIGA MARK II at the critical position. The results are summarised in Tables 6 and 7. We notice that clearly the results of ν, νp, νd and βeff are strongly depending on the nuclear data library used.
However, we can see that the calculated values are close between (ENDF/B-VII.1, ENDF/B-VIII.0) and (JEFF-3.2, JEFF-3.3) and as expected, it is highly likely due to the original ENDF data evaluation process. It can be seen likewise that the values of ν, νp, νd and βeff are slightly closer to the reference values (Boulaich et al., 2011). Anyway, both libraries agreed with a maximum relative error smaller than 3% with the reference value.
Table 6. Predicted values of keff, kp, kd, and βeff of various nuclear data libraries.
Reference ENDF/B-VII.1 JEFF-3.2 JEFF-3.3 ENDF/B-VIII.0
keff - 0.99964 ± 0.00010 1.00395 ± 0.00011 1.00474 ± 0.00010 1.00120 ± 0.00010 kp - 0.99241 ± 0.00011 0.99643 ± 0.00010 0.99724 ± 0.00011 0.99396 ± 0.00010 kd - 0.00723 ± 0.00021 0.00752 ± 0.00021 0.00750 ± 0.00021 0.00724 ± 0.00020 βeff 0.00730 0.00723 ± 0.00021 0.00749 ± 0.00021 0.00746 ± 0.00021 0.00723 ± 0.00020
Table 7. The comparison of the estimated values of average number of neutrons released per fission (ν), prompt neutron (νp) and delayed neutron yields (νd).
Reference(Keepin 1965) ENDF/B-VII.1 JEFF-3.2 JEFF-3.3 ENDF/B-VIII.0
ν 2.4300 2.4386 2.4272 2.4273 2.4314
νp 2.4142 2.4226 2.4108 2.4110 2.4155
νd 0.0158 0.0159 0.0164 0.0164 0.0159
Calculation of the mean neutron generation time (Ʌ).
The second important kinetic parameter that characterizes the time behavior of neutron population is the mean neutron generation time. Using the prompt neutron lifetime τr, which is the average time from the emission of a prompt neutron in fission to the removal of the neutron by some physical process such as fission, capture, or escape, and keff, Ʌ parameter can be determined as demonstrated in equation 4. The difference between the mean neutron generation time Ʌ and neutron prompt neutron lifetime τr is that Ʌ only takes into account the neutron absorptions inducing fission.
Λ =SXY
UVV. (5)
For Ʌ the reported value of mean neutron generation time for 8.5w/o fuel range from 49.2 to 54.8µs (Hassanzadeh and Feghhi, 2013), while our computed values are 52.250 µ s and 52.623 µs utilizing JEFF-3.3 and ENDF/B-VIII.0 data libraries, respectively (Table 8). The maximum difference value
M AN US CR IP T
AC CE PT ED
between the reference value and JEFF-3.3 is 2%. On the contrary, it can, therefore, be concluded that the mean neutron generation time Ʌ depends slightly on the nuclear data library used.
Table 8. The comparison of the calculated values of τr and Ʌbetween the libraries and also with the reference (Boulaich et al., 2011).
Kinetic
parameters Reference Calculation
ENDF/B-VII.1 JEFF-3.2 JEFF-3.3 ENDF/B-VIII.0
τr (µs) - 53.315 ± 0.39020 52.195 ± 0.38781 52.250 ± 0.40307 52.623 ± 0.37948 Λ (µs) 53.000 53.335 ± 0.44369 51.990 ± 0.44325 52.003 ± 0.45293 52.560 ± 0.43152
4. Conclusion
A comparative study has been performed with the newly continuous energy cross-section data JEFF-3.3 and ENDF/B-VIII.0, generated for MCNPX using the NJOY21 data processing system.
The correlation has carried out against some critically benchmarks lattices and TRIGA Mark II research reactor.
Based on criticality calculations using both libraries, the agreement overall between experimentally measured and computed values of keff is very good for the benchmark calculations performed here.
However, the largest deviations (more than 1000 pcm) were seen for “leu-sol-therm-001”, “mix- met-fast-008”, “pu-sol-therm-009-case3a” and “u233-sol-inter-001-case1”. Moreover, the beryllium is in particular considered as a problematic issue for different benchmark example for “u233-sol- inter-001-case1”. For this reason, we strongly recommended that the cross-sections for beryllium could be assessed in the future releases. It should be noted that ENDF/B-VIII.0 produces excellent agreement with the experimental for the majority of cases than JEFF-3.3 with the exception of a few benchmark cases, whereas the rest of the cases show a similar behavior between both libraries.
For TRIGA Mark II reactor, the objective of this study is to investigate the predictive capability the recent nuclear data libraries JEFF-3.3 and ENDF/B-VIII.0 by analyzing both neutronic and kinetic parameters, and then to compare the accuracy of the predicted results with those of the published values. For the effective multiplication factors, the analysis of the simulation results with the control rods in different positions highlights that the absolute value of ( − 1) is 374pcm using the JEFF- 3.3 cross-section library, and is 341pcm using the ENDF/B-VIII.0. In general, the calculation results for JEFF-3.3 and ENDF/B-VIII.0 data libraries were found to be very encouraging. During the comparison calculation, it is worth pointing out that the performance of JEFF-3.3 and ENDF/B- VIII.0 data libraries is quite satisfying. However, their influence could be qualitatively assessed.
Acknowledgments
The authors wish to thank Nuclear Reactor and Nuclear Security Group Energy Center members (Mohamed V University, Morocco) for their support, helpful suggestions and discussions.
References
Bakkari, B. El, T. El Bardouni, B Nacir, C. El Younoussi, Y Boulaich, H Boukhal, and M Zoubair. 2013. “Fuel Burnup Analysis for the Moroccan TRIGA Research Reactor.” Annals of Nuclear Energy 51 (January): 112–19.
https://doi.org/10.1016/j.anucene.2012.07.030.
Boulaich, Y., B. Nacir, T. El Bardouni, M. Zoubair, B. El Bakkari, O. Merroun, C. El Younoussi, A. Htet, H. Boukhal, and E. Chakir. 2011. “Steady-State Thermal-Hydraulic Analysis of the Moroccan TRIGA MARK II Reactor by
M AN US CR IP T
AC CE PT ED
Using PARET/ANL and COOLOD-N2 Codes.” Nuclear Engineering and Design 241 (1): 270–73.
https://doi.org/10.1016/j.nucengdes.2010.10.033.
Brown, D. A., M. B. Chadwick, R. Capote, A. C. Kahler, A. Trkov, M. W. Herman, A. A. Sonzogni, et al. 2018.
“ENDF/B-VIII.0: The 8th Major Release of the Nuclear Reaction Data Library with CIELO-Project Cross Sections, New Standards and Thermal Scattering Data.” Nuclear Data Sheets 148: 1–142.
https://doi.org/10.1016/j.nds.2018.02.001.
Cabellos, O., F. Alvarez-Velarde, M. Angelone, C.J. Diez, J. Dyrda, L. Fiorito, U. Fischer, et al. 2017. “Benchmarking and Validation Activities within JEFF Project.” EPJ Web of Conferences 146: 06004.
https://doi.org/10.1051/epjconf/201714606004.
Cabellos, Oscar. 2014. “Processing of the New JEFF-3.2T4 Cross Section Library with the NJOY Code System into Various Formats for Testing Purposes,” no. February.
Carlson, A. D., V. G. Pronyaev, R. Capote, G. M. Hale, Z. P. Chen, I. Duran, F. J. Hambsch, et al. 2018. “Evaluation of the Neutron Data Standards.” Nuclear Data Sheets 148: 143–88. https://doi.org/10.1016/j.nds.2018.02.002.
Chadwick, M. B., M. Herman, P. Obložinský, M. E. Dunn, Y. Danon, A. C. Kahler, D. L. Smith, et al. 2011. “ENDF/B- VII.1 Nuclear Data for Science and Technology: Cross Sections, Covariances, Fission Product Yields and Decay Data.” Nuclear Data Sheets 112 (12): 2887–2996. https://doi.org/10.1016/j.nds.2011.11.002.
Chham, E., T. El Bardouni, K. Benaalilou, H. Boukhal, B. El Bakkari, Y. Boulaich, C. El Younoussi, and B. Nacir.
2016. “Neutronic and Thermal–hydraulic Analysis of New Irradiation Channels inside the Moroccan TRIGA Mark II Research Reactor Core.” Applied Radiation and Isotopes 116: 178–84.
https://doi.org/10.1016/j.apradiso.2016.08.006.
Cornock, M. 2014. “Benchmarking of ENDF/B-VII.1 and JENDL-4.0 in the Fast Neutron Range.” Nuclear Data Sheets 118 (1): 430–32. https://doi.org/10.1016/j.nds.2014.04.098.
Engle, G B. 1977. “Assessment of Grade H-451 Graphite for Replaceable Fuel,” no. 17. https://doi.org/GA-Al4690.
Hassanzadeh, M., and S. A.H. Feghhi. 2013. “Analysis of Burn up Effects on Kinetic Parameters in an Accelerator Driven Subcritical TRIGA Reactor.” Annals of Nuclear Energy 62: 280–83.
https://doi.org/10.1016/j.anucene.2013.05.047.
Huda, M. Q., M. Rahman, M. M. Sarker, and S. I. Bhuiyan. 2004. “Benchmark Analysis of the TRIGA MARK II Research Reactor Using Monte Carlo Techniques.” Annals of Nuclear Energy 31 (11): 1299–1313.
https://doi.org/10.1016/j.anucene.2004.02.005.
IAEA. 2016. “History, Development and Future of TRIGA Research Reactors.” Technical Reports Series No. 482, no.
482.
International Atomic Energy Agency. 2018. “ENDF-Archive.” Available from: https://www- nds.iaea.org/public/download-endf/.
Keepin, G. Robert. 1965. “Physics of Nuclear Kinetics.” In Addison-Wesley Series in Nuclear Science and Engineering, 161–86. Addison-Wesley Pub. Co; 1st edition (1965).
Liang, Yongmei, Yongxia Jia, and Yaofang Liu. 2008. “Using Micro-Reactor and Gas Chromatograph Combined Apparatus to Study the Effect of LBO-A Co-Catalyst on Fcc Naphtha Upgrading.” Petroleum Processing and Petrochemicals 39 (11): 1–4.
Los Alamos National Laboratory. 2018. ““NJOY21 for the 21st Century.” Available from:
https://github.com/njoy/NJOY21.git.
MacFarlane, R. E., and A. C. Kahler. 2010. “Methods for Processing ENDF/B-VII with NJOY.” Nuclear Data Sheets 111 (12): 2739–2890. https://doi.org/10.1016/j.nds.2010.11.001.
Macfarlane, Robert, Douglas W. Muir, R. M. Boicourt, Albert Comstock Kahler, III, and Jeremy Lloyd Conlin. 2017.
“The NJOY Nuclear Data Processing System, Version 2016.” https://doi.org/10.2172/1338791.
Marshall, W J, and B T Rearden. 2008. Criticality Safety Validation of SCALE 6.1 with ENDF/B-VII.0 Libraries. Vol.
2013.
Matsumoto, Tetsuo. 1998. “Benchmark Analysis of Criticality Experiments in the TRIGA Mark II Using a Continuous Energy Monte Carlo Code MCNP.” Journal of Nuclear Science and Technology 35 (9): 662–70.
https://doi.org/10.1080/18811248.1998.9733922.
M AN US CR IP T
AC CE PT ED
Mosteller, R. 2014. “Comparison of ENDF/B-VII.1 and ENDF/B-VII.0 Results for the Expanded Criticality Validation Suite for MCNP and for Selected Additional Criticality Benchmarks.” Nuclear Data Sheets 118 (1): 442–45.
https://doi.org/10.1016/j.nds.2014.04.101.
Nuclear Energy Agency. 2010. International Handbook of Evaluated Criticality Safety Benchmark Experiments. Edited by Nuclear Energy Agency. 2010th ed.
Ouadie, Kabach, Chetaine Abdelouahed, Jalil Abdelhamid, Darif Abdelaziz, and Saidi Abdelmajid. 2017. “Processing and Benchmarking of Evaluated Nuclear Data File/b-Viii.0β4 Cross-Section Library by Analysis of a Series of Critical Experimental Benchmark Using the Monte Carlo Code MCNP(X) and NJOY2016.” Nuclear Engineering and Technology 49 (8): 1610–16. https://doi.org/10.1016/j.net.2017.08.017.
Pelowitz, Denise B. 2008. “MCNPX: User’s Manual, v2. 6.0.”
Trkov, A., M. Herman, and D. A. Brown. 2018. “ENDF-6 Formats Manual Data,Data Formats and Procedures for the Evaluated Nuclear Data Files ENDF/B-VI, ENDF/B-VII and ENDF/B-VIII.” www.nndc.bnl.gov.
X-5 Monte Carlo Team. 2003. “MCNP - Version 5, Vol. 1: Overview and Theory.” La-Ur-03-1987.
Appendix A
Tabulated below are the computed eigenvalues of the ICSBEP Benchmarks discussed in this paper.
These results were obtained after processing ENDF-6 cross-section libraries into the appropriate application suitable for transport codes MCNPX.
Table A1. MCNP results for HEU Benchmarks.
ICSBEP
Benchmark name Benchmark keff ENDF/B-VII.1 JEFF-3.2 JEFF-3.3 ENDF/B-VIII.0
heu-comp-inter-003-case7 1.00000 ±0.00470 0.99703 ±0.00014 0.99734 ±0.00014 1.00286 ±0.00014 0.99983 ±0.00014 heu-met-fast-001 1.00000 ±0.00100 1.00015 ±0.00011 0.99982 ±0.00011 1.00016 ±0.00011 0.99995 ±0.00011 heu-met-fast-002-case1 1.00000 ±0.00300 0.99987 ±0.00011 0.99998 ±0.00011 1.00014 ±0.00011 1.00001 ±0.00011 heu-met-fast-003-case1 1.00000 ±0.00500 0.99506 ±0.00011 0.99645 ±0.00011 0.99575 ±0.00011 0.99294 ±0.00011 heu-met-fast-003-case2 1.00000 ±0.00500 0.99433 ±0.00011 0.99593 ±0.00011 0.99524 ±0.00012 0.99191 ±0.00012 heu-met-fast-003-case3 1.00000 ±0.00500 0.99921 ±0.00012 1.00083 ±0.00012 0.99985 ±0.00011 0.99682 ±0.00011 heu-met-fast-003-case4 1.00000 ±0.00300 0.99716 ±0.00012 0.99866 ±0.00012 0.99805 ±0.00012 0.99490 ±0.00011 heu-met-fast-003-case5 1.00000 ±0.00300 1.00140 ±0.00012 1.00353 ±0.00012 1.00280 ±0.00012 0.99945 ±0.00012 heu-met-fast-003-case6 1.00000 ±0.00300 1.00164 ±0.00012 1.00370 ±0.00012 1.00306 ±0.00012 0.99947 ±0.00012 heu-met-fast-003-case7 1.00000 ±0.00300 1.00193 ±0.00012 1.00414 ±0.00012 1.00382 ±0.00012 1.00037 ±0.00012 heu-met-fast-003-case8 1.00000 ±0.00500 1.00153 ±0.00012 1.00315 ±0.00012 1.00144 ±0.00012 1.00050 ±0.00012 heu-met-fast-003-case9 1.00000 ±0.00500 1.00178 ±0.00012 1.00424 ±0.00012 1.00179 ±0.00012 1.00078 ±0.00012 heu-met-fast-003-case10 1.00000 ±0.00500 1.00531 ±0.00012 1.00845 ±0.00012 1.00531 ±0.00012 1.00396 ±0.00012 heu-met-fast-003-case11 1.00000 ±0.00500 1.01016 ±0.00012 1.01317 ±0.00012 1.00980 ±0.00012 1.00863 ±0.00012 heu-met-fast-003-case12 1.00000 ±0.00300 1.00859 ±0.00012 1.00542 ±0.00012 1.00535 ±0.00012 0.99965 ±0.00012 heu-met-fast-004-case1 1.00020 ±0.00100 1.00334 ±0.00014 1.00192 ±0.00014 1.00115 ±0.00014 1.00184 ±0.00014 heu-met-fast-008 0.99890 ±0.00160 0.99590 ±0.00015 0.99584 ±0.00016 0.99551 ±0.00015 0.99590 ±0.00015 heu-met-fast-009-case1 0.99920 ±0.00150 0.99730 ±0.00021 0.99715 ±0.00020 0.99644 ±0.00020 0.99636 ±0.00021 heu-met-fast-009-case2 0.99920 ±0.00150 0.99678 ±0.00012 0.99629 ±0.00011 0.99518 ±0.00012 0.99533 ±0.00011 heu-met-fast-011 0.99890 ±0.00150 0.99874 ±0.00014 0.99859 ±0.00014 0.99800 ±0.00014 0.99718 ±0.00014 heu-met-fast-012 0.99920 ±0.00180 0.99819 ±0.00011 0.99838 ±0.00011 0.99820 ±0.00011 0.99805 ±0.00011 heu-met-fast-013 0.99900 ±0.00150 0.99750 ±0.00011 0.99569 ±0.00011 0.99569 ±0.00011 0.99867 ±0.00011 heu-met-fast-014 0.99890 ±0.00170 0.99763 ±0.00011 0.99874 ±0.00011 0.99826 ±0.00011 0.99540 ±0.00011 heu-met-fast-015 0.99960 ±0.00170 0.99452 ±0.00011 0.99446 ±0.00011 0.99448 ±0.00011 0.99485 ±0.00011 heu-met-fast-018-case2 1.00000 ±0.00140 0.99940 ±0.00011 0.99970 ±0.00011 0.99991 ±0.00011 0.99954 ±0.00011 heu-met-fast-019-case1 0.99920 ±0.00150 1.00707 ±0.00011 1.00709 ±0.00011 1.00658 ±0.00011 1.00581 ±0.00012
M AN US CR IP T
AC CE PT ED
heu-met-fast-020-case2 1.00000 ±0.00280 1.00051 ±0.00012 1.00086 ±0.00012 1.00081 ±0.00012 1.00016 ±0.00012 heu-met-fast-021-case2 1.00000 ±0.00240 0.99752 ±0.00012 0.99691 ±0.00012 0.99642 ±0.00011 0.99977 ±0.00011 heu-met-fast-022-case2 1.00000 ±0.00190 0.99783 ±0.00011 0.99753 ±0.00011 0.99715 ±0.00012 0.99713 ±0.00011 heu-met-fast-026-case11 0.99820 ±0.00420 1.00377 ±0.00014 1.00317 ±0.00014 1.00209 ±0.00014 1.00214 ±0.00014 heu-met-fast-028 1.00000 ±0.00300 1.00257 ±0.00012 1.00486 ±0.00012 1.00433 ±0.00012 1.00073 ±0.00012 heu-met-inter-006-case-1 0.99770 ±0.00080 0.99335 ±0.00014 0.99472 ±0.00014 0.99605 ±0.00014 0.99599 ±0.00014 heu-met-inter-006-case-2 0.99970 ±0.00080 0.99685 ±0.00014 0.99904 ±0.00014 0.99944 ±0.00014 0.99954 ±0.00014 heu-met-inter-006-case-3 1.00150 ±0.00090 1.00068 ±0.00013 1.00322 ±0.00013 1.00020 ±0.00013 1.00316 ±0.00014 heu-met-inter-006-case-4 1.00160 ±0.00080 1.00739 ±0.00013 1.01102 ±0.00013 1.00136 ±0.00013 1.00541 ±0.00014 heu-sol-therm-013-case1 1.00120 ±0.00260 0.99893 ±0.00010 0.99928 ±0.00010 0.99807 ±0.00010 0.99926 ±0.00010 heu-sol-therm-013-case2 1.00070 ±0.00360 0.99789 ±0.00011 0.99833 ±0.00011 0.99736 ±0.00011 0.99836 ±0.00011 heu-sol-therm-013-case3 1.00090 ±0.00360 0.99426 ±0.00012 0.99438 ±0.00012 0.99371 ±0.00012 0.99484 ±0.00012 heu-sol-therm-013-case4 1.00030 ±0.00360 0.99612 ±0.00012 0.99605 ±0.00012 0.99550 ±0.00012 0.99653 ±0.00012 heu-sol-therm-032 1.00150 ±0.00260 0.99952 ±0.00007 0.99937 ±0.00007 0.99747 ±0.00007 0.99886 ±0.00007
Table A2. MCNP results for IEU Benchmarks.
ICSBEP
Benchmark name Benchmark keff ENDF/B-VII.1 JEFF-3.2 JEFF-3.3 ENDF/B-VIII.0
ieu-met-fast-001-case1 0.99885 ±0.00090 1.00033 ±0.00011 1.00017 ±0.00011 1.00007 ±0.00011 0.99919 ±0.00011 ieu-met-fast-001-case2 0.99970 ±0.00100 1.00061 ±0.00011 1.00040 ±0.00011 1.00025 ±0.00011 0.99917 ±0.00011 ieu-met-fast-001-case3 0.99930 ±0.00050 1.00096 ±0.00011 1.00076 ±0.00011 0.99989 ±0.00011 0.99880 ±0.00011 ieu-met-fast-001-case4 1.00020 ±0.00050 1.00149 ±0.00011 1.00085 ±0.00011 1.00034 ±0.00011 0.99948 ±0.00011 ieu-met-fast-002 1.00000 ±0.00300 0.99898 ±0.00010 0.99764 ±0.00010 0.99625 ±0.00010 0.99601 ±0.00010 ieu-met-fast-003-case2 1.00000 ±0.00170 1.00247 ±0.00011 1.00203 ±0.00011 1.00131 ±0.00011 1.00005 ±0.00011 ieu-met-fast-004-case2 1.00000 ±0.00300 1.00762 ±0.00012 1.00655 ±0.00012 1.00550 ±0.00012 1.00509 ±0.00011 ieu-met-fast-005-case2 1.00000 ±0.00210 1.00173 ±0.00011 1.00086 ±0.00011 0.99990 ±0.00011 1.00091 ±0.00011 ieu-met-fast-006-case2 1.00000 ±0.00230 0.99651 ±0.00011 0.99546 ±0.00011 0.99399 ±0.00011 0.99370 ±0.00011 ieu-met-fast-007-case4 1.00450 ±0.00070 1.00451 ±0.00009 1.00463 ±0.00009 1.00484 ±0.00009 1.00455 ±0.00010 leu-sol-therm-007-case14 0.99610 ±0.00090 0.99513 ±0.00013 0.99535 ±0.00013 0.99527 ±0.00013 0.99573 ±0.00013 leu-sol-therm-007-case30 0.99730 ±0.00090 0.99713 ±0.00013 0.99774 ±0.00012 0.99754 ±0.00012 0.99740 ±0.00012 leu-sol-therm-007-case32 0.99850 ±0.00100 0.99646 ±0.00012 0.99632 ±0.00012 0.99586 ±0.00012 0.99657 ±0.00012 leu-sol-therm-007-case36 0.99880 ±0.00110 0.99872 ±0.00011 0.99884 ±0.00011 0.99825 ±0.00011 0.99890 ±0.00011 leu-sol-therm-007-case49 0.99830 ±0.00110 0.99744 ±0.00011 0.99680 ±0.00011 0.99689 ±0.00011 0.99766 ±0.00011
Table A3 .MCNP results for LEU Benchmarks.
ICSBEP
Benchmark name Benchmark keff ENDF/B-VII.1 JEFF-3.2 JEFF-3.3 ENDF/B-VIII.0
leu-comp-therm-008-case1 1.00070 ±0.00120 1.00009 ±0.00030 0.99974 ±0.00031 1.00125 ±0.00030 1.00068 ±0.00030 leu-comp-therm-008-case2 1.00070 ±0.00120 1.00114 ±0.00030 1.00071 ±0.00029 1.00098 ±0.00028 1.00129 ±0.00030 leu-comp-therm-008-case5 1.00070 ±0.00120 1.00021 ±0.00031 0.99992 ±0.00029 1.00033 ±0.00030 1.00091 ±0.00030 leu-comp-therm-008-case7 1.00070 ±0.00120 1.00108 ±0.00031 0.99924 ±0.00030 1.00032 ±0.00028 1.00079 ±0.00030 leu-comp-therm-008-case8 1.00070 ±0.00120 1.00039 ±0.00029 0.99932 ±0.00030 1.00029 ±0.00029 0.99952 ±0.00032 leu-comp-therm-008-case11 1.00070 ±0.00120 1.00166 ±0.00029 1.00063 ±0.00029 1.00077 ±0.00029 1.00123 ±0.00030 leu-sol-therm-001 0.99910 ±0.00290 1.01214 ±0.00014 1.01229 ±0.00015 1.01269 ±0.00015 1.01234 ±0.00014
