Effect of Pu-Rich Agglomerate in MOX Fuel on a Lattice Calculation

The effect of Pu-rich agglomerates in U-Pu mixed oxide (MOX) fuel on a lattice calculation has been demonstrated. The Pu-rich agglomerate parameters are defined based on the measurement data of MIMAS-MOX and the focus is on the highly enriched MOX fuel in accordance with increased burnup resulting in a higher volume fraction of the Pu-rich agglomerates. The lattice calculations with a heterogeneous fuel model and a homogeneous fuel model are performed simulating the PWR 17 × 17 fuel assembly. The heterogeneous model individually treats the Pu-rich agglomerate and U-Pu matrix, whereas the homogeneous model homogenizes the compositions within the fuel pellet. A continuous-energy Monte Carlo burnup code, MVP-BURN, is used for burnup calculations up to 70 GWd/t. A statistical geometry model is applied in modeling a large number of Pu-rich agglomerates assuming that they are distributed randomly within the MOX fuel pellet. The calculated nuclear characteristics include k-inf, Pu isotopic compositions, power density and burnup of the Pu-rich agglomerates, as well as the pellet-averaged Pu compositions as a function of burnup. It is shown that the effect of Pu-rich agglomerates on the lattice calculation is negligibly small.     

Irradiation of a thorium–plutonium rodlet: Experiment and benchmark calculations

A benchmark exercise for thorium–plutonium fuel, based on experimental data, has been carried out. A thorium–plutonium oxide fuel rodlet was irradiated in a PWR for four consecutive cycles, to a burnup of about 37 MWd/kgHM. During the irradiation, the rodlet was inserted into a guide tube of a standard MOX fuel assembly. After the irradiation, the rod was subjected to several PIE measurements, including radiochemical analysis. Element concentrations and radial distributions in the rodlet, multiplication factors and distributions within the carrier assembly of burnup and power were calculated. Four participants in the study simulated the irradiation of the MOX fuel assemblies including the thorium–plutonium rodlet using their respective code systems; MCBurn, HELIOS, CASMO-5 and ECCO/ERANOS combined with TRAIN. The results of the simulations and the measured results of the radiochemical analysis were compared and found to be in fairly good agreement when the calculated results were calibrated to give the same burnup of the thorium–plutonium rodlet as that experimentally measured. Average concentrations of several minor actinides and fission products were well reproduced by all codes, to the extent that can be expected based on known uncertainties in the experimental setup and the cross section libraries. Calculated results which could not be confirmed by experimental measurement were compared and only two significant anomalies were found, which can probably be addressed by limited modifications of the codes.     

Analysis of MOX Fuel Behavior in Halden Reactor by FEMAXI-6 Code

Capabilities of the FEMAXI-6 code to analyze the behavior of high burnup MOX fuels in LWRs have been evaluated. Coolant conditions, detailed power histories and specifications of the MIMAS-MOX fuel rods, rod 10 and rod 11, of IFA-597.4–7 irradiated in the Halden reactor were input, and calculated rod internal pressures and pellet center temperatures were compared with the measured data for the range of 0-31 MWd/kgUO₂. Some sensitivity studies were conducted mainly with respect to pellet thermal conductivity and swelling rate to investigate the changes in thermal behavior and their effects on fission gas release.

In the irradiation period up to about 23 MWd/kgUO₂, the calculated pellet center temperatures sufficiently agreed with the measured data and also the calculated rod internal pressures reproduced the tendency of an increase in the measured rod internal pressures. These results suggest that fission gas release from MOX fuels can be reasonably predicted by a diffusion process that is modeled in UO₂ pellet grains. On the other hand, the steep increase in the measured rod internal pressures observed at the power ramp around 23 MWd/kgUO₂ cannot be reproduced by FEMAXI-6 and can be regarded as the result of a relatively large amount of gas release, which possibly caused a pellet-cladding-gap closure through pellet gas-bubble swelling.     

Thermal conductivity evaluation of high burnup mixed-oxide (MOX) fuel pellet

The thermal conductivity formula of fuel pellet which contains the effects of burnup and plutonium (Pu) addition was proposed based on the Klemens’ theory and reported thermal conductivities of unirradiated (U, Pu) O₂ and irradiated UO₂ pellets. The thermal conductivity of high burnup MOX pellet was formulated by applying a summation rule between phonon scattering parameters which show the effects of plutonium addition and burnup. Temperature of high burnup MOX fuel was evaluated based on the thermal conductivity integral which was calculated from the above-mentioned thermal conductivity formula. Calculated fuel temperatures were plotted against the linear heat rates of the fuel rods, and were compared with the fuel temperatures measured in a test reactor. Since both values agreed well, it was confirmed that the proposed thermal conductivity formula of MOX pellets is adequate.     

An Apparatus for Continuously Measuring Radon-222 Exhalation from Ground

High burnup MOX and UO₂ test rods were prepared from the fuel rods irradiated in commercial BWRs. Each test rod was equipped with a fuel center thermocouple and reirradiated in the Halden boiling water reactor (HBWR) in Norway. The burnups of MOX and UO₂ test rods reached about 84GWd/tHM and 72GWd/t, respectively. Fuel temperature was measured continuously during the re-irradiation tests. Thermal conductivity change in high burnup fuel was evaluated from the results of comparison between the measured fuel temperature and the data calculated by using the fuel analysis code FEMAXI-6. The comparison results suggested that the thermal conductivity of MOX fuel pellets is comparable to that of UO₂ fuel pellets in the high burnup region around 80 GWd/t. It is probable that the impurity effect of Pu atoms gradually diminishes with increasing burnup because other factors that affect pellet thermal conductivity, such as the accumulation effect of soluble fission products and irradiation-induced defects in crystal lattice, become dominant in a high burnup region.     

Thermal diffusivity of homogeneous SBR MOX fuel with a burn-up of 35 MWd/kgHM

The effect of burn-up on the thermal conductivity of homogeneous SBR MOX fuel is investigated and compared with standard UO₂ LWR fuel. New thermal diffusivity results obtained on SBR MOX fuel with a pellet burn-up of 35 MWd/kgHM are reported. The thermal diffusivity measurements were carried out at three radial positions using a shielded “laser-flash” device and show that the thermal diffusivity increases from the pellet periphery to the centre. The fuel thermal conductivity was found to be in the same range as for UO₂ of similar burn-up. The annealing behaviour was characterized in order to identify the degradation due to the out-of-pile auto-irradiation.     

Analysis of measured isotopic compositions of high-burnup PWR MOX and UO2 fuels in the MALIBU program

The measured isotopic compositions of fuel samples taken from high-burnup spent PWR MOX and UO₂ assemblies in the MALIBU program has been analyzed by lattice physics codes. The measured isotopes were U, Np, Pu, Am, and Cm isotopes and about 30 major fission product nuclides. The codes used in the present study were a continuous-energy Monte Carlo burnup calculation code (MVP-BURN) and a deterministic burnup calculation code (SRAC) based on the collision probability method. A two-dimensional multi-assembly geometrical model (2 × 2 model) was mainly adopted in the analysis in order to include the fuel assemblies adjoining the relevant fuel assembly, from which the samples were taken. For the MOX sample, the 2 × 2 model significantly reduces the deviations of the calculated results from the measurements compared with a single assembly model. The calculation results of MVP-BURN in the 2 × 2 model reproduce the measurements of U, Np, and Pu isotopes within 5% for the MOX sample of 67 GWd/t. The deviations of their calculated results of U, Np, and Pu isotopes from the measurements are less than 7% for the UO₂ sample of 72 GWd/t.     

Extension of the TRANSURANUS burn-up model

The validation range of the model in the TRANSURANUS fuel performance code for calculating the radial power density and burn-up in UO₂ fuel has been extended from 64 MWd/kgHM up to 102 MWd/kgHM, thereby improving also its precision. In addition, the first verification of calculations with post-irradiation examination data is reported for LWR-MOX fuel with a rod average burn-up up to 45 MWd/kgHM. The extension covers the inclusion of new isotopes in order to account for the production of ²³⁸Pu. The corresponding one-group cross-sections used in the equations rely on results obtained with ALEPH, a new Monte Carlo burn-up code. The experimental verification is based on electron probe microanalysis (EPMA) and on secondary ion mass spectrometry (SIMS) as well as radiochemical data of fuel irradiated in commercial power plants. The deviations are quantified in terms of frequency distributions of the relative errors. The relative errors on the burn-up distributions in both fuel types remain below 12%, corresponding to the experimental scatter.     

Analysis of Measured Isotopic Compositions of High-Burnup BWR MOX Fuel

Analysis of measured isotopic compositions of four high-burnup BWR MOX fuel samples was performed by using a general-purpose neutronic calculation code SRAC and a continuous-energy Monte Carlo burnup code MVP-BURN. The initial Pu fissile content of the samples was 5.52 wt%, and the burnups ranged from 50 to 80 GWd/t. It is confirmed that a geometrical model including the effect of UO₂ assemblies adjacent to the MOX assembly is necessary in the burnup calculations to obtain accurate calculated isotopic compositions. The calculated results of MVP-BURN with JENDL-3.3 taking such effect into account show more accurate results for major actinides (U, Pu, and Am isotopes) and most fission products than those of infinite assembly calculations. The paper also shows the results calculated using SRAC with JENDL-3.3, ENDF/B-VII, and JEFF-3.1.     

Behavior of Irradiated ATR/MOX Fuel under Reactivity Initiated Accident Conditions

Pulse irradiation experiments with irradiated ATR/MOX fuel rods of 20MWd/kgHM were conducted at the NSRR in Japan Atomic Energy Research Institute to study the transient behavior of MOX fuel rod under reactivity initiated accident conditions. Four pulse irradiation experiments were performed with peak fuel enthalpy ranging from 335 J/g to 586 J/g, resulted in no failure of fuel rods. Relatively large radial deformation of the fuel rods due to pellet-cladding mechanical interaction occurred in the experiments with peak fuel enthalpy above 500 J/g. Significant fission gas release up to 20% was measured by rod puncture measurement. The generation of fine radial cracks in pellet periphery, micro-cracks and boundary separation over the entire region of pellet were observed. These microstructure changes might contribute to the swelling of fuel pellets during the pulse irradiation. This could cause the large radial deformation of fuel rod and high fission gas release when the pulse irradiation conducted at relatively high peak fuel enthalpy. In addition, fine grain structures around the plutonium spot and cauliflower structure in cavity of the plutonium spot were observed in the outer region of the fuel pellet.     

Validation of a new continuous Monte Carlo burnup code using a Mox fuel assembly

The reactivity of nuclear fuel decreases with irradiation (or burnup) due to the transformation of heavy nuclides and the formation of fission products. Burnup credit studies aim at accounting for fuel irradiation in criticality studies of the nuclear fuel cycle (transport, storage, etc…). The principal objective of this study is to evaluate the potential capabilities of a newly developed burnup code called “BUCAL1”. BUCAL1 differs in comparison with other burnup codes as it does not use the calculated neutron flux as input to other computer codes to generate the nuclide inventory for the next time step. Instead, BUCAL1 directly uses the neutron reaction tally information generated by MCNP for each nuclide of interest to determine the new nuclides inventory. This allows the full capabilities of MCNP to be incorporated into the calculation and a more accurate and robust analysis to be performed. Validation of BUCAL1 was processed by code-to-code comparisons using predictions of several codes from the NEA/OCED. Infinite multiplication factors (k_∞) and important fission product and actinide concentrations were compared for a MOX core benchmark exercise. Results of calculations are analysed and discussed.     

Thermal conductivity degradation analyses of LWR MOX fuel by the quasi-two phase material model

The temperature measurements of mixed oxide (MOX) and UO₂ fuels during irradiation suggested that the thermal conductivity degradation rate of the MOX fuel with burnup should be slower than that of the UO₂ fuel. In order to explain the difference of the degradation rates, the quasi-two phase material model is proposed to assess the thermal conductivity degradation of the MIMAS MOX fuel, which takes into account the Pu agglomerate distributions in the MOX fuel matrix as fabricated. As a result, the quasi-two phase model calculation shows the gradual increase of the difference with burnup and may expect more than 10% higher thermal conductivity values around 75 GWd/t. While these results are not fully suitable for thermal conductivity degradation models implemented by some industrial fuel manufacturers, they are consistent with the results from the irradiation tests and indicate that the inhomogeneity of Pu content in the MOX fuel can be one of the major reasons for the moderation of the thermal conductivity degradation of the MOX fuel.     

Post-irradiation examinations focused on fuel integrity of spent BWR-MOX and PWR-UO2 fuels stored for 20 years

Post-irradiation examinations (PIEs) of spent BWR-MOX and PWR-UO₂ fuel rods irradiated in commercial LWRs and stored for 20 years were carried out to evaluate fuel integrity during storage. Average burn-up of five fuel rods of the BWR-MOX fuel was about 20 MWd/kgHM and that of the PWR-UO₂ fuel was 58 MWd/kgHM. The PIE items included: (a) visual inspection of the cladding surface, (b) puncture test, (c) ceramographic observation on the pellet and cladding, (d) pellet density, (e) electron probe microanalysis of the pellet, (f) cladding tensile test, (g) hydrogen content and hydride orientation in the cladding, and (h) hydrogen redistribution in the cladding under temperature gradient. The PIE results showed no marked difference in the visual inspection, fission gas release, oxide layer thickness, pellet microstructure, and cladding mechanical properties or hydride orientation after storage. The result of the hydrogen redistribution experiment showed that hydrogen migration had little effect on the fuel integrity during dry storage. Hydrogen migration on the fuel rod for 40 years of storage was evaluated using the heat of transport obtained in the hydrogen redistribution experiment and calculated result showed that hydrogen migration had little effect on the fuel integrity during dry storage.     

Performance of TRISO particles fueled with plutonium and minor actinides in a PBMR-400 core design

The TRISO particle design of high temperature reactors fueled with plutonium (Pu) and/or minor actinides (MAs) is investigated by calculating the failure fraction of TRISO particles during irradiation. For this purpose, a fuel depletion, neutronics and thermal-hydraulics code system, which delivers the fuel temperature, fast neutron flux and power density profiles, is coupled to an analytical stress analysis code. The latter is being further developed for the calculation of a reliable and realistic failure fraction. The code system has been applied to a PBMR-400 design containing TRISO particles fueled with 1st and 2nd generation plutonium and with a target burn-up of 700 and 600 MWd/kgHM, respectively. It is shown that the pebble-bed type high temperature reactor under consideration is a promising option for burning Pu and MAs if very high burn-ups can be achieved. The TRISO particle failure fraction is also calculated for both Pu and MA fuels, and compared to U-based fuel. It is shown by the present stress analysis code that the Pu-based fuel particles need a better design and this has been achieved for the MA-based fuel, in which helium gas atoms have a significant contribution to the buffer pressure.     

Determination of fissile fraction in MOX (mixed U + Pu oxides) fuels for different burnup values

When spent Light Water Reactor fuels are processed by the standard Purex method of reprocessing, plutonium (Pu) and uranium (U) in spent fuel are obtained as pure and separate streams. The recovered Pu has a fissile content (consisting of ²³⁹Pu and ²⁴¹Pu) greater than 60% typically (although it mainly depends on discharge burnup of spent fuel). The recovered Pu can be recycled as mixed-oxide (MOX) fuel after being blended with a fertile U makeup in a MOX fabrication plant. The burnup that can be obtained from MOX fuel depends on: (1) isotopic composition of Pu, which is closely related to the discharge burnup of spent fuel from which Pu is recovered; (2) the type of fertile U makeup material used (depleted U, natural U, or recovered U); and (3) fraction of makeup material in the mix (blending ratio), which in turn determines the total fissile fraction of MOX. Using the Non-linear Reactivity Model and the code MONTEBURNS, a step-by-step procedure for computing the total fissile content of MOX is introduced. As was intended, the resulting expression is simple enough for quick/hand calculations of total fissile content of MOX required to reach a desired burnup for a given discharge burnup of spent fuel and for a specified fertile U makeup. In any case, due to non-fissile (parasitic) content of recovered Pu, a greater fissile fraction in MOX than that in fresh U is required to obtain the same burnup as can be obtained by the fresh U fuel.     

Comparison of thorium-based fuels with different fissile components in existing boiling water reactors

With the aim of investigating the technical feasibility of fuelling a conventional BWR (Boiling Water Reactor) with thorium-based fuel, computer simulations were carried out in a 2D infinite lattice model using CASMO-5. Four different fissile components were each homogenously combined with thorium to form mixed oxide pellets: Uranium enriched to 20% U-235 (LEU), plutonium recovered from spent LWR fuel (RGPu), pure U-233 and a mixture of RGPu and uranium recovered from spent thorium-based fuel. Based on these fuel types, four BWR nuclear fuel assembly designs were formed, using a conventional assembly geometry (GE14-N). The fissile content was chosen to give a total energy release equivalent to that of a UOX fuel bundle reaching a discharge burnup of about 55 MWd/kgHM. The radial distribution of fissile material was optimized to achieve low bundle internal radial power peaking. Reactor physical parameters were computed, and the results were compared to those of reference LEU and MOX bundle designs. It was concluded that a viable thorium-based BWR nuclear fuel assembly design, based on any of the fissile components, can be achieved. Neutronic parameters that are essential for reactor safety, like reactivity coefficients and control rod worths, are in most cases similar to those of LEU and MOX fuel. This is also true for the decay heat produced in irradiated fuel. However when Th is mixed with U-233, the void coefficient (calculated in 2D) can be positive under some conditions. It was concluded that it is very difficult to make savings of natural uranium by mixing LEU (20% U-235) homogenously with thorium and that mixing RGPu with thorium leads to more efficient consumption of Pu compared to MOX fuel.     

Use of the local Pu-239 concentration as an indicator of burnup spectral history in DYN3D

Codes for reactor core calculations use few-group cross sections (XS) which depend on local burnup, given in terms of the energy produced per fuel mass (MWd/kgHM). However, a certain burnup value can be reached under different spectral conditions depending on moderator density and other local parameters. Neglecting these spectral effects, i.e. applying the summary-burnup value only, can cause considerable errors in the calculated power density.     

Spent fuel interrogation using delayed fast neutron spectrum at Missouri University of Science and Technology Reactor

Interrogation of nuclear fuel and Plutonium (Pu) and Uranium (U) discrimination was performed using Missouri University of Science and Technology Reactor (MSTR) fuel by non-destructive (NDA) method. Post-irradiated delayed fast neutron spectra were obtained for two pairs of burnt and fresh fuels. Burnup and ²³⁹Pu conversions were calculated based on neutron emission intensity ratios. After 100 kW high power runs, all fuel elements showed three distinct regions of neutron spectra; a distinct low energy peak followed by intermediate energy region without distinct peak but a wide hump, followed by a high energy peak with a long tail. At 10 kW low powers, intermediate energy hump and low energy peak seems to merge together while the high energy peak still remains distinct. Based on data from 10 kW power runs, the burnup values of F1 and F2 fuel elements were estimated to be 149 MWD/T and 196 MWD/T, respectively. ²³⁹Pu conversion since 1992 for low enriched (19.75%) fuel elements was calculated as 0.24 g for F1 and 0.32 g for F2. Results based on high power runs of 100 kW provided comparable burnup of 217 MWD/T for F2. However the results for F1 were approximately 10 times higher perhaps due to unique burnup history and consequently high poison buildup. These experimental burnup results compare well with the reactor burnup calculation as reported to the Nuclear Regulatory Commission (NRC).     

Analysis of high burnup fuel behavior in Halden reactor by FEMAXI–V code

The author developed a code FEMAXI–V to analyze the behaviors of high burnup LWR fuels. FEMAXI–V succeeded the basic structure of code FEMAXI–IV, and incorporated such new models and functions as fuel thermal conductivity degradation with burnup, alliance with burnup analysis code which gives radial power profile and fast neutron flux, etc. In the present analysis, coolant conditions, detailed power histories and specifications of the fuel rods DH and DK of IFA-519.9 irradiated in Halden reactor were input, and calculated rod internal pressures were compared with experimental data for the range of 25–93 MWd kg⁻¹ UO₂, and factors affecting pellet temperature were discussed. Also some sensitivity studies were conducted with respect to the effect of swelling rate and grain growth. As a result, it is found that the prediction is sensitive to the models of thermal conductivity and swelling rate of fuel, and FEMAXI–V analytical system proved to give a reasonable prediction even in the high burnup region.     

Analysis of Post Irradiation Experiments in PWR Using New Nuclear Data Libraries

For the precise calculation of the burnup of minor actinide isotopes, a code system-SWAT has been developed. This system analyzes burnup problems with neutron spectrum that depends on the type of a reactor and the irradiation history, using latest evaluated nuclear data files JENDL-3 or ENDF/B-Vl. The post irradiation test in TRINO and the recent experiment in typical PWRs in Japan were analyzed with SWAT. These analyses show that the results of U and Pu for high burnup fuels almost agree with experimental results but those for middle burnup fuels do not agree with them. The results for Am and Cm isotopes still have large discrepancy. The average C/E of ²⁴³Am is –0.79, and that of ²⁴⁴Cm is –0.70 for high burnup (–33,000 MWd/tU) samples.

For middle burnup (–25,000 MWd/tU) samples, the C/E for ²⁴⁴Cm is over 2.0. The discrepancy is partially explained by considering the power peaking history of first cycle and second cycle.