18th Symposium of AER on VVER Reactor Physics and Reactor Safety

Paks Unit 1 finished it?s 25th cycle in 2008. In the article different reload strategies and applied fuel on this unit are listed. Cycles recalculated using present reload design tools and safety limits checked using present safety limits. Correlation between advanced fuel utilization and margins to safety limits are analyzed.

The paper discusses a well-known condition [95%/95%], which is important for monitoring some limits of core parameters in the course of designing the reactors (such as PWR or VVER). The condition ensures the postulate ?there is at least a 95 % probability at a 95 % confidence level that? some parameter does not exceed the limit. Such conditions are stated, for instance, in US standards and IAEA norms as recommendations for DNBR and fuel temperature.

The present methodology of safety margins in NPP Dukovany design power distribution calculations is based on the philosophy of engineering factors with errors defined on the bases of statistical approach of standard (95%) confidence intervals. On the level of FA power distribution the normality (normal density distribution) of this approach is tested and comparison with errors defined on the 95-percent probability at a 95-percent confidence level (shortly in statistics 95%/95%)) is provided. Practical applications are presented for several NPP Dukovany fuel cycles. The paper also deals briefly with difference between confidence interval and tolerance interval, with the problems of density distribution of mechanical engineering factor variables and solution of axial and radial error distribution like bivariate problem.

Until recently, computer code certificates indicated no fuel burnup calculation errors. In the same time, these errors should be taken into account while checking the discharged fuel burnup limit, as well as while checking the local linear heat rate and local linear heat rate ramp limits depending on fuel burnup. This paper intends to connect the fuel burnup calculation error with the calculation error related to fuel power distribution for the contemporary VVER fuel cycle. This paper also attempts to analyze the burnup calculation error in the framework of random functions theory, on the basis of power density correlation function. This paper also contains examples of estimated errors for burnup calculations performed for characteristic VVER fuel loads.

Joint AER WG A and WG B held their seventeenth meeting in Modra-Harm?nia (near NPP Jaslovsk? Bohunice), Slovak Republic, during the period of 14-15 April 2008.
The objectives of the meeting content of presentations and future activities are shortly described in this paper.

ANDREA is a new macrocode for core design and safety evaluation, which is being developed at NRI. The core simulator is based on the nodal expansion method. Standard features such as pinwise power reconstruction and feedback iterations on critical control rod position, boron concentration and reactor power are implemented. Thermomechanical model of the fuel rod is based on FEMAXI-V calculations and the predictions were benchmarked against HRP experimental data. A branch of the code, designed for the calculation of reactivity induced transients, is currently being developed, based on GAKIN method.

The paper deals with detailed and accurate modeling of the WWER-440/V213 reactor type and precise criticality calculation using the MCNP5 transport code.

A new ?so called second generation- fuel type developed by Russian vendor [TVEL] for VVER-440 has been introduced. Even the outer parameters of the assembly remain unchanged, the length of the pins was increased and 6 of the 126 pins have been doped by gadolinium. The above mentioned modifications requiring more accurate calculations have necessitated the further development and validation of our KARATE code system. The paper summarizes the capabilities of the new version of the KARATE program concerning the properties of currently used and newly developed fuel types. On the basis of an international benchmark published from Dukovany NPP validation procedure has been started.

A new three-dimensional nodal model for neutronics calculation is currently under implementation into APROS – Advanced PROcess Simulation environment – to conform the increasing accuracy requirements. The new model is based on an advanced nodal code HEXTRAN and its static version HEXBU-3D by VTT, Technical Research Centre of Finland. Currently the new APROS is under a testing programme. Later a systematic validation will be performed. In the first phase, a goal is to obtain a fully validated model for VVER-440 calculations. Thus, all the current test calculations are performed by using Loviisa NPP?s VVER-440 model of APROS. In future, the model is planned to be applied for the calculations of VVER-1000 type reactors as well as in rectangular fuel geometry. The paper outlines first the general aspects of the method, and then the current situation of the implementation. Because of the identical model with the models of HEXTRAN and HEXBU-3D, the results in the test calculations are compared to the results of those. In the paper, results of two static test calculations are shown. Currently the model works well already in static analyses. Only minor problems with the control assemblies of VVER-440 type reactor still exist but the reasons are known and will be corrected in near future. Dynamical characteristics of the model are up to now tested only by some empirical tests.
?
INTRODUCTION

One of the limiting factors in terms of nuclear power plant lifetime is reactor pressure vessel condition. Its radiation embrittlement causes especially fast neutrons. We are focused on describing methodology of the neutron flux density, inhere in reactor VVER-440/V-213 pressure vessel area with simulation in code MCNP 4C2 in this work. Computer simulation of selected neutron-physical parameters will be verified for a consideration of the data from ionization chamber, received from mode department of nuclear power plant in Bohunice and Mochovce. Condition of reactor pressure vessel is usually monitored by surveillance specimen program. Between the irradiation channel which contain surveillance specimen and the reactor pressure vessel is large area containing water with boric acid and structural steel parts. This environment influence density of neutron flux in pressure vessel area and energy spectrum of neutrons. The goal of this work is improve assessment of reactor VVER-440/V-213 pressure vessel radiation degradation and following valuation possibility of its lifetime.

Major advances have been made in the HELIOS code, resulting in the impending release of a new version, HELIOS-2. The new code includes a method of characteristics (MOC) transport solver to supplement the existing collision probabilities (CP) solver. A 177-group, ENDF/B-VII nuclear data library has been developed for inclusion with the new code package. Computational tests have been performed to verify the performance of the MOC solver against the CP solver, and validation testing against computational and measured benchmarks is underway. Results to-date of the verification and validation testing are presented, demonstrating the excellent performance of the new transport solver and nuclear data library.

!
#$

The transport method of HELIOS is called the CCCP method, because it is based on current coupling and collision probabilities. The system to be calculated consists of space elements that are coupled with each other and with the boundaries by interface currents. The angular dependence of the interface or coupling currents can be discretized in various ways. This is done by partitioning the directional half-sphere into a number of polar ? levels, and each ? level into a number of azimuthal ? intervals. In two dimensional calculations the discretization of the azimuthal level is dominant. In the last issue of the HELIOS code (version 1.10) the maximum value of azimuthal discretization is increased from 4 to 12. This gives the possibility to calculate large (core-wide or near core-wide) systems with appropriate accuracy, which extends the applicability of the HELIOS program.

The critical experiments performed at the Hungarian ZR-6 reactor and experiments
performed at General Purpose P (GPP) critical facility at RRC ?Kurchatov Institute? are used
to benchmark HELIOS-2, as part of its ongoing validation and verification activities. The
comparisons presented in this paper are based on, ZR6-VVER-EXP-001, Vol. 1 (2007), LEUCOMP-THERM-015, Vol. 4 (2005), and LEU-COMP-THERM-061, Vol. 4 (2002).
On ZR-6, single cell, macro cell, and 2D calculations on selected experiments, regular and
perturbed, are made. In the 2D calculations, the radial leakage is treated by including the
reflector in the calculations, representing only axial leakage by a measured axial buckling.
keff and RMS reaction rates comparisons are presented.
On GPP critical facility, the entire experiment is modelled as 2D. Comparisons of keff and
fission rates are presented and the effect of the axial buckling on keff is investigated.

A?depletion calculation benchmark devoted to MOX fuel cycles in an ongoing objective of OECD/NEA WPRS (Working Party on Scientific Issues of Reactor Systems) following the study of depletion calculation concerning UOX fuel ¹. Four corresponding fuel structures as super-cells were calculated: the MOX PWR, WWER-1000, WWER-440 assembly surrounded by UOX assembly and MOX assembly in an infinite lattice. These scenarios enable to show the importance of the UOX environment in the mixed UOX – MOX core. Studies have shown that a?standard calculation based on an infinite medium pattern for MOX assembly can lead to a significant error in Pu-239 concentration at a high burn up. The model calculations were performed by a spectral code HELIOS 1.10.

!

The results of tests of the mode related to switching on one main coolant pump (MCP) when the other three MCPs are in operation that are obtained during commissioning work at ?Kozloduy? NPP, Unit 6, Bulgaria, and given in specification OECD/NEA VVER-1000 (V1000 CT-1) were used for calculations using the coupled thermohydraulic computer codes that are applied in OKB ?Gidropress? and include models of spatial kinetics and inter-loop mixing in the reactor.

The paper briefly describes the Specification of an international NEA/OECD benchmark based on measured plant data.
During the commissioning tests for nominal power at NPP Kalinin Unit #3 a lot of measurements of neutron and thermo-hydraulic parameters have been carried out in the reactor pressure vessel, primary and the secondary circuits. One of the measured data sets for the transient ?Switching-off of one Main Circulation Pump (MCP) at nominal power? has been chosen to be applied for validation of coupled thermal-hydraulic and neutron-kinetic system codes and additionally for performing of uncertainty analyses as a part of the NEA/OECD Uncertainty Analysis in Modeling (UAM) Benchmark.
The benchmark is opened for all countries and institutions. The experimental data and the final specification with the cross section libraries will be provided to the participants from NEA/OECD only after official declaration of real participation in the benchmark and delivery of the simulated results of the transient for comparison.
INTRODUCTION

The ATHLET-BIPR-VVER coupled system code is applied for performing of safety analysis for different VVER reactors. During the last years its validation matrix is continuously being enlarged. The measurements performed during the commissioning phase of NPP Kalinin Unit #3 for the transient ?Switching-off of one Main Circulation Pump (MCP) at nominal power? are very well documented and have a variety of recorded integral and local thermo-hydraulic and neutron-physic parameters including the measurements? errors. This data is being used for further validation of the coupled code system ATHLET-BIPR-VVER.
In the paper are discussed the problems and our solutions by the correct interpretation of the measured thermocouples? records at NPP Kalinin-3 and the comparison with the predicted results by the coupled thermal-hydraulic/neutron-kinetic code ATHLET-BIPR-VVER. Of primary importance by such comparisons is the correct accounting of the fluid mixing process that take place in the surrounding of the measuring sensors and also the consideration of the time delay (inertia term) of the measuring devices. On the bases of previous experience and many simulations of the defined transient a method is discussed and proposed to consider correctly the inertia term of the thermocouples? measurements. The new modelling is implemented in the coupled system code ATHLET-BIPR-VVER for further validation.

Processes that cause the scram actuation by period during WWER-1000 fast unloading in condition of application of new monitoring system AKNP-I are analyzed. Results of calculation investigation of reactor fast unloading, that have been occurred on Ukrainian NPP, and comparisons of results of modeling with experimental data are presented. Influence of such factors as dropping time of control rods, time moment of campaign, discretization step of period determination on the reactor fast unloading behavior are investigated.

The influence of the delayed neutrons on the reactor dynamics can be understood through their impact on the reactor power change rate. In spite of the fact that delayed neutrons constitute only a very small fraction of the total number of neutrons generated from fission, they play a dominant role in the fission chain reaction control. If only the prompt neutrons existed, the reactor operation would become impossible due to the fast reactor power changes. The exact determination of delayed neutrons main parameter, the delayed neutron fraction (?eff), is very important in the field of reactor physics. The interest in the delayed neutron data accuracy improvement started to increase at the end of 80-ties and the beginning of 90-ties, after discrepancies among the results of calculations and experiments. In consequence of difficulties in (eff experimental measurement, this value in exact state use to be determined by calculations. Subsequently, its reliability depends on the calculation method and the delayed neutron data used. Determination of (eff requires criticality calculations. In the past, keff used to be traditionally calculated by taking the ratio of the adjoint- and spectrum-weighted delayed neutron production rate to the adjoint- and spectrum- weighted total neutron production rate ¹. An alternative method has also been used in which ?eff is calculated from simple k-eigenvalue solutions.

CFD CODES AND THE ONSAGER RELATIONS
?
MIHALY MAKAI
Abstract. In the last decade, papers appeared in the literature discussing
a shortcoming of the basic equations of hydrodynamics: the Navier-Stokes
equations do not meet the Onsager symmetry relations. Recently R. Streater
wrote about the topic. The basic problem is that the solution to the Boltzmann equation f (r, v, t) depends on seven variable, nevertheless the solution
of the Navier-Stokes equation yields T (r, t), v(r, t) and ?(r, t)-the temperature, velocity and density distribution, altogether ?ve functions. Clearly, the
solution class of the Boltzmann equation is a broader class than the solution
class of the Navier-Stokes equation. What is the importance of that question?
Are the results of CFD codes questionable ones or, the contradiction can be
resolved by some ignored terms of second order?

Based on experiences of former validation and sensitivity studies, a CFD model for head part of real working fuel assemblies with pitch of 12.3 mm has been developed with the code ANSYS CFX. Calculations were performed for typical fuel assemblies used in the Paks NPP. Differences between the outlet average temperatures and thermocouple signals were determined. Effect of the mixing grid position on thermocouple signal was investigated also. Mixing was analyzed in details with using so-called mixing scalars and weight factors of the central tube and rod bundle regions for in-core thermocouple signal were determined. Sensitivity of the weight factors for pin power distribution and mixing grid position were investigated.

The final step in the safety analysis is the investigation of the fulfilment of the acceptance criteria using hot channel calculations. Recently, there has been under way at Paks NPP to introduce a new, higher enriched (4.2 %) fuel type containing Gd burnable poison. To do that, for some transients the DBA analyses must be repeated and last year, as one of the first steps in this process, it was needed to review the hot channel calculation methodologies used in the analyses.

The Budapest University of Technology and Economics, Institute of Nuclear Techniques (BUTE INT) has been working since 2001 on the three-dimensional CFD model of the reactor pressure vessel of the VVER-440 type reactor. During this time period ? due to the development of the available computational capacity ? a very complex and detailed model of the RPV has been developed.
The aim of the construction of the new model is to describe further internal structures of the RPV (e.g. correct modeling of brake tubes, or internals in the upper mixing chamber) and to perform an extensive sensitivity analysis on the different modeling and calculation parameters (e.g. porous region models vs. detailed modeling, or n different turbulence models).

The validation of depletion codes is an important task in spent fuel management, especially for burnup credit application in criticality safety analysis of spent fuel facilities. Because of lack of well documented experimental data for VVER-1000, the validation could be made on the basis of code intercomparison based on the numerical benchmark problems. Some years ago a VVER-1000 burnup credit benchmark (CB5) was proposed to the AER research community and the preliminary results from three depletion codes were compared. In the paper some new results for the isotopic concentrations of twelve actinides and fifteen fission products calculated by the depletion codes SCALE5.1, WIMS9, SCALE4.4 and NESSEL-NUKO are compared and evaluated.

A synthetic time dependent step-wise Fission Gas Release (FGR) algorithm – having only very few unknowns to be refreshed in each time step – is derived from the exact solution by using a Green function technique. The proposed method is easy to be built into the existing fuel behavior codes, and additionally gives the exact solution of the infinite series back with a high accuracy. Using this ?fix point? of the modeling, puncturing FGR measurements were used to clarify the mechanism of the further release from the inter-granular region to the free volume. The Halden diffusion coefficient proved satisfactory concerning the average measured FGR in case of a one step algorithm application. Nevertheless, a one step algorithm or a leaking two step method cannot give back the burnup dependence of the measured FGR. On the contrary, a two step algorithm with a saturation effect in the inter-granular region can lead to the measured burnup dependence, but only with an enhanced diffusion and increased inter-granular FGR.

The assembly VVER-440 type Gd-2 has radial profile of enrichment and has 6 pins with Gd2O3. The maximal enrichment is 4.4%. Some analyses are done for assembly with flat enrichment (4.4%) and without Gd2O3. In this article are compared some characteristics (decay heat, some nuclide concentration, photon and gamma sources) for real Gd-2 assembly and for flat 4.4% assembly. The TRITON module (in SCALE 5.1) was used.

Updated Results of Simpli?ed Benchmark Based on #2670
ISTC VVER PIE
?
Franti?ek Havluj, Nuclear Research Institute at Re?, Czech Republic, haf@nri.cz
?
18th AER Symposium on VVER Reactor Physics and Reactor Safety, 6-10 October
2008, Eger, Hungary

The analysis of spent nuclear fuel properties and isotopic compositions is a very important stage for neutron-physical calculations of VVER-type reactor cores and the subcriticality analysis of storage facilities. The reliable PIE measurements are necessary for testing composition calculations.

Database of publicly available VVER PIE data
?
Franti?ek Havluj, Nuclear Research Institute at Re?, Czech Republic, haf@nri.cz
?
18th AER Symposium on VVER Reactor Physics and Reactor Safety, 6-10 October
2008, Eger, Hungary

Ten institutions of nuclear industry, regulation and research from countries operating VVER reactors in the European Union established VVER PIE International Consortium to fund the radiochemical assays of VVER-440 fuel which should be performed in a hot cell laboratory in RIAR Dimitrovgrad, Russia. The measurements will be made on 12 samples cut out from 3 pins of VVER-440 fuel assembly with flat enrichment 4.4% 235U. The sample burnups lie in interval 30 – 60 MWd/kgU. The measured nuclides will be the ?burnup credit isotopes?, actinides and fission products.

The report deals with the results of simulating the experimental studies of the reactor plant behaviour and the core characteristics in the xenon transients caused by a small and large decrease in the reactor power and also a small increase thereof. The experimental studies were performed in the end of burn-up (T??? ? 244 eff. days) of the first fuel loading at Kalinin NPP, Unit 3. A comparative analysis of the testing data and the results of neutron- physical calculations was made with the use of code package SAPFIR_95 & RC.

In year 2006 we started data processing from the Dukovany NPP operating history database that contained data from the old measurement system VK3 and the new Scorpio-VVER. The work has been done in cooperation with the reactor physicists at Dukovany NPP. Obtained data from database were compared with calculated parameters from 3D diffusion macrocode Mobydick.

The SCORPIO-VVER core monitoring system has proved since the first installation at Dukovany NPP in 1999 to be a valuable tool for the reactor operators and reactor physicists. It is now installed on four units of Dukovany NPP and two units of Bohunice NPP replacing the original Russian VK3 system. By both Czech and Slovak nuclear regulatory bodies it was licensed as a Technical Specification Surveillance tool.

Starting from 1980s, the problem of bow of the VVER-1000 reactor FAs and the effect of that on the operational safety is being discussed. At the initial period of time, the extension of time for dropping control rods of the control and protection system associated with this bow posed the highest threat. Later on, new more rigid structures were developed for FAs that eliminated the problems of control rods.

18th Symposium of AER on VVER Reactor Physics and Reactor Safety,
Eger, Hungary, October 6-10, 2008

The High Performance Light Water Reactor (HPLWR) is the European version of the various supercritical water cooled reactor proposals. The paper presents the activity of KFKI-AEKI in the field of neutronic core design within the framework of the European funded ?HPLWR Phase 2? and the Hungarian sponsored NUKENERG projects. Preliminary loadings of the HPLWR core were assessed, which contain insulated assemblies with Gd absorbers. The initial reactivity excess was compensated by control rods. The reactivity coefficients of HPLWR with respect to fuel temperature and water density show similarity to VVER-440 at normal operating conditions but the power response of the core to reactivity changes is quite different.

This article compares two kinds of fuel cycles: one is the gadolinium UOX (4.25 % U-235 + 3.35 % Gd2O3) cycle in WWER-440 reactor and the other is the cycle in fast reactor – SUPER PHENIX loaded with MOX fuel (23.2 % Pu + 2.7 % Minor Actinides). The starting point is a single recycling of Pu from WWER-440 in FR. The next step is multirecycling of Pu in FR to obtain equilibrium cycle. The model of SUPER PHENIX includes a radial blanket. The results of weight concentration of WWER-440 are listed for the 1st cycle only. This papers is dealing with the solution of energy questions and also fuel cycle indicators attractive for the final disposal. All kinds of calculations were performed by computer code HELIOS 1.10.

In the paper, different advanced nuclear fuel cycles including thorium-based fuel and inert-matrix fuel are examined under light water reactor conditions, especially VVER-440, and compared. Two investigated thorium based fuels include one solely plutonium-thorium based fuel and the second one plutonium-thorium based fuel with initial uranium content. Both of them are used to carry and burn or transmute plutonium created in the classical UOX cycle. The inert-matrix fuel consist of plutonium and minor actinides separated from spent UOX fuel fixxed in yttria-stabilised zirconia matrix.

18th Symposium of AER on VVER Reactor Physics and Reactor Safety,

Review of AER Working Group B Meeting in Modra – Harm?nia, Slovakia is given.

The report shows comparisons of calculation results by code package KASKAD and experimental data for the last cycles of Kozloduy NPP Unit 5 and Unit 6. The results are related to:

Reactor power variation causes additional thermo-mechanical loads on fuel rod clads increasing with increase in a number of loading cycles and fuel burn-up fraction. Power distribution resulted from power variation and motion of control rods is of rather complicated character that can be correctly considered only in detailed (pin-by-pin) presentation of the core neutron physics characteristics.

This paper discusses the effect of feedback and boundary condition models used in the Loviisa core design code HEXBU-3D on the calculated reactor core characteristics. Two alternative feedback models and two boundary condition models are shortly demonstrated and their effects on the calculated reactor physical properties are considered. Computational results by these alternative models will also be compared with measured reactor physical properties of the recent Loviisa-1 and Loviisa-2 cores.

The WWER-440 control assemblies have significant influence on space power distribution (fission rate) and can cause power peaks in adjacent fuel assemblies. This is a consequence of a deficiency in design of joining structure of the absorber adapter to the fuel part of the control assembly, due to presence of water cavity in this space. Presence of the given cavity results in flash-up of thermal neutrons flux density in periphery fuel pins of the adjacent operating fuel assemblies, that is, power gradients can occur inside the mentioned fuel pins that could result in static and cyclic loads with some consequences, e.g. fuel pin bowing.

Results of Gd benchmark based on NPP Dukovany Gd fuel introduction is given. Calculated are kinetics parameters, reactivity coefficients, length of cycles and let-down curves for cycles.

Five years fuel cycle has been implemented in Dukovany NPP since the year 1999 including preparation of Fuel Supply Contract and licensing procedure. This paper describes main milestones of this process with regard to parameters of other projects realized in Dukovany NPP. Main fuel cycle parameters are presented and discussed. Second part of paper is dedicated to description of 5-years fuel cycle at new nominal power level 1444 MWth.