20th Symposium of AER on VVER Reactor Physics and Reactor Safety

The current Finnish national research programme on nuclear power plant safety SAFIR2010 for the years 20072010 as well as the coming SAFIR2014 programme for the years 2011-2014 are based on the chapter 7a,
?Ensuring expertise?, of the Finnish Nuclear Energy Act. The objective of this chapter is realised in the research
work and education of experts in the projects of these research programmes.
SAFIR2010 research programme is divided in eight research areas that are Organisation and human, Automation
and control room, Fuel and reactor physics, Thermal hydraulics, Severe accidents, Structural safety of reactor
circuit, Construction safety, and Probabilistic Safety Analysis (PSA). All the research areas include both projects
in their own area and interdisciplinary co-operational projects. Research projects of the programme are chosen
on the basis of annual call for proposals.
In 2010 research is carried out in 33 projects in SAFIR2010. VTT is the responsible research organisation in 26
of these projects and VTT is also the coordination unit of SAFIR2010 and SAFIR2014. In 2007-2009
SAFIR2010 produced 497 Specified research results (Deliverables), 618 Publications, and 33 Academic degrees.
SAFIR2010 programme covers approximately half of the reactor safety research volume in Finland currently. In
2010 the programme volume is ? 7.1 million and 47 person years. The major funding partners are VYR with ?
2.96 million, VTT with ? 2.66 million, Fortum with ? 0.28 million, TVO with ? 0.19 million, NKS with ? 0.15
million, EU with only ? 0.03 million and other partners with ? 0.85 million.
The new decisions-in-principle on Olkiluoto unit 4 for Teollisuuden Voima and new nuclear power plant for
Fennovoima ratified by the Finnish Parliament on 1 July 2010 increase the annual funding collected according to
the Finnish Nuclear Energy Act from Fennovoima, Fortum and Teollisuuden Voima for the SAFIR2014
programme to ? 5.2 million from the current level of ? 3.0 million in SAFIR2010. The anticipated annual
volume of SAFIR2014 is approximately ? 10 million. The increase in research needs and in training new
personnel is well recognised in the Framework Plan of SAFIR2014.

Joint AER Working Group A on ?Improvement, extension and validation of
parameterized few-group libraries for VVER-440 and VVER-1000? and AER Working
Group B on ?Core design? nineteenth meeting was hosted by V?JE a.s. in Modra – Harm?nia
(Slovakia) during the period of 20th to 22nd April 2010.
There were present altogether 12 participants from 8 member organizations and 9
papers were presented (8 of them in written form).
Objectives of the meeting of WG A are: Issues connected with spectral calculations
and few-groups libraries preparation, their accuracy and validation.
Presentations were devoted to some aspects of transport and diffusion calculations and
to the benchmark dealing with VVER-1000 core periphery power tilt.
Tam?s Park? (co-authors Istv?n P?s and S?ndor Patai Szab?) described in his
presentation ?Application of Discontinuity factors in C-PORCA 7 code?, Radoslav Zajac
(co-authors Petr Dalek and Vladim?r Ne?as) spoke about ?Fast Reactor Nodalisation in
HELIOS Code?, Gabriel Farkas presented ?Calculation of Spatial Weighting Functions of
Ex-Core Neutron Detectors for VVER-440 Using Monte Carlo Approach? and Daniel
Sprinzl (co-authors V?clav Kr?sl, Pavel Mikol and Ji?? ?varn?) provided a definition of a
benchmark in ??MIDICORE? VVER-1000 core periphery power tilt benchmark proposal?.
Future activities are also shortly described in the end of the paper.

Serpent is a three-dimensional continuous-energy Monte Carlo reactor physics burnup
calculation code, developed at VTT Technical Research Centre of Finland since 2004. The
code is specialized in lattice physics applications, but the universe-based geometry description
allows transport simulation to be carried out in complicated three-dimensional geometries as
well. The suggested applications of Serpent include generation of homogenized multi-group
constants for deterministic reactor simulator calculations, fuel cycle studies involving detailed
assembly-level burnup calculations, validation of deterministic lattice transport codes,
research reactor applications, educational purposes and demonstration of reactor physics
phenomena. The Serpent code has been publicly distributed by the OECD/NEA Data Bank
since May 2009 and RSICC in the U.S since March 2010. The code is being used in some 35
organizations in 20 countries around the world. This paper presents an overview of the
methods and capabilities of the Serpent code, with examples in the modelling of VVER-440
reactor physics.

The recent nodal reactor theory has improved the nodal reactor analyses to the point
where accurate three-dimensional nodal methods can successfully replace the detailed pin-bypin calculations, provided adequate homogenized parameters are available. That has been the
driving force behind the continuing development of transport methods and lattice codes from
one side and homogenization techniques from the other. The introduction of heterogeneous
factors by Koebke (1984) in the Equivalence Theory followed by the General Equivalent
Theory and the definition of the discontinuity factors by Smith (1985), significantly improved
the results from homogenized nodal calculations at the time and basically made possible the
subsequent advances in the nodal methods and higher order homogenization and rehomogenization techniques.
The purpose of this paper is to demonstrate the applicability of HELIOS2 (3) and its post
processor ZENITH in a simple, analytical procedure to derive and study homogenization
parameters in a multi-assembly domain of fuel assemblies and non-multiplying assembly
area. The two-group homogenized parameters are generated in the multi-assembly transport
calculation by HELIOS and then used in one-dimensional, two-group homogeneous diffusion
problem solved in ZENITH. The procedure is used to generate discontinuity factors and
albedo matrices for non-multiplying domain and to verify their dependence on exposure and
calculation conditions.

QUADRIGA and KIRK? ? front-end applications for HELIOS
?
?
Franti?ek HAVLUJ, Radim VOCKA
20th AER Symposium

1.6 Application of Discontinuity factors in C-PORCA 7 code
Istv?n P?s, Tam?s Park?
Paks NPP Ltd
E-mail:pos@npp.hu
S?ndor Patai Szab?
TS-Enercon Ltd
Abstract
During last years there were up-rated the reactor power up to 1485 MW and new fuel
types have been utilised at the Paks NPP. To fulfil the demand of the accuracy and
correctness of on-line core monitoring and off-line core analysis the HELIOS/C-PORCA
models have been modernised as well. The main step of this developing process was to
change the mathematics of the 3D two group diffusion model on the basis of hybrid finite
element method. The upgraded mathematics gave very good results comparing the C-PORCA
calculations against mathematical benchmarks and measurements of different units and
cycles of NPP Paks and Mohovce.
As a final step of the modernisation process the application of flux discontinuity
factors has been made. In the frame of VVER community the usage of this parameter in core
analyses codes is very unusual in contrast with codes for the same purpose in western
countries.
In this paper both the reason of the introduction of discontinuity factors into
HELIOS/C-PORCA models and its effect on the accuracy of calculation are also presented.
We tried to emphasise which kind of codes and which kind of reactor-physical parameters can
be influenced mainly by discontinuity factors. The method of the calculation of flux
discontinuity factors in fuel and non-fuel regions of the core is also described.
As the most important effect of the utilisation of this parameter was that almost all
fittings in C-PORCA code based on in-core measurements have became needless.

Presence of spacing grid in fuel assembly noticeably decreases local energy release
due to small local change of uranium-water ratio. Condition of total energy release
conservation leads to some increase in maximum of axial power distribution. With MCU
Monte Carlo code these increase/decrease were calculated for some VVER-440 and VVER1000 FA?s. Since geometry of spacing grid is very complicated, two different sensibly
simplified models were proposed. Both gave close results. Local minimums turn out to be
~5% lower than average and local maximums increase slightly more than 1%.

The paper deals with the problem of weighting factor calculation and the determination of
spatial weighting functions of ex-core detectors for VVER-440 using the Monte Carlo
method. The computational results were obtained by the well-known code MCNP5 allowing
high performance three-dimensional modeling of complex geometry of the in-vessel and exvessel reactor parts. Despite the fact that adjoint methods dominate in practice, forward mode
of code running was chosen and applied to provide more accurate results contrary to the
adjoint one. The calculation was performed for a boron lined proportional counter CPNB44
installed at the 3rd unit of NPP Jaslovsk? Bohunice. The base is the calculation of ex-core
detector reaction rate induced by a neutron generated in a given volume element of a fuel pin.
All the geometrical details and arising space heterogeneities were taken into account with the
highest accuracy in the complex reactor model. Having obtained computational results, the
weighted least square method was used to fit axial weighting functions. With respect to
horizontal direction, the polyhedral approximation of closed Jordan surfaces was used to find
the proper shape of horizontal weighting factor distribution. Sensitivity and parametric
analysis was performed to evaluate the influence of various reactor operational parameters as
well as the ex-core detector positioning on the weighting function values.

Reactor shut down margin is an important safety parameter. According to the Finnish regulations
recriticality is not allowed to occur in transient and accident cases after reactor scram. With the
Loviisa VVER-440 reactors the most limiting case is cooling in steam line brake accident in the
end of cycle with strong coolant temperature coefficient of reactivity. The reactor recriticality
temperature was measured in 1988 with Novovoronezh-4 end of cycle 15. The experiment was
used to validate our CASMO-4E and HEXBU-3D/mod6 calculation system. The results are
presented and discussed in this paper.

In this paper the results of KARATE-440 calculations on Novovoronezh NPP Recriticality
Experiment are presented, the corresponding parameters are analyzed. The simulation of
the processes and the comparison of the results with the measurements are of particular
interest as these efforts make our code to be validated in a higher level. Even if only some
well defined states of the transient were simulated, satisfactory agreement was found
between measured and calculated data. The results present evidence that the KARATE-440
code package can adequately model the reactor states in a wide range of performance
parameters and the special core type referred in the experiment so it is acceptable for
neutronic analysis of all the VVER-440 NPP`s

Nowadays, t here is a tenden cy to u se best estimate p lus uncertainty m ethods in th e
field of nucle ar energy . This implies th e application of b est estimate code sy stems and the
determination of the corresponding uncertainties. For the latter one an OECD bench mark was
set up. Th e object ive of t he O ECD/NEA Uncert ainty Analysis in Bes t-Estimate Modeling
(UAM) L WR ben chmark is to deter mine the un certainties o f t he coupled reactor
physics/thermal hy draulics LWR cal culations at all s tages. In th is pa per th e AEKI
participation i n Phase I w ill be pres ented. Th is Ph ase is d ealing with the e valuation of the
uncertainties of the neutronic calculations starting from the pin cell spectral calculations up to
the stand-alone neutronics core simulations.
1. INTRODUCTION
The following phases and tasks are defined in the benchmark ¹:
Phase I (Neutronics Phase)
? Exercise I-1: ?Cell Ph ysics? focused on t he d erivation of th e multi-group
microscopic cross-section libraries and their uncertainties
? Exercise I-2: ?Lattice Ph ysics? focused o n th e d erivation of the few group
macroscopic crosssection libraries and their uncertainties
? Exercise I-3: ?Core Physics? focused on the core steady state stand-alone neutronics
calculations and their uncertainties
Phase II (Core Phase)
? Exercise II-1: Fuel thermal properties relevant to transient performance
? Exer cise II-2: Neut ron kinetics stand alone perfor mance (kinet ics data, space- time
dependence treatment, etc.)
? Exercise II-3: Thermal-hydraulic fuel bundle performance
Phase III (System Phase)
? E xercise III-1: Coupled neutronics/thermal-hydraulics core pe rformance (coupled
steady state, coupled depletion, and coupled core transient with boundary conditions)
? Exercise III-2: Thermal-hydraulics system performance
The benchmark was started 4 y ears ago, and almost all participants are still dealing with the
exercises of Phase I (and mostly with Exercise I-1 and Exercise I-2). The main reason of that
is the complexity and the non trivia lity of t his Phase: e.g. there are a lot of ta sks relating to

The MIDICORE benchmark is a 2D calculation benchmark based on the VVER-1000 reactor
core cold state geometry with taking into the account the geometry of explicit radial reflector.
The main task of this benchmark is to test the pin by pin power distribution in selected fuel
assemblies that are placed mainly at the border of the VVER-1000 core. This is due to an
observed phenomenon at calculation of the ?first core loading? (completely composed from
TVEL TVSA-T fresh fuel assemblies ¹) at Temelin NPP (VVER-1000 core), where the
maximum of fuel pin power (FdH) was found in a peripheral FA in a fuel pin at assembly edge
in direction to the core centre. This phenomenon consist not only in position where this
maximum occurs, but especially in relatively big difference of FdH value observed when this
value was determined by codes based on pin to pin diffusion difference method on one side
and by codes based on nodal diffusion method with pin power reconstruction on the other
side. Because value of FdH is not directly measured by the core monitoring system, a decision
about a proposal of benchmark of this kind has been made. In this contribution we define the
MIDICORE benchmark; we present the preliminary reference Monte Carlo calculation results
and also preliminary MOBY-DICK macrocode calculation results.

For some years MCU specialists perform wide range of full-scale VVER-1000 neutron
physical calculations. One of interesting questions emerged in course of these calculations is
the elaboration degree of radial reflector geometry description. MIDICORE 2D approach
simplifies modeling of horizontal grooves in core basket. We modeled them not only exactly,
but also in two approximate ways, one of which was proposed in MCNP reference
calculation. Apart from this reference restricted core area all three cases considered were 60o
rotational symmetry segment of model core. keff, relative pin by pin power distribution in
proposed FA?s and integral fission power of all FA?s was obtained.

The unstable accuracy of power forecasts at periphery fuel pins and utilization of exploitation
data from VVER-440 reactors are main motivations for benchmark definition. Second
generation fuel assemblies with mean enrichment 4,25 % at 5-year cycle at Unit 4 of NPP
Bohunice are analyzed with emphasis on the last cycle. Starting point, calculated period and
results are characterized. SCORPIO data will be used for comparison.

One of the limiting factors in terms of nuclear power plant lifetime is the reactor pressure
vessel condition. The most important ageing effect on the reactor pressure vessel is the
embrittlement. Its radiation embrittlement cause especially fast neutrons. One of the
parameters describing the radiation damage is DPA (Displacement per Atom). In this work
we are focused on describing influence of operating parameters such as boric acid
concentration, control rod position, coolant temperature and burn-up on the neutron field and
DPA in the reactor pressure vessel of VVER-440/V-213 reactor. The calculation of DPA was
realized on detailed 3D model of reactor using MCNP5 code. The goal of this work is to
improve the assessment of the reactor pressure vessel degradation and following valuation
possibility of reactor lifetime extension.

The reactor plant service life is governed, to a high extent, by radiation life of the
reactor vessel that, in its turn, depends on maximum value of the fast neutron fluence (E>0.5
MeV) to the reactor vessel inner surface. Therefore the calculation of fast neutron fluence to
the reactor vessel, as well as determination of calculation error, is an important task in
justification of the reactor vessel service life.
Result of neutron fluence calculation and its error depend substantially on the
accuracy of assigning the source of fission neutrons in the core, and especially in the
peripheral fuel assemblies. The given paper deals with consideration of two main approaches
to assigning the source of fission neutrons ? rod-by-rod and assembly-by-assembly. The main
objective of this paper is to determine the difference in results of calculation of neutron
fluence to the reactor vessel using different methods of description of the fission neutron
source, and to determine which of these methods is the most preferable for justification of the
reactor vessel service life from the viewpoint of maintaining the reasonable conservatism.
Error of calculation of fast neutron fluence to the reactor vessel of VVER-440 and
VVER-1000 was estimated by a comparison of calculated and experimental values of activity
of templets taken from the inner surface of reactor vessels of VVER-440 (V-230) and activity
of neutron activation detectors on the outer surface of reactor vessels of VVER-440 and
VVER-1000. On the basis of the analysis of calculated and experimental data a conclusion
was made on the error of calculation results obtained with the use of assembly-by-assembly
and rod-by-rod distribution of fission neutron source.

Review of AER Working Group B Meeting in Modra – Harm?nia, Slovakia is given.
Regular meeting of Core Design Group was organized by VUJE, Inc. and held at Modra pension Harm?nia, Slovakia, April 20?22, 2010, together with Working Group A. Presented
papers (see List of papers and List of participants) covered topics as follows.

Loviisa NPP is currently in the transition phase from 3-batch to 4-batch loading scheme.
During the 3-batch loading 3.7 ? 4.0 % enriched fuel has been used. By the end of the
transition phase the cores of both NPP units will be consisting purely on TVEL second
generation fuel with average enrichment of 4,37 %. The new fuel assembly type includes six
gadolinium doped rods (3,35 % Gd2O3) that are located next to the corner rods.
Introduction of the new fuel type has an effect on certain reactor core characteristics.
From the point of view of reactor operation under normal conditions the most remarkable
change is seen in the decrease of the hot subchannel enthalpy rise margin. In the worst case
the reduced margin could lead to power limitation for some period of the cycle. In order to
reduce the risk of power limitation special attention is paid on modeling of the coolant mixing
in the fuel bundle and also in the upper part of the assembly. A short study on these modeling
issues is given in the paper.
The first transition cycle with Gd fuel reload batch started in the Loviisa-1 unit in September
2009 and ended in August 2010. In the Loviisa-2 unit the first transition cycle will be starting
in October 2010. The paper gives a discussion on core operating experience in the recent
cycles. Attention is paid on outlet temperature measurement readings for the Gd fuel
assemblies during the first transition cycle. Measured outlet temperatures are expected to
change slightly due to Gd induced change in assemblywise pinpower distribution.

Tianwan Nuclear Power Station (TNPS) owns two VVER-1000 reactors imported from
Russia at present. The two reactors are put into commercial operation in 2007 and operating
in the 4th fuel cycle in 2010. This report briefly describes the characteristics of fuel assemblies
and in-core loadings for TNPS? VVER-1000 reactors under annual refueling scheme. As well
as the plan of 18-month fuel cycle, which will be introduced to TNPS? VVER-1000 reactors
in the future, is presented in the report.

The Dukovany NPP has operated four VVER-440 reactor units. Starting 2009 power of 3rd unit has been uprated, so
reactor is operated at power level of 1444 MWt instead of 1375 MWt used up to now. Innovation of secondary circuit
and electrical equipment leads to electrical output of 500 MWe. During four years all four units will be uprated this
way. Power uprates together with outage shortages are challenges for fuel cycles optimization. Reactors of Dukovany
NPP have already operated in incomplete five-year fuel cycles based on modified Gd-2+ fuel design called Gd-2M
with 4.38 % average enrichment now. New fuel types with high enrichment were proposed by fuel manufacture
(4.87 % avg. enrich.) and by SKODA JS Company (4.76 % avg. enrich.). These fuels should allow complete fiveyear fuel cycle with higher power and has also ability to allow incomplete six-year fuel cycle. Fuels should
economize fuel assemblies and operational costs.
The SKODA JS Company together with the University of West Bohemia has developed new optimizing code
called Athena. The code has been used to optimize fuel cycles from 25th cycle of 3rd unit of Dukovany NPP up to 34th
cycle with new fuel types presented above. Comparison between fuel cycles with Gd-2M fuel and with mixed
loading patterns (Gd-2M mixed with new types) has been calculated. Positives and negatives are discussed from
physical, safety, operational, and economical points of view. New fuels allow 330 to 340 FPD long cycles with
uprated power, but are very challenging to radial power distribution parameter (FdH). Results show, that operational
parameter limit of FdH used until now should be increased from 1.54 to 1.58. Couple of fuel assemblies is
economized. Results are a part of study of applicability and feasibility of new fuels and might be important for
additional development of fuel design. Results will also be used as a comparison of presently used optimization code
Optimal and new Athena code.

A nearly equilibrium five-year cycle has been achieved at Dukovany NPP in the last years.
This means that working fuel assemblies (WFA) with an average enrichment of 4.25 w% of 235U
(control assemblies (CA) with an average enrichment of 3.82 w% of 235U) are normally loaded
and reloaded for five years. Operation at uprated thermal power (105% of the original one,
increase from 1375 MWth to 1444 MWth) started by use of WFA with an average enrichment of
4.38 w% of 235U (CA with an average enrichment of 4.25 w% of 235U) in 2009.
With the aim of fuel cycle economy improvement, the fuel residence time in the core has to be
prolonged up to six years with one cycle duration time and preserving loadings with very low
leakage. In order to achieve this goal, at least neutron-physical characteristics of FA must be
improved and such changes should be evaluated from other viewpoints. Some particular changes
have already been analyzed earlier.
Designs of new fuel assemblies with higher (and in the central part of a FA the highest
possible, i.e. 4.95 w% of 235U) enrichment with preserving low pin power non-uniformity are
described in the presented paper. An FA with an average enrichment of 4.76 w% of 235U (lower
than originally evaluated) containing six fuel pins with 3.35 w% of Gd2O3 content was selected
in the end. Fuel pins have bigger pellet diameter, but preserved central hole.
A newly designed FA?s were evaluated at first from the viewpoint of physics (pin power nonuniformity, cycle length etc.).
Possibilities of fuel cycles are evaluated on model loadings with the newly designed FA,
where the base are loadings for 27th ? 34th cycles of the third unit of Dukovany NPP for uprated
power. These cycles were prolonged (from approx 330 FPD to 370 FPD) using FA with higher
enrichment.
Also, a preliminary evaluation of FA with a quite new design is presented.

The Project ?Mochovce Units 3&4 Completion? started in 2009 and it will be finished in
2013. VUJE, Inc. is one of the five main Project contractors for the Nuclear Island and it is
responsible, inter alia, for Mochovce Units 3&4 commissioning. The commissioning of Units
3&4 includes the stages of Physics Commissioning and Power Commissioning. This paper
deals with the preparation of Mochovce Units 3&4 Physics Commissioning. In the paper there
is presented a preparation of some commissioning documents, e.g. ?Quality Assurance
Programme?, ?Commissioning Programme?, ?Stage Programme for Physics
Commissioning?, ?Test working programmes?, ?Neutron-physics characteristics for Physics
and Power Commissioning?, etc. The scope of Physics Commissioning is presented by list of
tests. For assessment of tests results so-called three-level acceptance criteria will be applied:
realization, design and safety criteria. In the paper there are also presented computer codes,
which will be used for neutron-physics characteristics calculation and the fuel loading scheme
for the reactor core of Mochovce Unit 3.

Ex-core neutron flux measurement system was modernized at Bohunice NPP in period
2007-2008. The previous system AKNT-2 was replaced by new system AKNT-17R. In spite
of the new modern system, neutron flux measurement accuracy is still influenced by the
changes of various parameters: control assemblies group 6 position, coolant temperature at
reactor inlet, power distribution in reactor core change because of fuel burn-up, etc. Therefore
AKE-02R system (neutron power on-line calibration system) was installed at Bohunice NPP
in 2008 and 2009. AKE-02R system was working in open-loop mode more than one year and
finally at the end of 2009 (Unit 3) and at the beginning of 2010 (Unit 4) was switched into
close-loop mode.
The purpose of AKE-02R system is to increase ex-core neutron power measurement
accuracy. AKE-02R system eliminates above mentioned dependencies by using correction
factors, determined on the basis of real control assemblies axial position, real coolant inlet
temperature and real burn-up. Correction factors are continually calculated in AKE-02R
system and next enter into AKNT-17R system. New corrected value of neutron power is
computed in AKNT-17R system by using correction factor. Corrected value of neutron power
is used as input value for all other systems (reactor control system, reactor trip system, reactor
limitation system, etc.).
In this paper brief description of AKNT-17R system and AKE-02R system is
presented. The process of commissioning of AKE-02R and also the results of tests are
explained. Operational experiences with AKE-02R system after switching into close-loop
mode are showed.

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 (EDU, Czech Republic) and two units of
Bohunice NPP (EBO, Slovak Republic) replacing the original Russian VK3 system. By both
Czech and Slovak nuclear regulatory bodies it was licensed as a Technical Specification
Surveillance tool.
The monitoring system operates in two modes: in core follow mode and in predictive mode.
In the core follow mode, the present core state is evaluated by a method combining the
instrumentation signals and the theoretical calculation of the power distribution done by the
core simulator. This procedure is followed by an automatic limit checking, where
characteristics of the current state are compared to the Technical Specifications. The operator
obtains relevant information on core status through the dedicated Man-Machine Interfaces.
In the predictive mode, the operator can visualize the core characteristics during the transients
forecasted for coming hours or days. Quick forecasts realized by the strategy generator are
deeply analyzed by the predictive simulator. Similarly as in the core follow mode,
characteristics of the evaluated states can be compared against Technical Specifications.
Since it?s first installation, the development of SCORPIO-VVER system continues along with
the changes in VVER reactors operation. The system is being adapted according the utility
needs and several notable improvements in physical modules of the system were introduced.
The latest most significant changes were done in connection with implementation of a new
digital I&C system, loading of the optimized Gd2 fuel assemblies, improvements in the area
of core design (neutron physics, core thermal hydraulics and fuel thermal mechanics) and
improvements in the predictive part of the system (Strategy Generator).
The currently finished upgrades (Upgrade 2 at EBO 06/2009, Upgrade 5 at EDU 12/2009)
includes the adaptation of the system to up-rated unit conditions as well as further
improvements of methods applied in physical modules, especially as are the improvements of
3D power reconstruction methods by using the SPND detectors in fuel assemblies, as are the
changes in design and methodology of the limit checking and as is the implementation of the
on-line shutdown margin calculation to the system.

The VVER-440 nuclear fuel vendor permanently improves the assortment of produced
nuclear fuel assemblies for achieving better fuel cycle economy and reactor operation safety.
During unit refuelling there also could be made some other changes in hydraulic parameters
of primary circuit (change of impeller wheels, hydraulic resistance coefficient changes of
internal parts of primary circuit, etc.). Therefore it is necessary to determine real coolant flow
rate through the reactor during units start-up after their refuelling, and also to have the skilled
methodology and computing code for analyzing factors, which affecting the inaccuracy of
coolant flow redistribution determination through reactor on flows through separate parts of
reactor core in any case of parallel operation of different assembly types. Computing code
TH-VCR and CORFLO are used for reactor core characteristics determination for one type of
fuel and control assemblies and also in case of parallel operation of different assembly types.
The code TH-VCR is able to calculate coolant flow rate for different combinations of three
different fuel assembly channel types and three different control assembly channel types. The
CORFLO code deals the area of the reactor core which consists of 312 fuel assemblies and 37
control assemblies. Regarding the rotational 60 o symmetry of reactor core only 1/6 of reactor
core with 59 fuel assemblies is taken into account. Computing code CORFLO is verified and
validated at this time. Paper presents some results from measurements of coolant flow rate
through reactors during start-up after unit refuelling and short description of computing code
TH-VCR and CORFLO with some calculated results.

The AER Working Group D on VVER reactor safety analysis held its 19th meeting in Pisa,
Italy, during the period 15-16 April, 2010. The meeting was hosted by the San Piero a Grado
Nuclear Research Group of the University of Pisa and was held in conjunction with the
second workshop on the OECD/NEA Benchmark for the Kalinin-3 VVER-1000 NPP and the
fourth workshop on the OECD Benchmark for Uncertainty Analysis in Best-Estimate
Modelling (UAM) for Design, Operation and Safety Analysis of LWRs. Altogether
12 participants attended the meeting of the working group D, 8 from AER member
organizations and 4 guests from non-member organization. The co-ordinator of the working
group, Mr. S. Kliem, served as chairman of the meeting.
The meeting started with a general information exchange about the recent activities in the
participating organizations.
The given presentations and the discussions can be attributed to the following topics:
?
?
?

Results of calculation study of transient connected with drop of WWER-1000 cluster of
working group are presented. Transient was considered in the mode of automatic power
control without forming of warning protection signal due to reaching of dropped cluster of
core bottom. Calculations are shown that given transient can cause valuable distortion of
power distribution in axial direction. At that main increase of pin power is occurred in upper
part of the core, whereas power in lower part is almost not changed. The additional increase
of power in the upper part of core makes conditions for initiation of DNB. This effect can be
observed if in initial state axial power distribution is displaced in upper part of core nearby to
rest of supported power clusters of working group. It is necessary to define conservatively
with taking into account assumed working group efficiency ? in which row from extracted
clusters of working group the displacement of axial power in the upper part is possible.
Probability of such displacement and its localization in plane of core must be properly
analyzed. The work was performed in framework of orders BMU SR 2511 and BMU
R0801504 (SR2611). The report describes the opinion and view of the contractor ? SSTC
N&RS – and does not necessarily represent the opinion of the ordering party – BMU-BfS/GRS
and T?V S?D.

Dependence of gas gap properties on bu
rnup has been obtained with use of
TRANSURANUS [ 2] code. Implem
ented dependency on burnup is based on
TRANSURANUS calculations of different fuel pins upon different linear power Ql. Obtained
dependence was implemented into DYN3D ¹ code and results of new dependence effect on
characteristics of WWER fuel loadings are presented.
The work was perform ed in fram ework of orders BMU SR 2511 and BMU R0801504
(SR2611). The report describes the opinion and vi ew of the contractor ? SSTC N& RS – and
does not necessarily represent the opinion of the ordering party – BMU-BfS/GRS and T?V
S?D.

The CFD code ANSYS CFX has been coupled with the neutron-kinetic core model DYN3D.
ANSYS CFX calculates the fluid dynamics and related transport phenomena in the reactor?s
coolant and provides the corresponding data to DYN3D. In the fluid flow simulation of the
coolant, the core itself is modeled within the porous body approach. DYN3D calculates the
neutron kinetics and the fuel behavior including the heat transfer to the coolant. The physical
data interface between the codes is the volumetric heat release rate into the coolant. In the
prototype that is currently available, the coupling is restricted to single-phase flow problems.
In the time domain an explicit coupling of the codes has been implemented so far.
Steady-state and transient verification calculations for two small-size test problems confirm
the correctness of the implementation of the prototype coupling. The first test problem was a
mini-core consisting of nine real-size fuel assemblies with quadratic cross section.
Comparison was performed with the DYN3D stand-alone code. In the steady state, the
effective multiplication factor obtained by the DYN3D/ANSYS CFX codes shows a deviation
of 9.8 pcm from the DYN3D stand-alone solution. This difference can be attributed to the use
of different water property packages in the two codes. The transient test case simulated the
withdrawal of the control rod from the central fuel assembly at hot zero power in the same
mini-core. Power increase during the introduction of positive reactivity and power reduction
due to fuel temperature increase are calculated in the same manner by the coupled and the
stand-alone codes. The maximum values reached during the power rise differ by about 1 MW
at a power level of 50 MW. Beside the different water property packages, these differences are
caused by the use of different flow solvers.
The same calculations were carried for a mini-core with seven real-size fuel assemblies with
hexagonal cross section in order to prove the applicability of the coupled code to cores with
hexagonal fuel assemblies. The differences between the results of coupled calculations and
those of the stand-alone DYN3D code are in the same range as for the quadratic mini-core.

Lappeenranta University of Technology (LUT) gives most extensive education in nuclear
engineering in Finland and it has the largest experimental facilities. The research concentrates
mostly on thermal hydraulics phenomena. The integral facility PACTEL models the Loviisa
VVER-440 primary circuit in height scale 1:1 and volume scale 1:305. Now it has been
modified to include two vertical type steam generators and new loops in order to be able to
also model phenomena of the EPR type PWR which is been constructed in Finland ? it has
become PWR PACTEL. However, the ability to carry out VVER tests has not been lost
because the original PACTEL can still be used due to the ingenious construction of the
complex. The analyzing calculations of the PACTEL and PWR PACTEL tests are carried out
in LUT both with the Finnish code APROS and the US code TRACE. Recently the work on
VVERs has mostly been carried out with different separate effect test facilities concentrating
on pressure vessel wall cooling, straining effects, and emergency core cooling. In these
connections the CFD codes, e.g. Fluent are the main calculation code types utilized.

APROS – Advanced PROcess Simulation environment ? is a widely used simulation tool for
nuclear power plant modelling. Earlier the three-dimensional neutronics calculation has been
based on model using the difference method. The original three-dimensional core model is
mainly used in power plant simulator applications, where it fits well because of its speed. For
safety analysis purposes, however, a new model was considered to be an important
improvement to meet the accuracy requirements. A sophisticated nodal model used already in
HEXTRAN and TRAB-3D was decided to be implemented into APROS. The hexagonal part
of the model has now been implemented and tested. For practical reasons, the model was
programmed from scratch into APROS and also some small improvements were added and
thus, an extensive validation program was necessary to prove the correct behaviour of the
model. In this paper, the most important results from AER kinetic benchmarks 2 & 3
calculations are shown as well as the calculation results against data achieved LR-0 test
reactor space-time kinetic experiments. Since the model is similar to the one in HEXTRAN,
the results in benchmarks are compared to the results by it. In LR-0 calculations, results by
both, original and new model are presented and compared to the measurements. The results
shows that the implementation of the model has been successful and the new model improves
the accuracy of three-dimensional neutronics calculation in APROS into the level required in
safety analyses.

A detailed RPV model of VVER-440/213 type reactors was developed in BME NTI in the last
years. This model contains the main structural elements as inlet and outlet nozzles, guide
baffles of hydro-accumulators, alignment drifts, perforated plates, brake- and guide tube
chamber and simplified core. For the meshing and simulations ANSYS softwares (ICEM 12.0
and CFX 12.0) were used. With the new vessel model a series of parameter studies were
performed considering turbulence models, discretisation schemes, and modeling methods. In
steady state the main results were presented on last AER Symposium in Varna. The model is
suitable for different transient calculations as well.
The purpose of the suggested new benchmark (7th Dynamic AER Benchmark ¹) is to
investigate the reactor dynamic effects of coolant mixing in the VVER-440/213 reactor vessel
and to compare the different codes. The task of this benchmark is to investigate the start up of
the 6th main coolant pump. The computation was carried out with the help of ATHLET/BIPRVVER code in Kurchatov Institute ² for this transient and was repeated with ANSYS CFX
12.0 at our Institute.

5.10 Comparison of In-Core Thermocouple and SPND Measured Data
with the ATHLET-BIPR-VVER Predictions
S. P. Nikonov, RRC Kurchatov Institute, Russia
K. Velkov and A. Pautz, GRS mbH, Germany

5.11 Peculiarity by Modeling of the Control Rod Movement by the
OECD Kalinin-3 Benchmark
S. P. Nikonov, RRC Kurchatov Institute, Russia
K. Velkov and A. Pautz, GRS mbH, Germany

Summary on the activity of working group G in 2010
Attila Asz?di, Coordinator of Working Group G
The AER Working G meeting was held on 31 May – 1 June 2010 in Balatonf?red, Hungary
together with Working Group C. The short summary of Working Group G presentations is the
following.
1. D. Tar (AEKI): Experimental Investigation of Coolant Mixing in VVER and PWR Reactor
Fuel Bundles by Laser Optical Techniques for CFD Validation
? PIV and LIF measurements of flow in a VVER-440 fuel assembly head at different
boundary conditions (five heating configuration)
? Test section: 1 m long rod bundle (36 heated rods, 72 kW total heat power) with
assembly head
? Presentation of some velocity and temperature fields in the assembly head
? Calculation of POD modes in 5?5 PWR rod bundle with spacer grids
2. B. Kiss, A. Asz?di (BME NTI): Benchmark Calculation with Improved VVER-440/213 RPV
CFD Model
? Detailed CFX model of VVER-440/213 reactor pressure vessels with a simplified core
? Transient calculations for starting up of the 6th main coolant pump
? Comparison between ATHLET and CFX results
? Development of hexahedral mesh for the downcomer
3. R. Szij?rt?, B. Yamaji, A. Asz?di (BME NTI): Study of Natural Convection Around a
Vertical Heated Rod Using PIV/LIF Technique
? Simultaneous measurements of velocity and temperature fields performed applying a
combined PIV and LIF setup to investigate natural convection around the vertical
heated rod.
? Local heat transfer coefficient calculated from measured data.
4. P. M?hlbauer (REZ): A Simplified Approach to VVER-440 Fuel Assembly Head
Benchmark
? Model for VVER-440 fuel assembly head without spacer grid
? Calculation of benchmark exercises with FLUENT code
? Unstructured mesh of about 14 million cells
? Presentation of calculated temperature and velocity distributions, thermocouple
signals
5. A. Shishov, O. Kudryavtsev, D. Posysaev (GIDROPRESS): VVER-440 Fuel Assembly
Head CFD benchmark ? Preliminary results
? Model for VVER-440 fuel assembly head
? Three hexahedral meshes of 10.5, 12 and 25 million cells
? Calculation of benchmark exercises with CFX code
? Presentation of calculated temperature and velocity distributions, thermocouple
signals
6. S. T?th, A. Asz?di (BME NTI): First Results of VVER-440 Fuel Assembly Head Benchmark
? Introduction of benchmark exercises
? First comparison of GIDROPRESS, REZ and VUJE results
? After discussions agreement with the participants: benchmark calculations should be
performed with and without central tube flow

Paks Nuclear Power Plant uses the REMIX code for the calculation of the coolant mixing in
case of the use of high pressure injection system while stagnating flow is present. The use of
the code for Russian type WWER-440 reactors needs strict conservative approach, and in
several cases the accuracy and the reserves to safety margins cannot be determined now. In
order to quantify and improve these characteristics experimental validation of the code is
needed.
An experimental program has been launched at the Institute of Nuclear Techniques with the
aim of investigating thermal stratification processes and the mixing of plumes in simple
geometries. With the comparison and evaluation of measurement data and computational fluid
dynamics (CFD) results computational models can be validated.
For the experiments a simple hexahedral plexiglas tank (250×500×100 mm ? HxLxD) was
fabricated with five nozzles attached, which can be set up as inlets or outlets. With different
inlet and outlet setups and temperature differences thermal stratification, plume mixing may
be investigated using Particle Image Velocimetry (PIV).
In the paper comparison of PIV measurements carried out on the plexigas tank and the results
of simulations will be presented. For the calculations the ANSYS CFX three-dimensional
CFD code was used.

Nonintrusive laser optical measurements have been carried out to investigate the coolant
mixing in a model of the head part of a fuel assembly of a VVER reactor. The goal of this
research was to investigate the coolant flow around the point based in-core thermocouple; and
also provide experimental database as a validation tool for CFD calculations. The experiments
have been carried out on a full size scale model of the head part of VVER-440/213 fuel
assembly. In this paper first the previous results of the research project is summarised, when
full field velocity vectors and temperature were obtained by particle image velocimetry (PIV)
and planar laser induced fluorescence (PLIF), respectively. Then, preliminary results of the
investigation of the influence of the flow in the central tube will be reported by presenting
velocity measurement results. In order to have well measurable effect, extreme flow rates
have been set in the central tube by applying an inner tube with controlled flow rates. Despite
the extreme conditions, the influence of the central tube to the velocity field proved to be
significant. Further measurement will be done for the investigation of the effect of the gaps at
the spacer fixings by displacing the inner tube vertically, and also the temperature distribution
will also be determined at similar geometries by LIF. The aim of the measurements was to
establish an experimental database, as well as the validation of CFD calculations.

The Nuclear Training Reactor of the Institute of Nuclear Techniques (Budapest
University of Technology and Economics, Hungary) is a pool-type reactor with light water
moderator and with a maximum thermal power of 100 kW. The fuel elements are cooled by
natural convection. An experimental setup was built to analyse the nature of the natural
convection around a heated rod. The flow field was investigated using an electrically heated
rod, which models the geometry of a fuel pin in the training reactor. The electric power of the
model rod is variable between 0?500 W. The rod was placed in a square-based glass tank.
PIV (Particle Image Velocimetry) and LIF (Laser Induced Fluorescence) measurement
techniques were used to study the velocity and temperature field in a two-dimensional area.
The thermal and the hydraulic boundary layers were detected near a rod in a lower
section of the aquarium. The laminar-turbulent transition of the flow regime was observed,
the maximum velocity of the up-flow was 0.025-0.05 m/s. From the temperature
measurements the local heat transfer coefficient was estimated.

In the VVER-440/213 type reactors, the core outlet temperature field is monitored with
in-core thermocouples, which are installed above 210 fuel assemblies. These measured
temperatures are used in determination of the fuel assembly powers and they have important
role in the reactor power limitation. For these reasons, correct interpretation of the
thermocouple signals is an important question. In order to interpret the signals in correct way,
knowledge of the coolant mixing in the assembly heads is necessary. Computational Fluid
Dynamics (CFD) codes and experiments can help to understand better these mixing processes
and they can provide information which can support the more adequate interpretation of the
thermocouple signals.
This benchmark deals with the 3D CFD modeling of the coolant mixing in the heads of
the profiled fuel assemblies with 12,2 mm rod pitch. Two assemblies of the 23rd cycle of the
Paks NPP?s Unit 3 are investigated. One of them has symmetrical pin power profile and
another possesses inclined profile. In this benchmark, the same fuel assemblies are
investigated by the participants thus the results calculated with different codes and models can
be compared with each other. Aims of benchmark was comparison of participants? results
with each other and with in-core measurement data of the Paks NPP in order to test the
different CFD codes and applied CFD models. This paper contains OKB ?GIDROPRESS?’s
results of CFD calculations this benchmark. Results are:
? In-core thermocouple signals above the selected assemblies;
? Deviations between the in-core thermocouple signals and the outlet average coolant
temperatures of the assemblies;
? Axial velocity and temperature profiles along three diameters at the level of the
thermocouple;
? Axial velocity and temperature distributions in the cross section at the level of the
thermocouple;
? Axial velocity and temperature distributions in the center plane of the assembly head
model.

The latest version of Russian fuel VVER-440 includes burnable absorber in 6 pins. In this article is
impact of burnable absorber on nuclide composition and criticality analyzed. In part 1 was analyzed
whole burnup interval 0 – 50 MWd/kgU ³. In present part 2 are detailed analysis only for first cycle
(burnup 0 – 10 MWd/kgU)

20th SYMPOSIUM of AER on VVER Reactor Physics and Reactor Safety
September 20 ? 24, 2010, Espoo, Finland

A very topical problem of nuclear power is the fuel cycle back-end. One of the options is a
LWR spent fuel reprocessing and fissile nuclides re-use in the same light water reactors. Under
the terms of LWR-DEPUTY project (Light Water Reactor Fuels for Deep Burning of Pu in
Thermal Systems) the new innovative VVER-440 transmutation fuel assembly by the paper
authors was developed. The innovative transmutation assembly includes two different types of
nuclear fuel. The first one is UO2 and the second fuel type consists of the metal molybdenum
and PuO2 mixture. The five period VVER-440 core was proposed by the molybdenum
transmutation assembly and also the transient process (control rod ejection) in this core was
analyzed. All kinds of calculations were performed by computer codes HELIOS 1.9, BIPR 7 and
DYN 3D.

20th SYMPOSIUM of AER on VVER Reactor Physics and Reactor Safety
September 20 ? 24, 2010, Espoo, Finland

20th SYMPOSIUM of AER on VVER Reactor Physics and Reactor Safety
September 20 ? 24, 2010, Espoo, Finland
12th session of the AER Working Group F – “Spent Fuel Transmutations”
3rd meeting INPRO IAEA Collaborative Projects RMI
“Meeting energy needs in the period of raw materials insufficiency during the 21st century”
Czech Republic, Liblice, April 6 ? 9, 2010.

10.1 SUMMARY OF THE SPECIAL AER GROUP MEETING
?ELABORATION OF THE METHODOLOGY
FOR CALCULATING THE CORE DESIGN ENGINEERING FACTORS?

One of the topics for discussion at special working group ?Elaboration of the
methodology for calculating the core design engineering factors? is the problem of
engineering factor components. The list of components corresponds to the phenomena that are
taken into account with the engineering factor. It is supposed the better understanding of the
influenced phenomena is important stage for developing unified methodology.
This paper presents some brief overview of components of the engineering factor for
VVER core peaking factors as they are in the Kurchatov Institute methodology. The evolution
of some components to less conservative values is observed. Author makes some assumptions
as for the further progress in components assessment.

From the late nineties Paks Nuclear Power Plant ? in collaboration with KFKI Atomic Energy
Research Institute (KFKI AEKI) ? is developing a system for determining the normal
operation local power engineering safety factors. The system is based on a Monte Carlo
sampling of the uncertain model input parameters. Additionally, the comparison of the
calculation to the in-core measurements plays essential role for determining some important
input parameters. By using new fuel types and the corresponding more recent detailed
technological data, the applied method is being improved from time to time. Presently, the
actually used and authorized engineering safety factors at Paks NPP are determined by using
this method.
In the paper, the system?s main properties are described (not going beyond the possible
extent). The main points are as follows:
?
?
?
?

The problem of the power distribution uncertainties is in general a multidimensional
problem of random vector and problem of multidimensional Probability Density Function
(PDF). The standard methodology of derivation VVER engineering factors is based on
representation of analyzed power peaking in the linear form of random factors and on
presumption about their normal PDF.
In this paper is presented the derivation of locally defined engineering factors and for
mechanical factors has been performed their reformulation. Final formulation of engineering
factors as a statistics of relative deviations involves new parameter ? mean.
Engineering factors definition from so called endpoints of uncertainty tolerance
interval is recommended. Approach (95%/95%) for normal PDF is discussed in detail, the
relation to present standard uncertainty methodology of power distribution is found and
problem of optimality in tolerance factor finding including limitation of sample size is
discussed.
On the bases of statistically based uncertainty kinf analysis for linear model has been
shown that multivariate outputs vector of power peaking has nearly normal PDF
independently on the PDF character of input multivariate vector from under its small
dimension (lower than number of FAs in 1/6 core symmetry).
Finally the development of the methodological part of the engineering factors for
VVER-1000 design macrocode MOBY-DICK is described and their calculation direct on the
bases of SPD experimental data of Temelin NPP and Volgodonsk NPP has been performed
with inclusion the variability of detector (SPD).