Reactor Physics Benchmarks
The following benchmarks concern the validation of computer
codes and data required for reactor design and operation, covering LWR
reactor core transients, power distribution within fuel assemblies and
reactor cores. Benchmarks relating to other fields are listed at the
end of this page.
Recent benchmarks in progress
Power Distribution within Assemblies
A final meeting of Phase II, addressing the comparison of production
codes in reconstructing the pin power in UOX/MOX assemblies in 2
dimensional models, was held in November 1997. This second phase deals
in particular with the cross section data reduction methods. Fourteen
solutions have been received and more than fifteen calculation schemes
have been tested. The final report is foreseen for December 1999. The
3D extension of this benchmark will be considered in the frame of a
3D VENUS-2 benchmark discussed by
the
WPPR.
Forsmark 1 & 2 BWR Stability Benchmark
The purpose of this benchmark is to compare different time series
analysis methods that can be applied to the study of BWR stability. It
is a follow-up activity of the
Ringhals 1 Stability
Benchmark. While the Ringhals 1 Stability Benchmark included both
time domain and frequency domain calculation models to predict
stability parameters, the new activity is focused on the analysis of
time series data by means of noise analysis techniques in the time
domain.
A first meeting of participants in the benchmarks was held
from 18-19 February 1999 at the Consejo de Seguridad Nuclear, Madrid. A
total of 17 participants from 8 countries and 13 organisations
attended. It was agreed that the major objective was met, namely to
verify that different methods give the same answer. The applicability
and reliability of the different methods were investigated. The six
cases chosen for this study were relatively difficult and were really
addressing the limits of the methods. Use of a full set of data could
be the subject of a different investigation involving reactor physics.
The final report will be submitted to the NSC and the CSNI/PWG2 for
review in the autumn 1999 and will be published by the end of 1999. For
detailed information see NEA/NSC/DOC(99)9.
PWR Main Steam-Line Break Benchmark
This study is co-organised by the NSC and the CSNI/PWG2 on Coolant
System behaviour. The third meeting was hosted by GRS, Garching,
Germany, from 24-25 March 1999:
NSC/DOC(99)6. It
was attended by 41 participants from 22 organisations from 11
countries. The main objective was to present the "final" Phase I report
concerned with the system model and point kinetics and to discuss Phase
II results concerning the 3D core model, for which thermal-hydraulic
boundary conditions are provided. Each of the two first phases involved
more than 10 different code systems. It was noticed that not all
participants have strictly complied with the specification provided for
the different phases of the MSLB benchmark and several participants
have considerable deviations in the modelling used. This resulted in a
relatively large spread of the results. It was therefore agreed to
issue a final version of the specification covering the three
exercises, with the aim of removing problems of interpretation and
sources of discrepancy. Strongly deviating solutions, resulting from
non-compliance with the benchmark specification will not be included in
the final report. It was agreed that the PWR-MSLB would be published in
4 volumes in the next two years:
It is the intention to prepare also a CD-ROM with the four reports and
supplementary tables and graphs not published in the paper versions.
An ad-hoc meeting on the PWR MSLB Benchmark was held on 26
September 1999 in Madrid in conjunction with the M&C'99-Madrid
Conference. It was attended by twenty-six participants from 16
organisations and 7 countries. The work on completing the final reports
on the first and second exercises is progressing fast and the reports
will be sent for approval at the beginning of 2000. The first two
exercises allowed participants to develop their own system and core
models and to test them against each other's. The major effort was
devoted to narrow down modelling differences in terms of
misinterpretations of the specifications. Additional benefits came from
the fact that deviations in different code predictions were observed
separately for system and core modelling, before comparing the "best
estimate" coupled results. It was concluded that the so-called "return
to power" transient scenario for the second exercise really helped to
identify local nodalisation and coupling effects.
The final PWR-MSLB meeting was held on 25-26 January 2000 in
Paris. The summary of that workshop is NEA/NSC/DOC(2000)2.
For detailed information access the
Summary of the MSLB project
BWR Turbine Trip Transient (BWR-3T) Benchmark
The Turbine Trip (TT) transient in a BWR is a pressurisation event in
which the coupling between core phenomena and system dynamics plays an
important role. The reference problem chosen for simulation in a BWR is
a Turbine Trip transient, which begins with a sudden Turbine Stop Valve
(TSV) closure. The pressure oscillation generated in the main steam
piping propagates with relatively little attenuation into the reactor
core. The objectives of the benchmark are: comprehensive feedback
testing and examination of capability of coupled codes to analyse
complex transients with coupled core/plant interactions by comparison
with actual experimental data.
The specifications are based on experimental data from Peach
Bottom - 2 BWR/4 Nuclear Power Plant. and were presented during the
ad-hoc PWR-MSLB meeting in Madrid. The CSNI-PWG2 on coolant system
behaviour endorsed this proposal at its meeting on 20-22 September
1999. The presentation of a specific work programme on the BWR-TTTB was
presented at the last PWR-MSLB meeting in January. The specifications
was issued in February 2001 as NEA/NSC/DOC(2001)1.
For detailed information access the
Summary of the BWRTT project.
This work is carried out in co-operation with the CSNI-PWG2.
3D Radiation Transport Benchmark
A meeting on the 3D Deterministic Radiation Transport Benchmark was
held on 1 October 1999 in Madrid, hosted by the Consejo de Seguridad
Nuclear. The Meeting was attended by 19 participants from 7 countries
and included all authors of 3D deterministic transport codes. The six
cases with void regions had been developed by Prof. K. Kobayashi from
the University of Kyoto. The main objective of the meeting was to
discuss the final draft report of the benchmark study, to draw
conclusions and discuss the publication of the report. It was agreed to
publish the summary report first. The different solutions provided will
be published in a special issue of Progress in Nuclear Energy. Overall
the solutions submitted were satisfactory, but have shown the
limitations due to ray effects of the methods such as SN when void
regions are present and what countermeasures have to be taken to avoid
shortcomings (first collision source). In the discussion, code authors
reviewed a series of additional problems that would be useful to tackle
in order to validate better 3D transport codes. Two were retained,
meeting the needs of different communities in radiation transport. The
first is a revisiting of case C5 from the first phase benchmark on pin
power distribution within core assemblies (PDWA1-C5). E. Lewis from NW
University together with ANL would prepare a 7-group cross-section set
for use in the benchmark. The second concerns streaming of radiation
around a 3D block. Both problems are particularly challenging and would
contribute to validation of algorithms used and to show the benefit of
transport methods for detailed full core modelling. The specifications
will be submitted to NSC for endorsement.
Recently completed benchmarks
Benchmarks carried out in other areas
Benchmark Exercise on e+/e-/photon Physics (BEEP)
In the framework of the NEA/NSC
SATIF
Task Force, a group has been created to organise an International
Benchmark Exercise on electron, positron and photon physics. The
purpose of this exercise is to constitute a group of experts in
general-purpose electron/photon Monte Carlo methods to design a
benchmark suite for comparison of existing public-domain Monte Carlo
codes (e.g. EGS4, ITS-3.0 or PENELOPE) or codes that are due to be
released later in 2000, namely EGS5, ITS-4 and others. Through use of
results of the benchmark suite, the community will have unprecedented
knowledge of the dynamic range of a given code, its range of
applicability and its accuracy. To this purpose, both theoretical and
experimental benchmark exercises are foreseen. Users will also be able
to gain some idea as to how efficient a code may be for certain
applications and some information as to the computing resources
required for certain broad classes of problems.
Reactor
surveillance and diagnostics, Reactor Noise
Nuclear Criticality Safety
Plutonium recycling
Safety: Thermal-Hydraulics
and CCVM
Dosimetry
Transmutation
Nuclear model
and code intercomparisons
JEF benchmarking and
Nuclear Data
See also the NEACRP archives for past benchmarks
Return to the NSC Home Page
