atw Vol. 63 (2018) | Issue 8/9 ı August/September
OPERATION AND NEW BUILD 450
There are three possibilities for an
excitation in general:
• Stochastic fluid forces from turbulent
flow would lead to oscillations
of components at their natural
frequencies. The lowest natural
frequency of the fuel assemblies is
reported around 2.6 to 4 Hz [6, 18],
which would not explain the coherence
maximum at 1 Hz. Calculations
with simplified finite element
models show that depending on
design and operational behavior,
i.e. lateral stiffness decrease due to
radiation induces spring relaxation
in the spacers, the lowest natural
frequency can be shifted significantly
towards lower values. In [6,
16] an additional mode of the fuel
assemblies around 1 Hz in form of
synchronously moving cantilevered
beams (fixed at the bottom) is supposed.
Nevertheless, regarding the
fixture of the fuel assemblies in the
grid plate, the manifestation of this
mode is questionable. A further
explanation is the excitation of the
coupled system of core barrel, grid
plate and fuel assemblies, which
might have additional natural
system frequencies below the natural
frequencies of the single fuel
assemblies.
• A second possibility would be the
existence of an excitation force,
which is oscillating at around 1 Hz
and evokes a subsequent transient
deflection of the fuel assemblies.
Pressure fluctuations from residual
imbalances of the coolant pump,
standing waves, cavity resonances
in the pressurizer or vibrations of
other components of the loop are
known to induce core barrel motions
which could propagate to the
fuel assemblies. Fluid mechanical
oscillating forces with direct effect
on the fuel assemblies, e.g. pressure
differences, are also possible.
• A third possibility would be a selfexcitation
of fuel assemblies in a
constant axial flow. Research on
fuel assembly bow gives hints that
in fluid-structure-interaction (FSI)
simulations local forces can arise
leading to instability of the zero
position of the fuel assembly [19].
To investigate and prove the mentioned
hypotheses, a coupled FSI
model of core components and the
surrounding fluid is essential.
Simulations of reflector
influence
Further, the reflector influence has
been studied by means of a simplified
2D core model, in which the reflector
Case description
Maximum (relative)
increase on the left side
cross-sections are manipulated in
order to simulate the effect of varying
water gap between core barrel and
reactor pressure vessel, which corresponds
to the reflector region. These
variations could be caused by mechanical
motions, e.g. of core barrel or
fuel assemblies at the core periphery,
and their effect increases with decreasing
boron concentration. In this
model the TH parameters are homogeneous
and representative of the
hot full power state at zero burnup.
Further assumptions are: fuel temperature
= 900 K, moderator density =
702 kg/m 3 and boron concentration
= 1,300 ppm.
Table 1 summarizes the results
obtained for different variations of
the thermal absorption and fast-tothermal
scattering crosssection. The
reflector is modified only in one half of
the core (the left side) to reproduce
the spatial oscillations observed in the
PWR. The results show that the effects
of thermal absorption and scattering
are additive. The amplitude of the
power variation can reach the same
order of magnitude as observed in the
PWR.
Additional study is necessary to
determine if actual mechanical motions
can cause such changes leading
to increase/decrease of the moderator
volume (coolant water) in the reflector
zone and in that way changing
the homogenized assembly crosssections.
In addition, time-dependent
simulations are needed to check if
the frequency observed in the PWR
can be reproduced. Nevertheless, this
preliminary result shows that this
hypothesis is very promising. The
recently published study [20] showed,
that a variation of the gap size
between fuel elements of about one
centimeter can result in changes
of the neutron flux amplitudes at
the ex-core detectors of up to the
order of magnitude of 10 %. Therefore,
the influence of mechanical
motions of the fuel elements relative
to each other and as an ensemble
relative to the reflector cannot be
ruled out as explanation of the observed
neutron flux oscillations.
Summary and outlook
Several models based on single
physical effects (TH fluctuations at
the core inlet, movement of a point
source, coupled oscillations of core
internals, changes in the reflector
coefficients) are used to simulate the
neutron flux. Each of these simple
models can reproduce some of the
characteristics of the observed neutron
flux fluctuations but does not
encompass all features observed in a
real reactor. This suggests that further
work on the combination of models
is needed. Thereby, the biggest challenges
will lie in FSI simulations of
fuel assemblies including further core
internals, neutron physics simulations
using time-dependent geometries,
and possibly the coupling of all three
physical models.
Acknowledgment
This work has been performed in the
framework of the German Reactor
Safety Research and was funded by
the German Federal Ministry for
Economic Affairs and Energy (BMWi,
project no. RS1533). The authors
would like to thank the operators of
one German Vorkonvoi PWR and one
Konvoi PWR for providing data of
neutron flux measurements.
References
Maximum (relative)
decrease on the right side
-10 % thermal absorption 4 % -3 %
-10 % scattering 7 % -5 %
-10 % thermal absorption
-10 % scattering
10 % -8 %
-20 % thermal absorption 11 % -7 %
-20 % scattering 14 % -11 %
| | Tab. 1.
Summary of the reflector study results.
1. M. Seidl et al., Review of the historic
neutron noise behavior in German
GWU built PWRs, Progress in Nuclear
Energy 85, pp 668-675, 2015.
2. Reaktor-Sicherheitskommission,
Stellungnahme DWR-Neutronenflussschwankungen,
457. Sitzung vom
11.04.2013.
3. Bundesamt für Strahlenschutz,
Kurzbeschreibung und Bewertung der
meldepflichtigen Ereignisse in Kernkraftwerken
und Forschungsreaktoren
der Bundesrepublik Deutschland im
Zeitraum Januar 2011, Stand
14.12.2012.
Operation and New Build
Analyses of Possible Explanations for the Neutron Flux Fluctuations in German PWR ı Joachim Herb, Christoph Bläsius, Yann Perin, Jürgen Sievers and Kiril Velkov