You are on page 1of 9

Nuclear Engineering and Design 344 (2019) 174–182

Contents lists available at ScienceDirect

Nuclear Engineering and Design


journal homepage: www.elsevier.com/locate/nucengdes

Thermal-hydraulics and neutronic code coupling for RELAP/SCDAPSIM/ T


MOD4.0
Sabahattin Akbasa, Victor Martinez-Quirogab, Fatih Aydoganc, , Chris Allisonb,

Abderrafi M. Ougouagd
a
Institute of Science, Gazi University, Milas No 15 Street, Teknikokullar, 06500 Ankara, Turkey
b
Innovative Systems Software 3585 Briar Creek, Ammon, ID 83406, USA
c
Department of Mechanical Engineering, Jacksonville University, USA
d
Idaho National Laboratory, Idaho Falls, ID, USA

ABSTRACT

The design and analysis of energy systems requires robust and reliable computer codes that produce realistic results. The governing equations and closure models
must represent the physical behaviors of the energy systems. Nuclear power plants, as the quintessential example of complex energy systems, must be modeled with
such high-fidelity and accurate nuclear system codes in which thermal-hydraulics, neutronics and fuel performance models must be coupled for in order to produce
high-fidelity predictions. Most commonly, the interplay between the various subsystems of a nuclear power plant (nuclear fuel, fluids, neutronics, control rods, power
conversion, etc.) is modeled through a combination of different codes that are externally coupled.
The objective of this paper is to present work that couples the RELAP/SCDAP SIM/MOD4.0 code with the NESTLE neutronic code within the 3DKIN package, thus
providing the latter with updated neutron kinetic capabilites. The version of the RELAP/SCDAPSIM/MOD4.0 code used in this work has been developed by Innovate
System Software (ISS) and is the latest in the series of such code versions developed as part of the international SCDAP Development and Training Program (SDTP) for
best-estimate analysis to model reactor transients including severe accident phenomena. The new feature of 3DKIN enhances the simulation of the Nuclear Power
Plants (NPP) response under accidental and operational scenarios in which high reliability of neutronic feedback is needed.
The RELAP/SCDAPSIM code is described and the improvements implemented in it are presented, including nodal kinetics library as well the coupling method that
is applied. Finally, a BWR transient with unexpected injection of subcooled water as specified in an OECD Benchmark has been simulated in order to assess the
reliability of the package and the performance of the coupling. Results are compared to those of the other OECD benchmark participants. The comparison of fuel
temperature, effective multiplication factor (keff) and power distribution results display good agreement with values within the range of those of other participants.

1. Introduction • fuel performance codes that simulate the nuclear fuel behavior
under irradiation.
The computational tools used in nuclear industry can simulate
complex nuclear systems. The nuclear systems include nuclear fuel, In the early decades of nuclear power, each of these types of codes
control rods, reactor internals, fluid systems, control systems, safety were developed and implemented individually. However, the strong
systems, energy conversion components (such as turbine), pump, con- interaction between the various physical phenomena of neutronics,
densers, preliminary heat exchangers, steam-seperators, valves, pipes reactor thermal hydraulics, fuel performance, containment physics, and
and other heat sources and heat sinks. To model all these systems and others, has been recognized to necessitate the coupling between these
components, some specific codes are utilized. For instance, some of the codes. Coipled codes capture the interplay between the various physics
following codes used commonly: phenomena and hence model the feedback between them. Feedback
effects can be observed in several transients in nuclear power plants.
• reactor physics codes for modeling core neutronics, cross-section Some example transients include:
preparation and/or lattice physics codes or cross section libraries,
• thermal hydraulic codes that encompass system codes, containment • Inadvertent control rod withdrawal (uneven feedback)
codes, computational fluid dynamics, sub-channel codes. All of these • Control rod ejection (strong local feedback)
can simulate the thermal and hydraulic behavior of the coolant and/ • Start-up of a cold or a boron free loop (uneven feedback)
or working fluid, • External asymmetrical boron dilution (uneven feedback)

Corresponding author at: Mechanical Engineering, Jacksonvile University, 2800 University Blvd N., Jacksonville, FL 32211, USA.
E-mail address: faydogan@gmail.com (F. Aydogan).

https://doi.org/10.1016/j.nucengdes.2019.01.009
Received 13 March 2018; Received in revised form 30 December 2018; Accepted 5 January 2019
Available online 08 February 2019
0029-5493/ © 2019 Published by Elsevier B.V.
S. Akbas, et al. Nuclear Engineering and Design 344 (2019) 174–182

Acronyms NGFM Nodal Green’s Function Method


NPP Nuclear Power Plant
ANM Analytic Nodal Method NK Neutron Kinetics
BE Best Estimate PWR Pressure Water Reactor
BWR Boiling Water Reactor SS Steady-State
FDM Finite Differential Method TH Thermal-Hydraulics
ISS Innovate System Software RS4 RELAP/SCDAPSIM/MOD4.0
SCDAP SCDAP Development and Training Program LWR Light Water Reactor
NEM Nodal Expansion Method LWRCT LWR Core Transient

• Transients with potential for inherent boron dilution (uneven complex configurations codes that incorporate TH and 3-D NK would be
feedback) required. Especially for full nuclear core applications integrated with
• Anticipated transients without scram (uneven feedback) primary and power conversion coolant cycles, detailed coupling be-
• Cool-down transients with re-criticality potential (steam or feed tween TH and NK codes is required for modeling strong interaction
lines break-uneven feedback) between NK and TH behaviors for steady-state and transients (design
basis accidents, anticipated transient without scram and severe acci-
In nuclear power plants there are strong feedback effects between dents). The following coupled codes are some of the existing codes used
thermal hydraulic (TH), fuel performance (FP), and neutron kinetics for the accurate and realistic evaluation of NK and TH behavior:
(NK) and static neutronic physical behaviors that require accordingly TRACE/PARCS (Xu et al., 2006), RELAP5/PARCS (Barber et al., 1999),
coupled codes. The temperature distribution in nuclear fuel is calcu- ATHLET/DYN3D (Kozmenkov et al., 2007) and RELAP/SCDAPSIM/
lated by the TH and neutronics codes. In turn, the temperature dis- MOD4.0/PARCS and NESTLE (Martinez-Quiroga et al., 2015, 2016).
tribution is used by the reactor physics codes since effective neutron The code prediction of this coupling method is also demonstrated for a
cross sections depend on the temperature distribution of fuel and Boiling Water Reactor (BWR) transient to demonstrate how developed
moderator. The NK code predicts the neutron flux and power dis- coupling works such as, TRACE/PARCS (Gajev et al., 2014), CATHARE-
tribution, which is used by the thermal hydraulic codes. The fuel per- CRONOS2-FLICA4 (Mignot et al., 2004), DYN3D/ATHLET (Grundmann
formance codes use information provided by both the TH and neu- et al., 2004), RELAP5/PARCS (Salah et al., 2005), TRAC-BF1/NEM/
tronics codes. Although there is a data exchange between all of these COBRA-TF (Solis et al., 2002).
codes, the coupling methods between them are different from one an- This paper presents a coupling of RELAP/SCDAPSIM/MOD4.0 and
other as the parameters that are exchanged vary and the time scales of NESTLE within 3DKIN package. The scenario of a BWR transient with
the various phenomena are substantially different. Thus, the coupling unexpected subcooled water injection was utilized to benchmark the
schemes can also vary substantially: for example some codes and si- code predictions with the other code predictions by using a OECD
tuations may require data exchanges only for certain specific para- benchmark.
meters that are not exchanged between other pairs of codes and under Section 2 shows a brief description of RELAP/SCDAPSIM/MOD4.0
other situations. Furhtermore, the exchanges have to occur with dif- and NESTLE in 3DKIN package. Section 3 describes the selected OECD
ferent frequencies (e.g., at every computational time step or not). scenario for this paper. Section 4 describes the code models and no-
To reduce the cost of the fully coupled codes using realistic feedback dalization diagrams. Section 5 demontrates the results of the code for
effects, some codes incorporate other code models to utilize limited the selected OECD bencmark.
feedback effects. For instance, NK codes include simple TH models and
some TH codes utilize basic 1-D NK modeling capability. Even though
2. Coupling of thermal-hydraulic code of RELAP5/SCDAPSIM/
these simple models for the representation of NK or TH does not pro-
MOD4.0 and 3DKIN package
vide realistic feedback effects, the combination of several coupled
models can provide simple treatment of feedback effects on the codes.
3DKIN is a nodal kinetics package that has been included into
However, inasmuch as the predictions of the neutronics models depend
RELAP/SCDAPSIM/MOD4.0 (RS4) as an option for RS4 users that do
strongly on fuel, coolant and moderator temperatures, and though for
not have their own 3D reactor kinetics package. 3DKIN package is
some problems NK with simple incorporated TH or TH codes with
based on models and correlations of NESTLE (Nodal Eigenvalue,
simple NK may suffice, for most realistic applications that involve
Steady-state, Transient, Le core Evaluator) 5.2.1, a NK code developed

Fig. 1. Diagram of 3DKIN series coupling.

175
S. Akbas, et al. Nuclear Engineering and Design 344 (2019) 174–182

at North Carolina State University (North Carolina State University, 3. Description of the selected BWR transient
2003). Code architect has been modified and reprogrammed in FOR-
TRAN 90/95/2000 standards to keep the implicit solution of the neu- In order to verify the programming of neutronic as well as the
tron diffusion equation as an internal library of RS4 (Fig. 1). New input proper coupled data exchange, a series of coupled calculations have
cards have been included in RS/M4 for user-friendly mapping of LWRs been carried out with RS4-3DKIN codes. Since international LWR core
Cartesian geometries (Hexagonal geometries are also possible to be si- transient (LWRCT) benchmarks, based on well-defined problems with a
mulated). Although it is not as powerful as the NESTLE code, it is a complete set of input data, are used to assess the discrepancies between
useful option for coupled calculations. three-dimensional space-time kinetics codes in transient calculations, a
3DKIN package solves few-group neutron diffusion equation uti- specific case of an OECD benchmark has been selected for this article.
lizing Nodal Expansion Method (NEM) for spatial discretization. It can The chosen scenario/case is a BWR transient with unexpected injection
simulate either eigenvalue or eigenvalue initiated transient problems, of subcooled water (Fig. 3) without scramming the reactor (case D1 of
thus they can be applied for simulating operation transients at constant the NEACRP LWR core transients benchmark (Finnemann and Galati,
power (by the use of the eigenvalue simulation at each time advance- 1992; Finnemann et al., 1993)). One of the goals of this transient is to
ment) or transient problems for an initial neutron flux distribution test the reactivity feedback effects of density variation and Doppler
(eigenvalue initiated transients). Depletion is also available for opera- broadening in core power distribution.
tional simulations. Two or four energy groups can be utilized for solving The selected benchmark provides results of eight national and in-
the neutron diffusion equation. NEM is based upon quartic polynomial dustrial institutions that will be used to evaluate the result of RS4-
expansion equations and a non-linear iterative strategy (outer-inner 3DKIN coupled simulations. Just as a reminder, the aim of this com-
iteration) is employed for solving the resulting matrix system. The parison is not to validate the accuracy of supplied sample input decks
utilized package solves either Nodal or Finite Differential Method and the neutronic code but to demonstrate 3DKIN reliability and to test
(FDM). Quartic polynomial expansion is employed for the transverse the performance by using a BWR nuclear power plant.
integrated fluxes, and transverse leakage is represented by a quadratic In a BWR reactor vessel (Fig. 4), there are a reactor core that mainly
polynomial. Furthermore, discontinuity factors are applied to correct consists of fuel assemblies and control rods in the center, equipment for
for homogenization errors. For transient problems, diffusion equation generating steam for a turbine, such as a steam-water separator and a
employs a first-order difference operator and precursor equations are steam dryer in the upper part of the vessel, equipment for reactor-
analytically solved assuming that fission rate behaves linearly over a power control, such as control rod guide tubes and control rod drive
time-step. In these problems, the delayed neutron precursor groups are housings in the lower part of the vessel, and a core shroud, jet pumps
specified by the user. As regards XS libraries, they are externally etc. that surrounds the reactor core and composes the coolant flow path
parameterized in macroscopic models that includes composition type, in the periphery of reactor core.
control rod state (in/out) and burnup (fuel depletion can be also
modeled for operational simulations). In addition, all cross sections are
expressed in terms of a Taylor series expansion depending on coolant 4. Modelling and simulation of the nuclear power plant
density, coolant temperature, effective fuel temperature and soluble
poison concentration. Finally, I-Xe and Pm-Sm chains are modeled with The selected BWR case for modelling and simulation consists of sub-
several options for calculating their number densities (i.e. equilibrium, prompt critical reactivity excursions generated by rapid cold water
transient, peak Sm- no Xe, no Sm nor Xe, or frozen), and decay groups injection or core pressurization events. The selected case has neutronic
are used to model decay heat. and thermohydraulic feedback-effects. This section demonstrates two
As regards neutron flux solution, FDM outer-inner iteration strategy main components of modelling that will be described in this section in
is employed taking advantage of the Chebyshev polynomials for ac- detail: (1) neutronic modelling based on 3DKIN module for the BWR
celerating the convergence. For NEM, a non linear iterative method is core and (2) thermal hydraulic modelling of RELAP/SCDAPSIM/
also applied after a sufficient number of outer iterations are completed. MOD4.0 for the BWR’s TH system.
Fig. 2 shows the structure of the iterative method. As it can be seen in
the diagram, for achieving the convergence it is required a continuous
updating of TH data after each outer iteration. Hence, it was decided to
define an implicit coupling between TH and neutronics for RS4 time
advancement in order to enhance the stability, accuracy and robustness
of the iterative method. With this approach, if there is no convergence
in the neutron flux solution, TH data are updated with the provisional
estimation of the power distribution without advancing to a new time
step. Once new TH values are computed, outer-inner iteration is re-
started at the same level and new cycle is run. This process is iteratively
repeated until convergence is achieved. After that, RS4 starts a new
time advancement following the same numerical scheme. As it can be
seen, the coupling is completely implicit because both neutronics and
TH data solutions are computed by the use of a coupled set of equations
that are function of the current and previous time step solutions esti-
mates.
In order to verify the programming and the integration of the 3DKIN
package within RS4, three testing problems were carried out based on
the sample XS libraries and core configurations supplied with NESTLE
neutronic model (Martinez-Quiroga et al., 2016). In our previous pub-
lication (published in NUTHOS conference), we have already studied
and published three cases, which are (1) Eigenvalue problem, (2) Ei-
genvalue initiated transient problem, and (3) Depletion problem. Cold
water injection transient without screaming the reactor core is selected
as the validation case in the scope of this paper. Fig. 2. Diagram of the NEM non-linear outer inner iteration method.

176
S. Akbas, et al. Nuclear Engineering and Design 344 (2019) 174–182

neutron flux (a measure of the neutron population) as a consequence of


the neutrons interaction with the fuel and moderator (and other ma-
terials) within the reactor core. The computed value of the flux, in turn,
is used to calculate reaction rates, including the volumetric rate at
which fission reactions take place and hence the power density dis-
tribution, and essential input to safety analysis modeling.
The governing equations solved by NK codes are the time-dependent
multi-group neutron diffusion equations and their associated delayed
neutron precursor equations (Duderstadt and Hamilton, 1976; Bandini,
1990).
For light water applications most NK codes use two neutron energy
groups, though occasionally as many as four groups may be used. In this
study, two neutron energy groups (thermal and fast neutron groups)
Fig. 3. Inlet subcooling vs. time for D1 transient (Finnemann and Galati, 1992; and six delayed neutron groups are used for neutron modeling in the
Finnemann et al., 1993). calculations. 3D cross-section (transport, absorption, nu-fission, kappa-
fission, nu) values (XSs) of NEACRP LWR core are used in the selected
4.1. Neutronics modelling transient benchmark (Finnemann and Galati, 1992; Finnemann et al.,
1993)
In coupled NK-TH codes (or code systems), the NK part, which The 3D NK model has been prepared based on the 3D cross-sections
computes the neutron flux as a function of time (and space in most given for the selected OECD benchmark. The cross-sectional view
instances), uses the neutron diffusion equation as a model for the (based on Xand Y dimensions) of the BWR reactor is shown in Figs. 5
physics. Furthermore, most of coupled NK-TH codes use one of the and 6. The full core consist of ten different fuel macroelement com-
various modern nodal methods as their neutronic solver for the multi- positions (as shown in Fig. 5) and 19 composition types. The length of
dimensional, multi-group neutron diffusion equations (Lawrence, 1986; each square fuel macroelement and macrocell in the radial reflector, is
Sutton and Aviles, 1996). These nodal methods such as Nodal Expan- 30.48 cm. Axially, the reactor is subdivided into 14 layers, each
sion Methods (NEM) (Finnemann, 1975), Analytic Nodal Method 30.48 cm high, as shown in Fig. 7. The fuel composition of each axial
(ANM) (Smith, 1979), the Nodal Green’s Function Method (NGFM) layer is individually defined (Fig. 7).
(Lawrence 1979), and the more efficient Semi-Analytic Nodal Method
(Rajic and Ougouag, 1989), have been implemented using different 4.2. Thermal-hydraulics model
approaches in current NK codes. Neutron kinetics (NK), or time-de-
pendent neutronics, deals with modeling of time evolution of the RELAP/SCDAPSIM/MOD4.0 is a Best Estimate (BE) code designed

Fig. 4. Typical BWR reactor system (Roth and Aydogan, 2015).

177
S. Akbas, et al. Nuclear Engineering and Design 344 (2019) 174–182

Fig. 5. BWR initial map.

Fig. 7. BWR macroelement type 1.

MOD3.2 models, developed by the US Nuclear Regulatory Commission,


in combination with advanced numerics, advanced programming,
techniques and SDTP member-developed models and user options
(more detailed information can be found in reference (Pérez et al.,
2015)). In particular, RELAP/SCDAPSIM/MOD4.0, the latest in the
series of the SDTP-developed versions, is the first version of RELAP or
SCDAP/RELAP5 completely rewritten to FORTRAN 90/95/2000 stan-
dards. It implies that RELAP5 database is dynamically allocated,
making easier and faster the coupling with other codes with the use of
pointers and structures. The input model (Fig. 8) of RELAP/SCDAPSIM/
MOD4.0 to simulate the TH behaviour of the the core and the BWR
system has been prepared based on the selected OECD benchmark. The
channels (from 1 to 4) shown on the right side of Fig. 8 are the re-
presentative channels for the channels of Fig. 6.
The core has been simulated with 4 concentric channels linked to 4
different active fuel heat structures, and a bypass. It has also been in-
cluded for simulating the peripheral region. As regards -3DKIN and
RELAP/SCDAPSIM model, main features of the supplied decks are given
in Table 1.
Fig. 6. BWR core map.
5. Results

to run a wide range of conditions from normal operating conditions up


NK and TH modelling and the developed coupling codes have been
through severe accidents (Allison and Hohorst, 2010). RELAP/SCDAP-
demonstrate in the previous sections of the paper. In this section, the
SIM uses the publicly available RELAP5/MOD3.3 and SCDAP/RELAP5/
results will be demonstrated for the selected OECD Benchmark. Even

178
S. Akbas, et al. Nuclear Engineering and Design 344 (2019) 174–182

Fig. 8. RS4 nodalization.

though the selected OECD benchmark does not demonsrate all the re- change from 0.952429 to 0.99999.
sults of TH and NK parameters, the benchmark of OECD participants For the steady state calculation, the axial and radial power dis-
were given based on selected parameters of NK and TH. Firstly, the keff tributions are compared with the results given in the OECD benchmark
results of the developed coupled code is benchmarked with the OECD (Figs. 9 and 10). The cross-sectional modelling of NK code can be re-
participants for the steady state BWR core results in Table 2. The keff visited to improve the top-half results of normalized axial power. Even

Table 1
Main design parameters of RS4-3DKIN for core modeling (Finnemann and Galati, 1992; Finnemann et al., 1993).
Parameters Value

Neutronic Number of fuel rods 196


Number of reflector elements 64
UO2 density 10.412 g/cm3
Zircaloy density 6.6 g/cm3
Outer clad diameter 1.430 cm
Inner clad diameter 1.267 cm
Pellet diameter 1.237 cm
Fuel rod pitch 1.875 cm
Reference effective fuel temp. 573.15 K
Number of composition types 19 (10 fuel XS macro elements)
Fast neutron inverse velocity 3.58 × 10−8 cm−1/s
Thermal neutron inverse velocitv 2.27 × 10−6 cm−1/s
Prompt energy release per fission 3.20 × 10−11 J/fission
Total delayed neutron fraction 0.00760

Thermal-hydraulic Core thermal power 1800 MW


Total inlet mass flow rate 13,000 Kg/s
Core pressure 67.0 bar
Core inlet subcooling 46.52 KJ/Kg
Core configuration 17 × 17 × 14 (30.48 cm × 30.48 cm × 30.48 cm)
Reference coolant density 47.627 lb/ft3

179
S. Akbas, et al. Nuclear Engineering and Design 344 (2019) 174–182

Table 2
Keff results.
ARROTTA DYNAS TNK-XC KICOM QUANDRY-EN STAND QUABOX RS4-PARCS RS4-3DKIN

keff 0.99999 0.98563 0.98764 0.9844 0.96788 0.991 0.98639 0.961342 0.952429

Fig. 9. Axial core power distribution.


Fig. 11. Core power for transient problem.

Fig. 10. Radial distribution of core averaged densities for steady state problem.

Fig. 12. Core averaged fuel temperature difference with respect to steady-state.
though our coupling results are in the range of other code predictions
range, the top half axial power can be improved by using more detailed
reliability of RS4-3DKIN coupling.
NK models.
After steady state runs, transient runs for cold water injection (as
Conclusions of the benchmark point out TH modelling as the main
described in section 2) have been performed. Time dependent problem
factor of the dispersed results for the SS power distributions. It is also
was executed in a Intel i5-640 PC (3,3 GHz and 4 GB of RAM) with
worth to mention that supplied RS4 radial core configuration for TH
Windows 7 64 bits OS. RS4 was compiled with Intel Fortran 2016 XE
was significantly simple (4 concentric channels for 10 fuel composition
Composer 32 bits compiler. The total CPU time was 70.28 s for a si-
types and 185 neutronic nodes) in comparison with the standalone si-
mulation transient of 30 s. The time step used for the whole simulation
mulations of the participants (one fluid channel per each fuel bundle).
was 2.5·10−2 s
With that modelling, no one-to-one TH/NK feedback exists for each fuel
With respect to the simulation, transient problem starts with a
bundle composition type, and greater deviations in criticality and keff
sudden injection of subcooled water at core inlet. For these conditions,
value must be expected. Fig. 10 shows how the limited number of radial
density is suddenly reduced also increasing the thermal neutrons
TH channels affects to the reactivity feedback. RS4-3DKIN simulation
moderation and core power. A maximum peak (165.6% of the initial
results of average density are relatively similar to the average results of
power) is achieved 1.8 s after the initiation of the water injection
the benchmark participants. On the contrary, results for the axial dis-
(Fig. 11). As a result of the power excursion, fuel average temperature
tribution of the core power showed a good agreement (see Fig. 9). This
starts to increase (Fig. 12). It causes a negative reactivity feedback
was due to a finer meshing coupling between RS/MOD4.0 and 3DKIN
because of Doppler broadening dependency that compensates the effect
(12 TH and NK nodes for the active fuel bundles). Therefore, the re-
of the water subcooling and stops the power excursion. Finally core
ported discrepancies are related with the TH modeling and not with the

180
S. Akbas, et al. Nuclear Engineering and Design 344 (2019) 174–182

Fig. 16. Difference of radial relative power (Δ = t0-t20).

Fig. 13. Radial power at t = 0 s (transient, initial).


similar trends for the increasing of the fuel temperature. The trend of
fuel-temperature change depend on several parameters: the detailed
modelling of fuel bundle, fuel rods and the physical modelling of the
core which effect the convective and conductive heat transfer char-
acteristics as well as heat capacitance of the fuel and non-fuel elements.
Very detailed comparison is necessary to figure out all these models
among these codes but the benchmark used does not provide detailed
information to clarify which parameter is the dominant on for defining
why RS4-3DKIN has more fuel temperature than other ones.
Figs. 13–16 demonstrates the difference of the radial power of the
reactor core for different time of the transient.

6. Conclusions

In this study, RELAP/SCDAPSIM/MOD4.0 and NESTLE codes are


coupled. The code results have been benchmarked with other codes by
using a OECD benchmark.
3DKIN package including NESTLE has been demonstrated as the
Fig. 14. Difference of radial relative power (Δ = t0-t5). neutronic interface between NK and TH modesuls to simplify the cou-
pling process. This coupling allows usre friendly coupling TH/NK cal-
culations. Steady-state and transient (with unexpected injection of
subcooled water) problems in OECD Benchmark have been simulated.
Our results have been compared with the benchmark results of other
seven codes.
RS4-3DKIN simulation results of average density are higher than
other benchmark participants at the mid-section of the radial core since
only 4 representative core channels have been used. The top axial core
power distribution is higher than other benchmark calculations even
though our code predictions are between other code prediction range.
Even though our coupling results are in the range of other code pre-
dictions range, the top half axial power can be improved by using more
detailed NK models. Despite the differences between the results of
OECD benchmark’s participants and the limitations of the meshing
modeling in the supplied RELAP5 input decks, RS4-3DKIN results
showed that the transient results are in the other code results range.
Continuation of the validations of RS4-3DKIN coupled code will de-
monstrate more code capabilities for other validation cases in the fu-
ture.
Fig. 15. Difference of radial relative power (Δ = t0-t10).

power is stabilized around the 132% of the initial core power at Appendix A. Supplementary data
t = 11 s.
The code of ARROTTA has decreasing fuel temperature. Some of the Supplementary data to this article can be found online at https://
codes have stable fuel temperatures at the last part of the transient. doi.org/10.1016/j.nucengdes.2019.01.009.
The codes of RS4-3DKIN, TNK_XC, STAND and QUANDRY-EN have

181
S. Akbas, et al. Nuclear Engineering and Design 344 (2019) 174–182

References Japan, 2015.


Martinez-Quiroga, V., Akbas, S., Aydogan, F., Ougouag, A.M., Allison, C., 2015. Coupling
of RELAP5-SCDAP Mod4.0 and Neutronic Codes. IMECE2015-52991, ASME 2015
Allison, C.M., Hohorst, J.K., 2010. Role of RELAP/SCDAPSIM in nuclear safety. Sci. International Mechanical Engineering Congress & Exposition, Houston, Texas,
Technol. Nucl. Installations 172010. https://doi.org/10.1155/2010. 425658. November 13–19, 2015.
Bandini, B., 1990. A Three-dimensional Transient Neutronics Routine for the TRAC-PF1 Martinez-Quiroga, V., Allison, C., Wagner, R.J., Aydogan, F., Akbas, S., 2016. NIRK3D
Reactor Thermal Hydraulic Computer Code. Ph.D. Thesis. Pennsylvania State and 3DKIN: General Description and Current Status of the New 3D Kinetics
University, Pennsylvania. Capabilities of RELAP5/SCDAPSIM/MOD4.0. The 11th International Topical Meeting
Barber, D., Miller, R.M., Joo, H., Downar, T., Wang, W., Mousseau, V., Ebert, D., 1999. on Nuclear Reactor Thermal Hydraulics, Operation and Safety, Gyeongju, Korea,
Coupled 3D-Reactor Kinetics and Thermal-Hydraulic Code Development Activities at October 9–13, 2016.
the U.S Nuclear Regulatory Commission. LA-UR-99-1208, M&C;99, Spain, Madrid. Mignot, G., Royer, E., Rameau, B., Todorova, N., 2004. Computation of a BWR turbine
Duderstadt, J.J., Hamilton, L.J., 1976. Nuclear Reactor Analysis. Wiley Interscience, New trip with CATHARE-CRONOS2-FLICA4 coupled codes. Nucl. Sci. Eng. 148, 235–246.
York. North Carolina State University, 2003. NESTLE Few-group Neutron Diffusion Equation
Finnemann, H., Galati, A., 1992. NEACRP-L-335: 3-D LWR Core Transient Benchmark Solver Utilizing the Nodal Expansion Method for Eigenvalue, Adjoint, Fixed-source
Specification, NEACRP-L-335 (revision 1), OECD Nuclear Energy Agency, 1992. Steady-state and Transient Problems, NC 27695-7909, North Carolina State
Finnemann, H., 1975, A consistent nodal method for the analysis of space-time effects in University, Raleigh, 2003.
large LWRs, in: Proc. of the Joint NEACRP/CSNI Specialists Meeting on New Rajic, H.L., Ougouag, A.M., 1989. ILLICO: a nodal neutron diffusion method for modern
Developments in Three-Dimensional Neutron Kinetics and Review of Kinetics computer architectures. Nucl. Sci. Eng. 103 (4), 392–408.
Benchmark Calculations, NEACRP/CSNI, Munich, pp. 133–172. Roth, G.A., Aydogan, F., 2015. Derivation of new mass, momentum, and energy con-
Finnemann, H., Bauer, H., Galati, A., Martinelli, R., 1993. Results of the LWR Core servation equations for two-phase flows. Prog. Nucl. Energy 80, 90–101.
Transient Benchmarks, NEA/NSC/DOC(92)25, AEN/NEA, 1993. Salah, A.B., Galassi, M.G., D’Auria, F., Koncar, B., 2005. Assessment study of the coupled
Gajev, I., And, Ma.W., Kozlowski, T., 2014. Sensitivity analysis of input uncertain para- code RELAP5/PARCS against the Peach Bottom BWR turbine trip test. Nucl. Eng. Des.
meters on BWR stability using TRACE/PARCS. Ann. Nucl. Energy 67 (2014), 49–58. 235, 1727–1736.
Grundmann, U., Kliem, S., Rohde, U., 2004. Analysis of the boiling water reactor turbine Smith, K.S., 1979. An Analytic Nodal Method for Solving the Two-group,
trip benchmark with the codes DYN3D and ATHLET/DYN3D. Nucl. Sci. Eng. 148, Multidimensional, Static and Transient Neutron Diffusion Equations. MS. Thesis.
226–234. Massachusetts Institute of Technology, Cambridge.
Kozmenkov, Y., Kliem, S., Grundmann, U., Rohde, U., Weiss, F.-P., 2007. Calculation of Solis, J., Avramova, M.N., Ivanov, K.N., 2002. Temporal adaptive algorithm for TRAC-
the VVER-1000 coolant transient benchmark using the coupled code systems DYN3D/ BF1/NEM/COBRA-TF coupled calculations in BWR safety analysis. Ann. Nucl. Energy
RELAP5 and DYN3D/ATHLET. Nucl. Eng. Des. 237, 1938–1951. 29, 2127–2141.
Lawrence, R.D., 1979. A Nodal Green’s Function Method for Multidimensional Neutron Sutton, T.M., Aviles, B.N., 1996. Diffusion theory methods for spatial kinetics calcula-
Diffusion Calculations. Ph.D. Thesis. University of Illinois, Urbana. tions. Prog. Nucl. Energy 30 (2), 119–182.
Lawrence, R.D., 1986. Progress in nodal methods for the solution of the neutron diffusion Xu, Y., Downar, T., Ward, A., Kozlowski, T., Ivanov K., 2006. Multi-physics coupled code
and transport equations. Prog. Nucl. Energy 17 (3), 271–301. reactor analysis with the U.S. NRC code system TRACE/PARCS, PHYSOR-2006, ANS
Pérez, M., et al., 2015. The development of RELAP/SCDAPSIM/MOD4.0 for advanced Topical Meeting on Reactor Physics, Vancouver, BC, Canada.D, 2006.
fluid systems design analysis. INCONE23-1623. In: Proceeding of INCONE-23, Chiba,

182

You might also like