atw Vol. 63 (2018) | Issue 8/9 ı August/September
pictured. Typically, four control rods
are required for reactivity control in
TRIGA reactors with thermal power
levels of less than 1 MW [IAEA2016B].
Further, graphite elements are at the
outer positions. For MTR research
reactors, there are often empty places
at the centre of the core grid for
radiation samples. If the input number
of fuel assemblies or elements
does not match the number of grid
positions, the implemented algorithm
considers these typical core characteristics.
The assemblies or elements are
positioned in respect of this information.
Furthermore, the free flow path
is calculated as a function of total core
area and number of fuel assemblies,
elements and other components
inside the research reactor core.
3 Generated input decks
of exemplary MTR and
TRIGA reactors
In this part, first functionality of
the new modelling system is demonstrated
by generating an exemplary
MTR and TRIGA research reactor
model. For this purpose, two reference
research reactors were chosen.
Providing technical details in
[ABD2008A] and comparative data in
[ABD2008B], the ETRR-2 was identified
as a MTR reference facility. The
ETRR-2 is a multipurpose research
reactor located in Inshas, the Arab
Republic of Egypt. It corresponds to
the rightmost research reactor design
in Figure 2-1. The ETRR-2 reactor
consists of 29 fuel assemblies of MTR
type with 19 fuel plates each and has
22 MW nominal power. Further
description is presented in [ABD2008].
The main nodalisation of the generated
ETRR-2 model in ATHLET is
pictured in Figure 3-1. On the left
side, the coolant loop is presented
in bright blue. The reactor pool is
modelled with two pipes interconnected
by cross-connections. The
inner pool pipe is connected to the reactor
chimney, which is marked in
brown, by a single junction pipe. The
reactor core is modelled with two
representative assemblies. Each is
composed of 18 core cooling channels.
One assembly is representing 28
grouped average assemblies. The
other assembly considers a hot channel
factor on the 19 fuel plates plus
one extra penalised fuel plate. The
nodalisation of both assemblies is
identically and shown in Figure 3-2.
To check the capability of the
nodalisation to reproduce the thermal
hydraulic plant conditions, steady
state calculations were performed.
| | Fig. 2-5.
MTR core layout (left) and TRIGA core layout (right), generated by software for input deck generation.
| | Fig. 3-1.
Overview of whole Nodalisation of the ETRR-2 (left) and one fuel assembly (right) with 18 core channels generated by the software
for input deck generation.
Power
[MW]
Loop mass
flow
[kg/s]
The initial conditions of the experiment
and the calculated parameters
are compared in Table 3-1. The
experiment was performed at 9.5 MW
thermal power. There is good agreement
between the calculated and
experimental stationary data.
As an exemplary TRIGA research
reactor, the IPR-R1 was identified.
The IPR-R1 is a TRIGA Mark I model,
installed in Belo Horizonte in Brazil
and operated since 1960. Several
analytic and experimental studies
were performed and published. As
reference data, experimental results
in [REI2009] were used. The IPR-R1
corresponds to the leftmost research
reactor design in Figure 2-1. It is
operating at 250 kW and consists of
63 fuel elements of TRIGA type.
Further description is presented in
[REI2009]. The main nodalisation of
the generated IPR-R1 model in
ATHLET is shown in Figure 3-2. On
the left side, the coolant loop is
Core mass
flow
[kg/s]
Core outlet
temperature
[°C]
Core pressure
drop
[bar]
| | Tab. 3-1.
Overview of whole Nodalisation of the ETRR-2 (left) and one fuel assembly (right) with 18 core channels generated by the software
for input deck generation.
presented in bright blue. The reactor
pool is modelled with two pipes interconnected
by cross-connections. The
inner pool pipe is connected to the
core entrance and core outlet. 13 core
channels, interconnected by crossconnections,
with 63 fuel elements
represent the reactor core (see Figure
3-2 right). The core nodalisation
based on the nodalisation presented
in Figure 2-4.
The experiment was performed at
50 kW thermal power. In Table 3-2,
the calculated steady state results are
compared to measured core inlet and
outlet temperatures. At different
positions, measuring devices were
installed (see [REI2009]). There are
small deviations but overall the results
are consistent.
Further, the ATHLET simulation is
compared to published RELAP steady
state calculation in [REI2009], which
reaches steady state conditions after
about 2000 s simulation time.
Reference
pressure
[bar]
Calculation 9.5 309.24 302.86 35.01 0.42 2.2
Reference
[ABD2015]
9.4 309.24 302.87 34.9 0.31* 2.0
*in [ABD2015] core pressure drop of 3.1 bar is mentioned, but in /IAEA2005/ 0.6 bar pressure drop at 100 % core power is referred
467
AMNT 2018 | YOUNG SCIENTISTS' WORKSHOP
AMNT 2018 | Young Scientists' Workshop
Heuristic Methods in Modelling Research Reactors for Deterministic Safety Analysis ı Vera Koppers and Marco K. Koch
Change language