APPLICATION OF THE RELAP5/Pb-Bi CODE TO SAFETY STUDIES
FOR THE ADS MYRRHA FACILITY
S. Heusdains and B. Arien
SCK-CEN
Boeretang 200, B-2400 Mol, Belgium
[email protected] or
[email protected] Abstract
RELAP (Reactor Excursion and Leak Analysis Program) is a code developed by Idaho National
Engineering Laboratory (INEL) for best estimate transient simulations of LWR coolant systems in a
broad range of normal or accidental situations such as LOCAs (Loss Of Coolant Accidents ).
The code is widely used for high pressurised reactors PWR (150 bars, 300C) and BWR (70 bars,
280C) safety studies for which it has been validated. A version bis of the code with the Pb-Bi as
coolant was developed by Ansaldo Nucleare (Genova, Italy) and we applied it to simulate the
behaviour of the ADS MYRRHA core in the framework of safety studies.
The paper describes this model and provides first calculation results in steady state conditions.
1073
Introduction
This paper concerns the modelling of the core of the MYRRHA facility in its pre-design phase.
The model was built to allow thermohydraulic calculations with a version of RELAP5.3.2 for
the coolant eutectic Pb-Bi. The version 3.2 of RELAP5 which normally deals with water as coolant,
was modified by the ANSALDO Nucleare (Genova, Italy).
The description of a partial model is showed here as well as first calculation results corresponding
to the nominal core operation.
To perform safety studies, the model has to include pumps, heat exchangers and control elements.
This work is still in progress.
Model description
In a first step this partial model is used to perform the simulation in steady state conditions and to
analyse the influence of some design parameters on the temperature and flow distributions in the core.
More particularly it addresses:
the thermohydraulic behaviour in a hot, a medium and a cold channels (see the
definition below);
the heat exchange between the core and the vessel;
the physical parameters at the core outlet;
the flow circulation in the medium plenum;
the total power and power distribution effects on the thermohydraulic behaviour inside and
outside the core.
At the present stage, the model does not include the pumps, the heat exchangers, the vessel and
the vessel structures. The spallation loop is partially modelled, only the part in the core is represented
(the pump and heat exchanger of the spallation loop are not included).
The left part of Figure 1 given below represents the general sketch of the MYRRHA primary
circuit and the right figure corresponds to the equivalent RELAP model of the core.
The RELAP building blocks showed on the right Figure can be divided into four fundamental
groups: thermal-hydraulic, heat structures, trips, and control variables.
The thermal-hydraulic group is composed of elements designed to simulate fluid passages and
fluid-handling equipment.
The volumes (rectangular boxes) represent the parts of the flow in which temperature,
pressure and phase concentrations are considered to be uniform. Temperature and pressure
are the average values in a volume taking into account an average flux and the heat transfer
between the structures and the flow. The type of flow in the volume is chosen in function of
some criteria (velocity.) calculated on the basis of these average parameters.
1074
Figure 1. Comparison between a sketch of the MYRRHA primary circuit (a)
and the equivalent RELAP model of the core (b)
a. b.
Junctions (arrows) connect two adjacent volumes and represent the pressure drop and the
mass flow rate.
Heat structures are designed to simulate solid structures and the interactions between the
wall structures and the fluid flowing along it. They simulate the behaviour not only of the
core fuel rods in a reactor system, but also the various plant structures
The heat structures are represented by slabs, which may be finely nodalised to provide a
rather detailed temperature distribution in one dimension. Plane, cylindrical or spherical
structures are allowed for any slab. The code assumes that energy flow to and from the heat
structures is in a direction normal to the stream-tube flow direction. Consequently, the heat
structure nodes are aligned in the direction normal to the fluid flow. The radial and axial
distributions of the flux are introduced with each structure if necessary.
Special components such as pumps, accumulators, tanks or valves are modelled separately.
Trips are designed to simulate the signals that initiate equipment actions of various sorts
(e.g., turning on a pump at a desired time or causing a valve to open at one pressure but to
close at another pressure).
Control systems are designed to give the code modelling added capability by allowing
equipment control systems (e.g., proportional- integral-differential controllers and lead-lag
controllers) and lumped-node systems to be simulated.
These three element types are not yet included in the model. They are necessary to simulate
transients behaviour for safety studies.
1075
Core model
Description of an assembly
Each assembly is vertically subdivided in 9 sub-volumes:
2 vertical sub-volumes for the assembly inlet;
5 vertical sub-volumes for the active fuel length;
2 vertical sub-volumes for the assembly outlet.
The hydraulic parameters are calculated in each sub-volume. Figure 3 shows the correspondence
between the axial zones in an assembly and the axial zones in the core. The inert parts of the fuel are
included in the top and bottom structures of the assembly. The active structure (fuel length) is
modelled such as in a PWR core model, i.e. in a half cylindrical geometry with conservation of the
area exchanging heat with the flow in the adjacent hydrodynamic volume. The centre of the fuel is an
axis of symmetry and the power is distributed in the 5 fuel sub-volumes via axial and vertical power
factors (see Figure 2 below ).
Figure 2. Correspondence between the axial zones in an assembly (a),
the length of volumes in the equivalent pipe (b) and the axial zones in the core (c)
a b c
1076
Figure 3. Core fuel distribution
Description of a group of assemblies
The sub-critical core of MYRRHA with 45 assemblies is subdivided into 3 zones (or groups): hot
(3 assemblies), medium (21 ass.) and cold channel (21 ass.), as shown on Figure 4. Each group is
modelled by one representative channel supposed to reproduce the average behaviour of the
corresponding group.
A group of n assemblies is simulated by only one pipe, (group of adjacent volumes) and
3 structures.
The pipe corresponds to a channel having a coolant flow equivalent to the flow of the n
assemblies in which the flow velocity and the pressure drop are kept. (left Figure 2).
The inert metallic parts of the assemblies in a group are simulated by 3 structures:
the wall of the assemblies, in contact with the core by-pass flow at the outer side and with
the assemblies cooling flow at the inner side;
the top grids including inert parts of the fuel elements in contact with it;
the bottom grids including inert parts of the fuel elements in contact with it.
The third structure represents the total active fuel parts of the n assemblies in which the power is
distributed according to the explanation given in chapter 4. The core by-pass is the volume between
the assemblies and the annular free volume around the core assemblies. It is represented by a dummy
pipe-volume.
Power distribution
The model is frozen with the configuration described in Figure 3. The total power is radially and
axially shared by means of distribution factors defined for each radial and axial nodes.
We have therefore 3 groups with 5 sub-structures for each one, i.e. in total 15 sub-structures
characterised every one by its group number g and its axial position i. The partial power Pgi
delivered in the sub-structure gi is equal to:
1077
Pgi = a g a gi Ptot where Ptot is the total power of the core (38.8 MW), ag is a radial distribution factor
agi is an axial distribution factor. The values of the ag and agi are given in Table 1. In this table the
ai factors are over-all axial distribution factors defined by:
3 3 5
ai = a
g =1
g a gi as a = a
g =1
g
i =1
gi = 1,
3 5
we may also write: Pgi = Ptot
g =1 i =1
Table 1. Radial and axial power distribution factors
Finally the fuel region of each sub-structure is cut in 4 radial intervals with a uniform power
generation in it. It means that 25% of the sub-structure power is delivered in each interval. Two radial
intervals are added to represent the gap and the cladding thickness respectively.
These radial node intervals are showed on Figure 4.
One should remark that all the assemblies within a group (hot, medium or cool) have the same
core-radial coefficient distribution.
Figure 4. Radial nodes in the fuel
1078
Modelling of the plena
The lower plenum is represented by the volume in the vessel under the core. It receives the flow
from the pumps and it supplies the core in Pb-Bi. Some calculations with the FLOW 3D CFD code are
in progress and will give information on the flow circulation in this volume. Later the lower plenum
will be remodelled by taking account this information.
The volume of fluid around the core chassis between the low assemblies level and the medium
level of the diaphragm is the medium plenum. In this volume free convection is induced by the heat
exchange with the core chassis. To simulate the fluid motion, this plenum must be modelled by
2 parallel channels interconnected via cross-flows. Until the modelling of the primary loop is
completed, a small fictitious cooling structure is introduced to remove the heat and to allow the free
convection.
The upper plenum is represented by the hot volumes above the core; one part is included in the
diaphragm and the other one is outside the chassis above the diaphragm walls. The 2 volumes are
connected by a junction simulating the flow through the openings in the wall chassis above the core.
The spallation loop
At the present stage, the spallation loop is modelled by a vertical pipe with inlet and outlet
horizontal pipes. The vertical pipe represents the part of the spallation loop through the core. The
pressure, temperature and coolant flow are given at the initial state. A small power could be introduced
to represent the thermal power produced in the loop. The wall of the loop, submitted to the thermal
exchange between the spallation flow and the by-pass core volume representing the flow between the
assemblies, is modelled with a structure.
First results of the calculations in nominal conditions.
A classical calculation with a nominal power of 38.8 MW and the power distribution described in
section 3.3 was done.
After stabilisation of the flow in the channels (between 0 and 300 s.), at zero power, a power
ramp is applied between 300 and 400 s. (Figure 5)
Figure 5. Power transient
1079
Figure 6. Distribution of the flow in the hot, medium and cold channels
The total coolant flow is 1 400 kg/sec and distributed as follows:
94 kg/s in the hot channel;
653 kg/s in the medium and cold channels;
653 kg/s in the cold channel considering that the dummy assemblies are plugged (the fluid
flows in 21 assemblies only with a small power 54 assemblies are plugged).
The velocities of the coolant are identical in the tree channels.
The repartition of the pressures in the vessel and core is shown in Figure 7 below. The initial
pressure (pressure at the pump outlet) was 0.7 MPa, and the pressure drop between the middle axes of
the upper and the lower plena was 0.56 Mpa.
Figure 7. Pressure distribution in the different channels
The stabilisation at 38.8 MW with the power distribution given in Table 1, shows a maximum
temperature of the coolant at the channel outlets equal to 730, 697, 611K respectively in the hot,
medium and cold channels.
1080
Figure 8. Pressure drop between the upper and the lower plena
The average temperatures of the fluid in the core axial level are: 634, 611, 556 K respectively in
the hot, medium and cold channels. The heat transfer coefficient and the heat flux, at the outer
cladding surface of the fuel elements, at the mid core level, are shown on Figures 9 and 10 for each
group of assemblies. The heat transfer coefficient is about 3.25e+04 W/mK.
Figure 9. Fuel surface heat transfer coefficient in the hot, medium and cold channels
(mid core level)
In the hot channel, the maximum fuel temperatures are 1 500K and 1 035K at the centre and at
the pellet surface respectively. The cladding temperature at the inner and outer surfaces, are 691 and
658 K (mid core level). The maximum temperature of the core vessel chassis walls is 623K in the
upper level.
Figures 11,12,13 and 14 show the temperature profiles at various levels and in the different
channels.
1081
Figure 10. Fuel surface heat flux in the hot, medium and cold channels (mid core level)
Figure 11. Temperature profile in the hot channel fuel (mid core level)
Fuel temperature profile
in the hot channel
K
1600
1400
HCH-inlet
1200
HCH-under axe
HCH-medium axe
HCH-upper axe
1000 HCH-outlet
temperature (K)
800
600
cladding
Fuel
outer surface
outer surface
400
vertical axe
200
0 nodal points
0 1 2 3 4 5 6 7 8
1082
Figure 12. Temperature profile in the medium channel fuel
Fuel temperature profile
in the medium channel
K
1600
1400
MCH-inlet
1200 MCH-under axe
MCH-medium axe
MCH-upper axe
1000
MCH-outlet
temperature (K)
800
600
vertical axe
cladding
outer surface
outer surface
Fuel
400
200
0
nodal points
0 1 2 3 4 5 6 7 8
Figure 13. Temperature profile in the cold channel fuel
Fuel temperature profile
in the cold channel
K
1600
1400
CCH-inlet
1200 CCH-under axe
CCH-medium axe
CCH-upper axe
temperature (K)
1000 CCH-outlet
800
600
cladding
Fuel
outer surface
outer surface
vertical axe
400
200
0
nodal points
0 1 2 3 4 5 6 7 8
1083
Figure 14. Temperature profile in the dummy fuel of the cold channel
Temperature profile
in the dummy fuel of the cold channel
K
1600
1400
DMCH-inlet
DMCH-under axe
1200
DMCH-medium
axe
1000
temperature (K)
DMCH-upper axe
800
600
cladding
Fuel
outer surface
outer surface
400
vertical axe
200
nodal points
0
0 1 2 3 4 5 6 7 8
Figure 15. Pb-Bi temperature at various position in the spallation loop
In these first calculations, no power was injected in the spallation loop. The temperature is only
function of the thermal exchange between the core and the loop.
At the outlet of the spallation pipe the temperature reaches 700K (see Figure 15).
1084
Conclusions
This first calculations in nominal conditions are in agreement with the preliminary calculation of
the pre-design (Ref 1 and 2). The pressure drops were calculated approximately, especially in the
sections of the grid and support plate. Only the restrictions and expansions of the cross -sections were
taken into account. These pressure drops have to be calculated more accurately.
The by-pass flows through narrow paths in the diaphragm were not taken into account neither.
The free convection was not included, and some sensitive pre-calculations must to be done. The
medium plenum is considered as a stagnant volume.
A lot of parameters could be tested and evaluated. The pumps, the heat exchangers and the
control system must be included in the model to perform safety studies.
Acknowledgement
We are grateful to Ansaldo Nucleare, in particular Doctor Alessandro Alemberti, for their
contributions and for the version Pb-Bi of RELAP/mod 3.2.2 to be tested by the MYRRHA team.
REFERENCES
[1] Ph. Benot (1999/2001), MYRRHA Project Pre-Design-Phase General Desription of the Primary
Systems and Associated Equipment, SCKCEN.
[2] H. At Abderrahim, P. Kupschus (2002) A Multipurpose Accelerator-driven System (ADS) for
Research & Development, Pre-Design Report R-3595 SCKCEN.
[3] V. Sobolev (2001), Reference Fuel Pin and Assembly, 3 RF&M SCKCEN.
[4] G. van den Eynde, Ch. de Raedt, E.M. Malambus, Th. Aoust, N. Messaoudi (2002), Status of
Neutronics Analysis of the MYRRHA ADS, SCKCEN Internal Report.
1085
