Nuclear Engineering and Design 305 (2016) 179190

Contents lists available at ScienceDirect

Nuclear Engineering and Design

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

A preliminary stability analysis of MYRRHA Primary Heat Exchanger

two-phase tube bundle
Diego Castelliti a,b, Guglielmo Lomonaco b,c,
a
Studiecentrum voor kernenergie Centre dtude de lnergie nuclaire (SCK
CEN), Boeretang 200, Mol, Belgium
b
GeNERG DIME/TEC, University of Genova, Via allOpera Pia 15/a, 16145 Genova, Italy
c
INFN, Via Dodecaneso 33, 16146 Genova, Italy

h i g h l i g h t s

 MYRRHA is a pool-type LBE-cooled ADS, operable also as a critical reactor.

 MYRRHA is a high priority infrastructure for nuclear research in Europe.
 PHX (primary side: LBE, secondary side: two-phase water), is a MYRRHA key component.
 The original MYRRHA PHX design does not offer a fully satisfying response to DWO.
 The adoption of an orifice allows extending considerably the stability of MYRRHA PHX.

a r t i c l e i n f o a b s t r a c t

Article history: The MYRRHA (Multi-purpose hYbrid Research Reactor for High-tech Applications) project, started at
Received 17 February 2016 SCK
CEN since 1999, aims at the construction of a pool-type sub-critical Accelerator-Driven System
Received in revised form 6 May 2016 (ADS) which could also operate as a critical reactor. The primary system, enclosed in the primary vessel,
Accepted 17 May 2016
is filled with Lead Bismuth Eutectic (LBE) which acts as primary coolant. The power is then delivered
Available online 3 June 2016
through four heat exchangers to four secondary loops. The secondary cooling fluid is two-phase water
operating at relatively low pressure (16 bar). Four aero-condensers act as heat sinks, since MYRRHA
JEL classification:
design does not foresee any electricity generation.
K. Thermal Hydraulics
The MYRRHA Primary Heat eXchangers (PHXs) cover a role of fundamental importance in normal oper-
ation and accidental conditions, being part of the primary and secondary cooling system and of the Decay
Heat Removal (DHR) system. It is thus highly relevant to understand the PHXs behavior under all the
potential working conditions.
In particular, the stability of the PHXs must be guaranteed under all operating conditions.
System code models play an important role in understanding and predicting the behavior of the reactor
in all conditions, from steady state to operational and accidental transients, and simulating all the pos-
tulated scenarios.
A solid PHX design requires a complete assessment of two-phase flow instabilities in the secondary
system water tube bundle and the potential implementation of a suitable stabilizing device (orifice) to
reduce the impact of the perturbations along the channel. The stability assessment should take in consid-
eration all the possible reactor operational power levels in order to prove the stable behavior under all
operational conditions.
The tube bundle stability assessment has been carried out by following a similar procedure used for
BWR fuel channels, through a specific RELAP5-3D model representing the PHX and able to evaluate
the propagation of a density wave in the tube length.

Abbreviations: ADS, Accelerator-Driven System; DHR, Decay Heat Removal; DWO, Density Wave Oscillations; EoL, End of Life; ETPP, European Technology Pilot Plant;
HLM, Heavy Liquid Metals; HTC, Heat Transfer Coefficient; HX, Heat eXchanger; IPS, In-Pile Section; IVST, In Vessel Storage Tank; LBE, LeadBismuth Eutectic; LFR, Lead Fast
Reactor; MYRRHA, Multi-purpose hYbrid Research Reactor for High-tech Applications; PHX, Primary Heat eXchanger; PP, Primary Pump; RVACS, Reactor Vessel Auxiliary
Cooling System; SCS, Shutdown Cooling System.
Corresponding author at: GeNERG DIME/TEC, University of Genova, Via allOpera Pia 15/a, 16145 Genova, Italy.
E-mail address: guglielmo.lomonaco@unige.it (G. Lomonaco).

http://dx.doi.org/10.1016/j.nucengdes.2016.05.019
0029-5493/ 2016 Elsevier B.V. All rights reserved.
180 D. Castelliti, G. Lomonaco / Nuclear Engineering and Design 305 (2016) 179190

A series of suitable boundary conditions, on both primary and secondary side, and perturbation triggers
have been foreseen into the model, so to discover all kind of unstable behavior and to dimension the
needed orifice to guarantee the flow stability in all operating conditions.
The PHX stability analysis is initially performed on the original tube bundle without the adoption of
any stabilizing devices, in order to check the natural behavior of the system. The possible adoption
and design of an orifice is then conducted on the basis of this preliminary study. The system response
against the various types of instabilities, before the introduction of an orifice, is not completely satisfac-
tory: a stable flow is found within certain specific system parameters ranges. After the introduction of a
suitable orifice, the system behavior becomes stable under all operating conditions against all types of
two-phase flow instabilities.
2016 Elsevier B.V. All rights reserved.

1. Introduction typical liquid metals used in nuclear applications (mainly Na and

MYRRHA (Multi-purpose hYbrid Research Reactor for High-tech
Applications) is a pool-type Accelerator Driven System (ADS) with
the ability to operate also as a critical reactor.
MYRRHA main targets can be summarized as:
 Flexible fast-spectrum irradiation facility (Abderrahim, 2011).
 Minor Actinides (MAs) transmutation demonstrator (Cerullo
and Lomonaco, 2012).
 ADS demonstrator (Lomonaco et al., 2014).
 GEN-IV European Technology Pilot Plant (ETPP) in the roadmap
for Lead Fast Reactor (LFR) (SNETP, 2013).
The MYRRHA project has been recognized as a high priority
infrastructure for nuclear research in Europe. Several European
FP6 and FP7 projects had, as main target, to finalize a preliminary
design of the MYRRHA reactor:
 FP6 IP-EUROTRANS (De Bruyn et al., 2010), leading to the final-
ization of MYRRHA/XT-ADS version of MYRRHA in June 2008.
 FP7 Central Design Team (CDT) (Sarotto et al., 2013), defining
the MYRRHA/FASTEF version in March 2012.
 FP7 MAXSIMA (Castelliti et al., 2015) (started in November
2012, ongoing), more focused on the MYRRHA safety analyses
and component qualification.
The outcome of these European FP projects has been partly used
to define the latest version of the MYRRHA design, which has been
finalized in June 2014 (De Bruyn et al., 2016) and is currently in the
verification phase. Though representing the current status, such
version is not definitive: the MYRRHA design is still evolving taking
into account results from the parallel R&D program.
SCK
CEN has actively participated in these FP6 and FP7 projects
focusing on the safety analysis through use of system codes by per-
forming code-to-code comparison of steady-state and transient
calculations on the MYRRHA reactor operating in sub-critical and
critical mode.
2. MYRRHA plant general description
A general overview of the MYRRHA primary system and its
main components is provided in Fig. 1:1
MYRRHA plant primary system is cooled by liquid LeadBis-
muth Eutectic (LBE, 45% Pb, 55% Bi) (Nuclear Science Committee,
2015). This choice shows several advantages with respect to other
1 All three systems are designed to operate in forced circulation
Some of the MYRRHA plant components are still under development; more
generally, many studies on the development of components for HLM nuclear systems
are still ongoing (e.g. Mangialardo et al., 2014; Ferrini et al., 2016).
pure Pb):
 More operation flexibility (and limited problems towards pri-
mary coolant freezing) thanks to the low melting temperature
of the eutectic (125 C), which allows to operate a fast-
spectrum irradiation facility with a relatively high core temper-
ature difference without incurring in corrosion problems.
 Low chemical interaction with water and air excluding the pos-
sibility for fire or explosions.
A drawback connected with use of LBE as primary coolant is the
accumulation of radioactive isotopes (mainly Po210), which could
pose difficulties during primary system maintenance or in case of
accidental conditions in terms of radiological releases.
Further details on LBE properties are available in Nuclear
Science Committee, (2015) and Tucek et al. (2006).
The primary system is completely enclosed in the primary
vessel (pool-type system). The primary LBE coolant flows from
the lower plenum into the core (T  270 C) to remove the core
power (100 MW in critical mode) and, from there, into the
upper plenum where it mixes with the cold by-pass flow. The
average upper plenum temperature is 325 C. Four Primary Heat
eXchanger (PHXs) units receive the LBE from the upper plenum,
which then flows into two Primary Pumps (PPs) (each PP serving
two PHXs). From the PPs the LBE is reinserted into the lower
plenum.
The cold lower plenum is separated from the hot upper plenum
by the Diaphragm, an inner vessel structure supporting the core
barrel and the penetrations for the PHXs and the PPs. Above the
LBE free surface level an inert gas layer (nitrogen) separates the
primary coolant from the reactor cover.
The primary system is linked to four independent secondary
systems through the four PHX units. Each secondary system is
operated in a forced-flow regime with a two-phase water mixture
at 16 bar (200 C): the water enters the PHX in almost saturated
conditions and exits with a quality 0.3. The moisture is then
separated in a steam drum, from where the steam is directed
towards an air condenser (one per secondary loop) and the water
is recirculated to the PHX. In normal operation the secondary
water temperature is kept constant by the control system, letting
the primary LBE conditions to change in function of the core
loading.
The steam dissipates the heat to the external environment
through the tertiary system air condenser and is then recirculated
into the steam drum. Each tertiary system contains an air fan oper-
ated in forced circulation and logically connected to the steam
drum pressure for power removal balance (Fig. 2).
(active mode) during normal operation. Nevertheless, the plant
must also be able to remove the decay heat in accidental
 D. Castelliti, G. Lomonaco / Nuclear Engineering and Design 305 (2016) 179190 181

1. Reactor vessel
2. Reactor cover
3. Diaphragm
4. Primary heat exchanger
5. Pump
6. In-vessel fuel handling Machine
7. Core barrel
8. Above core structure
9. Core plug
10. Spallation window

Fig. 1. Overview of the MYRRHA reactor (in ADS mode) (De Bruyn et al., 2016).

Fig. 2. Secondary system (single loop) schematic concept (De Bruyn et al., 2016).

conditions in passive mode. Two systems are devoted to DHR func- Tank (IVST), pump power, Po-decay heat, c-heating, spallation tar-
tion in accidental conditions: get power, etc.

 DHR system 1 function is accomplished by the secondary and

tertiary cooling systems, assumed able to operate in passive
mode (if required).
 DHR system 2 function relies on the Reactor Vessel Auxiliary
Cooling System (RVACS): it relies in liquid water flooding the
reactor cavity in order to remove the heat from the vessel exter-
nal surface and passively deliver it to a series of heat
exchangers.
Despite the maximum core power of 100 MW, the plant has
been designed for a maximum nominal power of 110 MW to take
into account all additional heat sources, such as In Vessel Storage
3. Primary Heat Exchanger general description
The PHX design (Fig. 3) chosen for MYRRHA is a counter-current
shell-and-tube concept consisting of:
 684 stainless steel (AISI 316L) tubes.
 2 tube plates.
 A double-walled central feedwater pipe connected to the sec-
ondary system recirculation line.
 A double-walled bottom head, collecting feedwater and con-
nected to the tube bundle.
182 D. Castelliti, G. Lomonaco / Nuclear Engineering and Design 305 (2016) 179190

Table 1
MYRRHA PHX main geometrical parameters.

Parameter Unit Value

Shroud internal diameter m 0.82
Feed water pipe external diameter m 0.2
Water tubes number 684
Water tubes pitch m 0.0256
Water tubes external diameter m 0.016
Water tubes internal diameter m 0.014
Thickness of water tubes m 0.001
Total length of water tubes m 10.92
Active length of water tubes m 2.1

Table 2
MYRRHA PHX main thermalhydraulic parameters.

Parameter Unit Value

PHX unit power MW 27.5
PHX LBE inlet temperature C 325
PHX LBE outlet temperature C 270
PHX LBE mass flow rate kg/s 3450
PHX water inlet temperature C 200
PHX water mass flow rate kg/s 47
PHX water pressure bar 16
PHX water outlet quality 0.3
PHX water outlet void fraction 0.9

0.3. A summary of the main geometrical and thermalhydraulic

PHX parameters is shown in Tables 1 and 2.
As shown in Fig. 3, the tube bundle is extended from the bottom
tube plate to the top tube plate, as normal for shell-and-tube HXs.
But, in MYRRHA design, the LBE inlet is placed at 2.5 m from the
bottom plate, instead of being located at the top of the component.
This configuration defines an active length for the tube bundle of
2.1 m where the LBE flow actually takes place.
The hot LBE free surface level is in contact with the tube bundle.
Above the free surface, cover gas fills the space between shroud
and feedwater pipe.2 Several other objectives can be achieved by
adopting such configuration:load conditions and in any
Fig. 3. Primary Heat Exchanger (De Bruyn et al., 2016).
 The two-phase flow is well developed inside the PHX tubes
from the inlet up to the top of the component, with no phase
 A top head, providing connection with the riser pipe of the sec-
separation to be expected within the tubes.
ondary system.
 High aspect ratio providing a better counter-current flow devel-
 An external shroud separating LBE in the hot plenum from LBE
opment through the bundle.
flowing in the PHX.
 Only one tube plate (and hence one set of weldings) is located
under LBE. The upper tube plate is positioned above the hot free
All metallic surfaces separating primary LBE from secondary
surface.
water, with the exception of tube bundle for Heat Transfer Coeffi-
cient efficiency reasons, present a double-walled structure as
There are, on the other hand, several possible disadvantages
countermeasure against LBE-water interaction in case of leakage/
coming from this design approach:
break: as a consequence, the bottom head and the feed-water pipe
are double walled, while the external shroud and the top head
 High two-phase pressure drop in the tube bundle, with poten-
maintain a single wall structure because no risk of interaction of
tial increase of dynamic instabilities and consequent need to
LBE and water is involved in case of failure.
design a suitable orifice to generate enough pressure drop in
In normal operation conditions, LBE from the hot plenum
the single phase (inlet) zone.
(325 C) enters the PHX from the inlet openings in the external
 The notable tube length could lead to important mechanical
shroud. The flow is then directed downwards, through the tube
stresses (weight and thermal induced) in the tube plates.
bundle, where the actual heat exchange takes place. Outlet open-
 The tube bundle is in contact with the free surface zone leading
ings, directing the LBE flow towards the PPs, provides the exit path
to possible problems due to differential thermal expansion and
for the cold (270 C) LBE.
level fluctuations resulting in thermal fatigue.
On the secondary side, water at a pressure of 16 bar at nearly
saturated conditions (200 C) flows down the central down-
comer pipe into the PHX bottom head and then upwards through
the tubes where it is heated by the counter-current flowing LBE,
thus producing a water steam mixture with a final quality of 2
Heat exchange through cover gas and tube bundle is neglected.
 D. Castelliti, G. Lomonaco / Nuclear Engineering and Design 305 (2016) 179190
Preliminary reactor control strategy
380
Core average LBE outlet temperature
360
PHX LBE inlet temperature
340
PHX LBE outlet temperature
320
Temperature (C)
Secondary water temperature
300
280
260
240
220
200
180
0 10 20 30
Fig. 4. MYRRHA preliminary control strategy.
3.1. MYRRHA preliminary control strategy
The PHX, as main interface component between primary and
secondary cooling system, plays a fundamental role for what con-
cerns the reactor control. Several design choices adopted for the
PHX are motivated by requirements coming from the reactor con-
trol strategy. It is then important to outline the main concepts for a
complete understanding of the PHX behavior.
The MYRRHA application catalog, based on the reactor being a
material testing and demonstration plant (Abderrahim, 2011)
and emphasizing the correct functioning of the In-Pile Sections
(IPSs), requires the primary system to be flexible with respect to
the operation power and by this to be able to vary thermalhy-
draulic conditions. The system must be able to operate at different
power levels in compatibility with performances needed. To meet
the irradiation performance needs, however, differently from
power reactors, it is possible to vary the primary system ther-
malhydraulic conditions without any major consequence on
these performances.
Hence, the reactor control, no more bound to preserve fixed
thermalhydraulic conditions in the primary vessel, can be designed
in order to achieve better operational and safety requirements.
The MYRRHA reactor control strategy adopted follows the
following logic:
 Primary mass flow rate is kept constant in any condition.
 Secondary mass flow rate is also kept constant in any condition.
 Secondary system pressure and temperature are always kept
constant, in any core power load conditions and in any PHX
condition:3
o Water pressure: 16 bar at the lowest SCS point.
o Water temperature: 200 C (saturation temperature or
slightly subcooling degree).
 Primary system temperatures may vary, according to the core
power, between 200 C (safe shutdown temperature) and the
nominal temperatures at maximum power.
 Secondary water PHX outlet quality also varying in function of
core power.
Fig. 4 shows the temperature variation in function of the core
power at every power levels.
3
PHX EoL conditions are defined basing on oxide layer formations and tube fouling.
The nominal conditions are evaluated assuming a PHX with 5 lm of oxide layer on
both inner and outer tube side and no fouled tubes.
183
40 50 60 70 80 90 100
Primary power (%)
4. Primary Heat Exchanger RELAP5-3D model
The MYRRHA plant has been modeled in its integrity (primary,
secondary, tertiary systems) through a RELAP5-3D (RELAP5-3D
Code, 2012) model (Castelliti and Hamidouche, 2016). The PHX is
a relevant section of the model, connecting primary LBE system
with the four secondary water loops.
Looking forward to the stability analysis, the PHX model has been
extracted from a general plant nodalization, making it self-standing
through the application of six boundary conditions (Fig. 5):
 Primary LBE inlet volume.
 Primary LBE inlet junction.
 Primary LBE outlet volume.
 Secondary water inlet volume.
 Secondary water inlet junction.
 Secondary water outlet volume.
The PHX tube bundle has been simulated with three pipe com-
ponents and two heat structures. One pipe, in the primary side,
represents two primary channels.4 Other two pipes, in the sec-
ondary side, represent two secondary tubes. The two heat structures,
simulating the AISI 316L tube material, connect the primary pipe
with each of the secondary pipes.
The number of nodes used in the simulation is 3005 (50 for
the primary channels, 250 for the secondary channels), with a
volume length always maintained in the range between 0.04 m
and 0.09 m. Adjacent volume length variation is always kept within
a factor 2.
The RELAP5-3D standard semi-explicit numerical scheme has
been chosen for the simulation; the time step has been kept con-
stant and well below the Courant limit for this problem. As a result,
the typical calculation time for a run is 1 h on a standard laptop
PC equipped with a 2.53 GHz processor.
LBE flow rate is kept always constant6, while the inlet and outlet
temperatures are varied according to the reactor control strategy to
simulate different power levels.
4
The LBE pipe element has been modeled according to inner elementary cell
geometry, thus neglecting edge tubes effect. This has an impact on flow area,
hydraulic diameter and, as a consequence, on mass flow rate.
5
For the sensitivity analysis on the number of nodes (see below), this value has
been reduced.
6
The mass flow rate has been evaluated as twice the mass flow rate of a single PHX
unit divided by the total number of secondary tubes.
184 D. Castelliti, G. Lomonaco / Nuclear Engineering and Design 305 (2016) 179190

Being MYRRHA PHX featuring two-phase water in the

tubes, it becomes important to well understand the potential
unstable behavior of the component, properly mapping the
flow regimes and conditions where the instabilities are more
likely to incur.
It is important to note that the PHX, differently from a BWR
core, is not a system with an imposed heat flux but with imposed
wall temperature. The way the boundary conditions are imposed
(see RELAP5-3D model description) to the system provides a
proper simulation of this feature.
Generally speaking, a number of different two-phase instabili-
ties can be found in a boiling channel (Castelliti and
Hamidouche, 2016; Boure et al., 1973):

 Ledinegg instability The flow excursion or Ledinegg insta-

Fig. 5. PHX simplified RELAP5-3D model scheme.
The secondary water inlet conditions, in terms of mass flow rate
and temperature, are varied as sensitivity parameters.
A series of features have been introduced in the RELAP5-3D PHX
model to achieve a higher accuracy in the instability region
definition:
 The physical properties adopted for the heat structures, thermal
conductivity and volumetric heat capacity in the stainless steel
tubes, have been modified on purpose to try to eliminate any
transient effect due to the thermal inertia of such solid struc-
tures. In particular, the thermal conductivity assumed a very
high value, while the volumetric heat capacity has been low-
ered to negligible importance.
 The inlet and outlet water collector volumes at the bottom
and the top of the PHX tubes have been greatly reduced
in size in order to reduce any lag in density wave
propagation.
All the local pressure drops coefficients have been derived from
Idelchik (2008) and applied in the RELAP5-3D model.
A series of Control Variables (RELAP5-3D Code, 2012) has been
programmed in order to evaluate the parameters of relevant
interest for the problem and to allow an accurate definition of
the stability curve.
5. Primary Heat Exchanger stability analysis
The two-phase flow instability problem has been extensively
studied for several cases and applications, one of them being the
BWR boiling channels (Todreas and Kazimi, 2012).
bility belong to static type instability involves a sudden
change in the flow rate to a lower value (flow undergoes
sudden, large amplitude excursion to a new, stable operating
condition). It can occur when the slope of the channel
demand pressure-drop flow-rate curve becomes in some
portion negative. This behavior requires that the channel
characteristics exhibit a region where the pressure drop
decreases with increasing flow. Ledinegg instability repre-
sents the limiting condition for a large bundle of parallel
tubes between common headers (like in the PHX), since
any individual tube sees an essentially constant pressure
drop. Stable operation beyond the minimum and up to the
critical heat flux can be achieved by throttling individual
channels at their inlet.
 Flow pattern transition instabilities Flow pattern transition
instabilities belong to static type instability. These instabilities
have been postulated as occurring when the flow conditions
are closed to the point of transition between bubbly flow
and annular flow. A temporary increase in bubble population
in bubbly-slug flow (arising from a temporary reduction in
flow rate) may change the flow pattern to annular flow with
its characteristically lower pressure drop. Thus the excess driv-
ing pressure drop will speed up the flow rate momentarily. As
the flow rate increases, however, the vapor generated may
become insufficient to maintain the annular flow, and the flow
pattern then reverts to that of bubbly slug flow. The cycle can
be repeated. This oscillatory behavior is partly due to the delay
incurred in acceleration and deceleration of the flow. In
essence, each of the hydro dynamically compatible sets of con-
ditions induce the transition towards the other one. This is
typically a relaxation mechanism, and it results in a periodic
behavior characterized by finite flow amplitudes at the
threshold.
 Density Wave Oscillations (DWO) instability Density Wave
Oscillations instability belongs to dynamic type instability.
The mechanism involves the propagation of disturbances. A
temporary reduction of inlet flow in a heated channel increases
the rate of enthalpy rise, thereby reducing the average density.
This disturbance affects the pressure drop as well as the heat
transfer behavior. For certain combinations of geometrical
arrangement, operating conditions, and boundary conditions,
the perturbations can acquire a 180 out-of-phase pressure fluc-
tuation at the exit, immediately transmitted to the inlet flow
rate and become self-sustained. For boiling systems, the oscilla-
tions are due to multiple regenerative feedbacks between the
flow rate, vapor generation rate, and pressure drop. DWO are
low frequency oscillations in which the period is approximately
equal to the time required for a fluid particle to travel through
the channel. Accordingly, the term Density Wave Oscillations
are used to describe the phenomenon.
 D. Castelliti, G. Lomonaco / Nuclear Engineering and Design 305 (2016) 179190
In addition to this, by introducing in the system a perturbation,
i.e. perturbing the flow in one channel through, for instance, a mass
flow rate disturbance or a power spike, it is possible to induce flow
oscillations that can dump or amplify in time depending if the sys-
tem is in the stable region or over the instability threshold (i.e. in
the unstable region).
The main scope of the PHX instability analysis can be divided
into two different steps:
 To carry out a complete assessment of the tube bundle behavior
in its original configuration, in order to understand the reaction
of the system against the various types of instabilities.
 To design an adequate stabilizing device to extend as much as
possible the stability range to include in it all the PHX operating
conditions.
5.1. Two-phase instability typologies found in PHX analysis
5.1.1. Ledinegg instability
The PHX behavior against Ledinegg instability has been studied
through a simplified version of the RELAP5-3D PHX model, only
featuring one water tube heated by one LBE channel.
In order to investigate Ledinegg instability, the water channel
pressure difference in function of the water mass flow rate must
be evaluated. Thus, assuming constant primary power (i.e., con-
stant LBE temperatures), the water mass flow rate has been pro-
gressively reduced.
The evaluation has been performed at power levels of 100% and
7%, respectively representing full power conditions and Decay Heat
Removal conditions.
The main findings are shown in Fig. 6.
From the evaluation, it appears that the system undergoes Ledi-
negg instabilities only at very low mass flow rates ranges (15% of
nominal value), which are not considered during normal operation.
The stability margin is assured by the low nominal exit quality
(30%) which prevents the system to operate in proximity of
unstable regimes.
In case of DHR, the instability threshold shifts to 3% of the
nominal mass flow rate value, which is well above the natural cir-
culation value (Castelliti et al., 2015).
In conclusion, Ledinegg instability does not represent an issue
for the MYRRHA PHX.
However, this type of instability will be better assessed when
the MYRRHA SCS detailed design will be available. It has to be
noticed that the adoption of a properly designed pump in the
SCS, for normal operation, could result in a stabilizing mean for this
type of instability.
5.1.2. Density wave oscillation instability
In order to study the PHX response to DWO instabilities, a series
of tests have been run by varying several parameters in the PHX
RELAP5-3D model and performing different runs.
The following three parameters variations have been included
in the DWO study:
 The primary system temperatures (providing the primary sys-
tem conditions representing the different power input).
 The subcooling temperature7.
 The secondary mass flow rate.
A number of different input decks have been run, each one with
a different combination of the three mentioned parameters. A ser-
7
With subcooling temperature the temperature difference between the satura-
tion temperature and the actual tube inlet temperature is assumed.
185
ies of Control Variables inserted in the input deck has allowed
judging the stability of the selected configuration.
In general, the following effects of different parameters on sys-
tem stability have been noted:
 Primary power increase shows a destabilizing effect.
 Water mass flow rate decrease shows a destabilizing effect.
 Inlet subcooling temperature increase has a stabilizing effect
when near saturation but a destabilizing effect at higher sub-
cooling levels.
An increase in outlet quality seems to shift the system towards
a less stable condition, while the effect of inlet subcooling temper-
ature is not always affecting the system in the same way.
These results are in agreement with the approach described in
Colombo et al. (2012) and with the general DWO stability theory.
In particular, it is possible to draw a single 2-D stability map mak-
ing use of two non-dimensional numbers including, in their formu-
lation, all the effects described above8
q
 Phase change number N pch mh  vvlv
lv l
 vvlv
hhf
 Subcooling number N sub hlv l
From the formulation of the two non-dimensional numbers, it
appears clear how two parameters combinations are enough to
understand the DWO instability: the (q/m) ratio identifies the
quality evolution in the water tube, while the (h  hf) difference
provides a measurement for the subcooling.
These two factors are weighted on the water pressure through
the density and the vaporization enthalpy: it appears that the pres-
sure has, in general, a slightly positive effect on the system stability
but the final outcome is not always obvious and the variation mag-
nitude is smaller compared to the two previously mentioned
factors.
By reshaping all the input parameters combinations (LBE tem-
perature, water subcooling temperature, water mass flow rate)
into the two non-dimensional numbers above mentioned, the fol-
lowing stability map (Fig. 7) is obtained.
A sensitivity study on the number of spatial nodes used to sim-
ulate the PHX tube bundle active length has also been performed:
three different sets of simulations (respectively with 50, 25 and 10
nodes) have been run to draw the PHX stability map.
The main conclusions can be summarized as follows:
 At nominal mass flow rate and subcooling conditions, it is pos-
sible to run the reactor at 30% of maximum power before
incurring in DWO instabilities.
 In case of increased subcooling (up to 8 C less than saturation
temperature vs. 1 C in normal operation), the maximum
achievable power could be reduced to <5% nominal power. By
furtherly increasing the subcooling, the stability range increases
(Nsub > 2).
 The three different nodalizations are nearly equivalent at low
Nsub values (tube inlet temperature close to saturation), while
some differences can be noted while increasing Nsub. The overall
qualitative trend is, however, the same.
In conclusion, the PHX does not seem to offer a satisfying
response against DWO instabilities. The adoption of corrective
measures is thus strongly recommended and actuated through
8
q = primary power; m = water mass flow rate; hlv = water vaporization enthalpy;
vlv = water specific volume phasic difference; vl = saturated liquid specific volume;
h = water inlet enthalpy; hf = saturated liquid enthalpy.
186 D. Castelliti, G. Lomonaco / Nuclear Engineering and Design 305 (2016) 179190

Ledinegg instability
2.50E+04
Full power
2.00E+04 DH power
Nominal PHX channel mass ow rate
Pressure (Pa) 1.50E+04

1.00E+04

5.00E+03

0.00E+00
0 0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08 0.09 0.1
Mass ow rate (kg/s)

Fig. 6. PHX Ledinegg instability chart.

10
PHX stability map - Sensivity on nodalizaon
9
8
Stable region
Stability curve 50 nodes
7
Stability curve 25 nodes
6 Stability curve 10 nodes
Nsub

5
4
3 Unstable region
2
1
0
0 2 4 6 8 10 12 14 16 18
Npch

Fig. 7. PHX non-dimensional stability map.

the insertion of an additional pressure drop throttling device (ori-
fice) in the single phase region.
5.1.3. Flow pattern transition instability
The PHX tube could also be affected by a series of instabilities
induced by some specific two-phase flow regimes. In particular,
in case PHX active length majority is filled with two-phase water
in slug flow or churn flow regime (that is, within certain specific
intervals of P/m9 and at certain inlet subcooling levels), the channel
flow becomes unstable. Being this instability linked to the channel
prevalent flow regime, it cannot be solved through orificing (Boure
et al., 1973). It is important to limit as much as possible the unstable
P/m ranges in all the operational conditions by shifting the channel
into annular flow.
This kind of instability could represent an issue especially dur-
ing the reactor operational transients (start-up) or, in general, dur-
ing operations at low power regimes.
To inspect this instability, the same RELAP5-3D model used for
DWO instabilities has been used. In this case, a power ramp has
been simulated through the LBE primary boundary conditions
changing linearly from 0% to 100%. Thus, all the relevant flow
regimes are encountered and it is possible to see an unstable
9
P/m = power over water flow rate ratio.
behavior of the PHX tubes. The water has been considered flowing
at full flow and entering the PHX at 200 C.
From the RELAP5-3D analyses, it appears that, in the considered
water conditions, the PHX tube active section is mainly lying in the
slug flow regime with primary power between 10% and 20%
(Fig. 8).
At lower inlet temperatures (higher subcooling), the flow pat-
tern transition instability is incorporated into the DWO instabil-
ity range extending on a range of 10100% power. Thus, it is not
possible to distinguish between the different kinds of instabilities:
the pattern transition induced instabilities are studied only when
clearly separated from other effects.
In Fig. 7 it is also shown the effect of different nodalization
refinements on flow pattern transition instabilities: it can be noted
that all models can predict the insurgency of such instabilities in
the same range, with some variability in the highest end of the
range.
5.2. Orifice dimensioning
As partially anticipated, in order to limit unstable channel
behavior and to extend the reactor operational ranges, especially
looking at DWO instabilities, it is possible to place an inlet throt-
tling device (orifice) at the inlet of the tubes in the bundle, so
 D. Castelliti, G. Lomonaco / Nuclear Engineering and Design 305 (2016) 179190 187

0.2 Flow paern transion instabilies

Flow regime-induced instabilies - 50 nodes
0.18
Flow regime-induced instabilies - 25 nodes
0.16
Flow regime-induced instabilies - 10 nodes
0.14
Mass ow rate (kg/s)

0.12

0.1

0.08

0.06

0.04

0.02

0
10 11 12 13 14 15 16 17 18 19 20
Power (%)

Fig. 8. PHX slug flow pattern transition instabilities (100% water mass flow rate).

10
PHX stability map - Orice eect
9 Stable
8 region
7
Stability curve without orice
6
Stability curve with orice
Nsub

5
4 Unstable region
3
Stable region Unstable
2
1
region
0
0 10 20 30 40 50 60 70 80
Npch

Fig. 9. PHX tube bundle stability map (with orifice).

increasing the local pressure drop in the single phase region, thus
compensating the two-phase pressure drops and increasing the
stability region in the map.
The further analysis has been conducted by relying on the more
refined model.
By assuming in the PHX tube bundle inlet junction a local pres-
sure drop factor K = 35, an orifice with the following dimensions
has been identified (Idelchik, 2008):
 Length: 80 mm.
 Diameter: 6 mm (about 40% of the tube internal diameter).
The overall effect can be noted on the DWO stability map,
where the stability line is shifted towards right, thus increasing
the operating parameters ranges.
It could be possible to further increase the local pressure drop
factor by decreasing the orifice diameter, but the influence on
the stability region extension would be limited and potential prob-
lems with local water velocity could arise.
The orifice has been represented in the RELAP5-3D model
through the adoption of correct section, pressure drop factor and
hydraulic diameter in the junction connecting the PHX lower ple-
num with the two PHX tubes.
The stability map of the orificed tube bundle can be seen in
Fig. 9.
From the new stability map it is possible to note how the stabil-
ity range has been greatly increased towards high values of Npch,
which means that high P/m ratios can be achieved without incur-
ring into DWO instabilities. In particular, the nominal operation
condition for the PHX is now stable, with stability range extending
also with increased DTsub. In general, all conditions within the
nominal P/m ratio are stable against DWO, with a limit occurring
at 110% power when Nsub  0.3.
5.3. Perturbation induced instability
The adoption of an orifice allows extending the stable region
considerably. However, resolving or mitigating the different insta-
bilities in a boiling channel configuration proves to be not enough
to guarantee the stability of the continuous PHX operation. An
additional requirement concerns the stability to induced perturba-
tions through local mass flow rate disturbances or local power
188 D. Castelliti, G. Lomonaco / Nuclear Engineering and Design 305 (2016) 179190
0.14
Damping oscillaons aer ow perturbaons
0.12
Mass ow rate (kg/s) 0.1
0.08
0.06
0.04
0.02
0
0 5
Fig. 10. Damping oscillations following a mass flow rate perturbation.
spikes. Such induced instability usually appears at power levels
found to be stable for DWO but relatively close to the instability
threshold.
An induced flow perturbation has been simulated in the
RELAP5-3D PHX model through a sinusoidal closure valve semi-
cycle with duration of 0.2 s applied to a servo-valve component
placed at the inlet of one of the two tubes. Second tube flow is
maintained unperturbed.
When the plant is operating in normal operation conditions,
including the previously dimensioned orifice, the behavior of the
system subject to the mass flow rate perturbation can be repre-
sented as follows (Fig. 10).
In this specific case (not representative of all possible configura-
tions), the mass flow rate perturbation has been imposed during a
simulation of normal operating conditions, which resulted to be
stable to DWO instabilities after the adoption of the orifice.
A series of oscillations is triggered in the two channels after the
perturbation. The induced oscillations can be considered com-
pletely extinguished after 20 s from the mass flow rate perturba-
tion. This system behavior is considered qualitatively acceptable
since no amplitude increase is noted in oscillations. However, a
stability criterion has to be chosen in order to maintain in any
operating conditions an adequate margin from the instability
threshold and this has to be established in conjunction with the
control system requirements.
Without the orifice placed at the tube bundle inlet, the system
in normal operating conditions would have been in the unstable
region for DWO instabilities, while the adoption of the orifice shifts
the system in the stable zone, also reaching a satisfying behavior
against flow perturbations.
In general, an induced perturbation is damped more quickly at
lower power, while it tends to oscillate slightly more in case of
higher subcooling.
While approaching the DWOs instability limits (see Figs. 7 and
9), a slower oscillation damping or even the insurgency of amplify-
ing oscillations following a single channel perturbation can be
noted. This phenomenon imposes a more stringent constrain to
the PHX design against instability, which can be quantified accord-
ing to the chosen stability criterion.
5.4. Stability margins
Looking forward to a robust plant design, it is important to con-
sider the stability margins available against the instabilities when
Channel 1
Channel 2
10 15 20
Time (s)
operating in normal conditions. A stability margin can be properly
evaluated once a stability criterion is defined.
As an example of the procedure to be considered and followed,
it is possible to refer to a stability criterion often used in BWR and
PHWR technology (Mansani, 1978) for the stability of the core
channels10:
X 2 =X 0 6 0:24
This criterion implies that the amplitude of the mass flow rate
oscillations should be reduced to 25% compared to the previous
cycle.
By multiple RELAP5-3D runs involving a number of variations of
model parameters, it is possible to draw the stability margins
maps. As an example of the potential applications, two stability
margin maps have been derived (Figs. 11 and 12):
 Stability margin at constant power by varying the inlet subcool-
ing temperature.
 Stability margin at constant subcooling temperature by varying
the reactor power.
As it can be seen from figures above, the assumed stability cri-
terion is only partly satisfied by the MYRRHA plant: the criterion is
fulfilled as long as the power remains below 85%. Above this
threshold, the damping ratio remains above 0.25. Approaching
the saturation temperature (reducing subcooling) is also providing
a contribution towards instability.
However, this criterion is very restrictive and, as already men-
tioned, has been derived from BWR and PHWR technology having
considered also the impact of flow instability on the reactor control
system behavior; so, its direct extension to the two-phase heat
exchangers design is not obvious (the possibility to apply a less
restrictive one should be exploited).
An upper boundary of the criterion could be to accept the
requirements to the non-amplification of oscillations generated
following a perturbation (X2/X0 < 1): in this case, the system would
be considered as stable until a maximum power of 110% (assum-
ing an inlet water temperature of 200 C). However, in these con-
ditions, the secondary system and in general the whole plant
could be very difficult to control. Hence the stability margin will
have to fall between the two limits proposed.
10
Xn identifies the amplitude of a channel oscillation after the perturbation.
 D. Castelliti, G. Lomonaco / Nuclear Engineering and Design 305 (2016) 179190 189

1.2
Stability margin vs. subcooling map
Stability criterion
110% power
1 100% power
90% power
75% power
0.8
60% power
X0/X2

0.6

0.4

0.2

0
192 193 194 195 196 197 198 199 200
Temperature (C)

Fig. 11. PHX stability margin vs. inlet subcooling temperature (constant power).

1.2
Stability margin vs. power map
Stability criterion
Inlet temperature = 200 C
1
Inlet temperature = 196 C
Inlet temperature = 192 C
0.8
X2/X0

0.6

0.4

0.2

0
60 65 70 75 80 85 90 95 100 105 110
Power (%)

Fig. 12. PHX stability margin vs. primary power (constant inlet subcooling temperature).

At present, the definition of a more suitable stability criterion
for the MYRRHA PHX is still ongoing.
6. Conclusions
One of the major MYRRHA design component is represented by
the Primary Heat eXchanger (PHX), which is characterized by
LeadBismuth Eutectic (LBE) flowing in the primary side and
two-phase water in the secondary side.
Being MYRRHA PHX featuring two-phase water in the tubes, it
becomes important to well understand the potential unstable
behavior of the component, properly mapping the flow regimes
and conditions where the instabilities are more likely to incur.
For the stability analysis calculations, the PHX model has been
extracted from the general RELAP5-3D plant model, making it
self-standing through the application of six boundary conditions.
The PHX stability against Ledinegg-type instabilities is assured
by the low nominal exit quality (30%).
For what concerns DWO, the PHX in its original design configu-
ration does not offer a fully satisfying response: in nominal mass
flow rate and subcooling conditions, it is possible to run the reactor
at 30% maximum power, while in case of increased subcooling
(4 C less than saturation temperature vs. 1 C), the maximum
achievable power could be reduced to <5% maximum power.
From the analyses of the flow regimes maps, it appears that, in
nominal mass flow rate and inlet subcooling conditions, the PHX
tube active section is mainly lying in the unstable slug flow regime
with primary power between 10% and 20%. This can cause instabil-
ities to occur during operational transients.
In order to limit unstable channel behavior and to extend the
reactor operational ranges, it is possible to place an orifice at the
inlet of the tubes in the bundle, so to increase the local pressure
drop in the single phase region, thus compensating the two-phase
pressure drops and increasing the stability region in the map. An
orifice corresponding to a pressure drop factor K = 120 was found
to be enough to guarantee stability in normal operating conditions.
The adoption of an orifice allows extending considerably the
stable region. However, an additional requirement concerns the
stability to induced perturbations through local flow disturbances.
Such induced instability usually appears at power levels found to
be stable for DWO but relatively close to the instability threshold.
An induced flow perturbation has been simulated in the
RELAP5-3D PHX model through a sinusoidal valve closure semi-
cycle with duration of 0.2 s.
190 D. Castelliti, G. Lomonaco / Nuclear Engineering and Design 305 (2016) 179190
The induced oscillations in normal operation conditions can be
considered completely extinguished after less than 10 s from the
mass flow rate perturbation. This system behavior is considered
acceptable. In general, an induced perturbation is damped more
quickly at lower power, while it tends to oscillate slightly more
in case of higher subcooling.
Looking forward to a robust plant design, it is important to con-
sider the stability margins available against the instabilities when
operating in normal conditions. A stability margin can be properly
evaluated once a stability criterion is defined.
A number of preliminary studies have been performed to assess
the PHX stability margins against a stability criterion used in BWR
and PHWR technology, which proved to be not fully fulfilled by the
MYRRHA plant operation conditions. Such a criterion could, how-
ever, prove to be overly conservative, since the definition of a more
suitable stability criterion for the MYRRHA PHX is still ongoing.
Acknowledgments
The authors would like to thank Mr. Baudoiun Arien of SCK
CEN
for the useful suggestions provided. Moreover, we want to thank
Mr. Luigi Mansani, whose expertise added considerably to this
paper.
References
Abderrahim, H.A., 2011. Multi-purpose hYbrid Research Reactor for High-tech
Applications a multipurpose fast spectrum research reactor. Int. J. Energy Res.
36, 13311337.
Boure, J.A., Bergles, A.E., Tong, L.S., 1973. Review of two-phase flow instability. Nucl.
Eng. Des. 25, 165192. http://dx.doi.org/10.1016/0029-5493(73)90043-5.
Castelliti, D., Hamidouche, T., 2016. Comparison of MYRRHA RELAP5 mod 3.3 and
RELAP5-3D Models on Steady State and PLOF transient. Nucl. Technol. 193, 36
46.
Castelliti, D., Van Den Eynde, G., Van Tichelen, K., Arien, B., 2015. Safety analysis of
the MYRRHA reactor. In: NURETH-16 conference, AugustSeptember 2015,
Chicago.
Cerullo, N., Lomonaco, G., 2012. Generation IV reactor designs, operation and fuel
cycle. In: Crossland, I. (Ed.), Nuclear Fuel Cycle Science and Engineering.
Woodhead Publishing, pp. 333395. http://dx.doi.org/10.1533/9780857096
388.3.333.
Colombo, M., Cammi, A., Papini, D., Ricotti, M.E., 2012. RELAP5/MOD3.3 study on
density wave instabilities in single channel and two parallel channels. Prog.
Nucl. Energy.
De Bruyn, D., Larmignat, S., Woaye Hune, A., Mansani, L., Rimpault, G., Artioli, C.,
2010. Accelerator Driven Systems for Transmutation: Main Design
Achievements for the XT-ADS and EFIT Systems within the FP6 IP-
EUROTRANS Integrated Project. In: ICAPP10 conference, June 2010, San Diego.
De Bruyn, D., Fernandez, R., Engelen, J., 2016. Recent Developments in the Design of
the Belgian MYRRHA ADS Facility. In: 2016 International Congress on Advances
in Nuclear Power Plants (ICAPP16), San Francisco, California, April 1620, 2016.
Proceedings published by American Nuclear Society, Paper 16140., accepted for
publication.
Ferrini, M., Borreani, W., Lomonaco, G., Magugliani, F., 2016. Design by theoretical
and CFD analyses of a multi-blade screw pump evolving liquid lead for a
Generation IV LFR. Nucl. Eng. Des. 297, 276290. http://dx.doi.org/10.1016/j.
nucengdes.2015.12.006.
Idelchik, I.E., 2008. Handbook of Hydraulic Resistances, fourth ed. Revised and
Augmented. fourth ed. Revised and Augmented. Begell.
Lomonaco, G., Frasciello, O., Osipenko, M., Ricco, G., Ripani, M., 2014. An intrinsically
safe facility for forefront research and training on nuclear technologies
Burnup and transmutation. Eur. Phys. J. vol. 129, 10. http://dx.doi.org/10.1140/
epjp/i2014-14074-6, Springer.
Mangialardo, A., Borreani, W., Lomonaco, G., Magugliani, F., 2014. Numerical
investigation on a jet pump evolving liquid lead for GEN-IV reactors. Nucl. Eng.
Des. 280, 608618. http://dx.doi.org/10.1016/j.nucengdes.2014.09.028.
Mansani, L., 1978. Flow stability studies for the CIRENE reactor. In: JUICE Meeting,
Toronto, 1415 August 1978.
Nuclear Science Committee Working Party on Scientific Issues of the Fuel Cycle
Working Group on LeadBismuth Eutectic, in: Handbook on LeadBismuth
Eutectic Alloy and Lead Properties, Materials Compatibility, Thermal
Hydraulics and Technologies, Nuclear Energy Agency, 2015, <http://www.
oecd-nea.org/science/pubs/2015/7268-lead-bismuth-2015.pdf>.
RELAP5-3D Code Manual Volume II: Users Guide and Input Requirements, INEEL-
EXT-98-00834, Rev. 4.0, Idaho national Laboratory, 2012.
Sarotto, M., Castelliti, D., Fernandez, R., Lamberts, D., Malambu, E., Stankovskiy, A.,
Jaeger, W., Ottolini, M., Martin-Fuertes, F., Sabath, L., Mansani, L., Baeten, P.,
2013. The MYRRHA-FASTEF cores design for critical and sub-critical operational
modes (EU FP7 Central Design Team project). Nucl. Eng. Des. 265, 184200.
http://dx.doi.org/10.1016/j.nucengdes.2013.08.055.
SNETP Strategic Research and Innovation Agenda, February 2013, <http://www.
snetp.eu/wp-content/uploads/2014/05/sria2013_web.pdf>.
Todreas, N., Kazimi, M., 2012. Nuclear Systems Volume 1 Second Edition. CRC
Press Taylor & Francis Group (Chapter 11).
Tucek, K., Carlsson, J., Wider, H., 2006. Comparison of sodium and lead-cooled fast
reactors regarding reactor physics aspects, severe safety and economical issues.
Nucl. Eng. Des. 236, 15891598. http://dx.doi.org/10.1016/j.nucengdes.2006.
04.019.