J. GUIDEZ et al.

OPTIMIZATION OF THE SECONDARY LOOPS ON THE ESFR SMART

J. GUIDEZ
CEA
CEN SACLAY 91190 Gif-sur-Yvette France
Email: [email protected]

E. GIRARDI
EDF Lab Paris-Saclay
7, bd Gaspard Monge, 91120, Palaiseau, France

J. BODI
Paul Scherrer Institut
5232 Villigen PSI, Switzerland

K. MIKITYUK
Paul Scherrer Institut
5232 Villigen PSI, Switzerland

Abstract

Based on feedback from existing reactors and current projects, the ESFR SMART project proposes an optimization of
the secondary circuit with the aim of improving safety. This optimization also leads to a simplification of the circuits and an
optimization of the costs. Some points require R&D to validate the options offered.

1. INTRODUCTION

The ESFR SMART project has proposed for the primary circuit a number of safety options based on the SFRs
experience feedback [1-3]. These options are based on simplifications as well as passive and forgiving systems.

In the same spirit, this paper proposes an optimization of the secondary circuits to improve their intrinsic safety
and their compactness.

2. OPTIMIZATION OF SAFETY AT THE LEVEL OF THE SECONDARY LOOPS

In terms of safety, three points are mainly to be optimized in these circuits

First, these circuits have the role of evacuating the power of the reactor, and in case of shutdown will actively
participate in the evacuation of the residual power. The use of these circuits has been favoured because it is the
loop commonly used by the operator for this purpose, and in all operating circumstances. They were therefore
pre-dimensioned [4, 5] to be able to evacuate this power alone even after the loss of the water circuits, only by
natural convection of the air around the modules of steam generators. On the other hand, this pre-sizing showed
[4,5] that natural convection in the circuits was not sufficient and that an operation of the secondary pumps at
reduced speed (100 rpm) was necessary, at least at the start of the event, to assure 100% function. Note also the
presence in each secondary loop of a DHRS1 system connected to the exchanger and capable of performing this
function on its own, in passive natural convection and even in the event of draining of the secondary circuit.

Second, the feedback experience of the past SFRs shows that sodium leaks mainly take place at the level of the
secondary circuits [6-8] . Note also that for ESFR SMART specific measures have been taken to avoid any
leakage of primary sodium. These possible leaks of non-active secondary sodium are more of a security concern
than a safety concern. That said, we will see in the chapter on secondary piping, the proposals made to both
minimize the risk of rupture and increase the possibilities of rapid detection and mitigation

1
 FR21: IAEA-CN-291/201

Finally, sodium / water reactions are a problem to be tackled at the level of steam generators. We will see in this
chapter that the choice of type of steam generator has been made accordingly. It will also be seen that
corresponding provisions are to be taken at the level of the casings containing the modules.

All these points have been taken into account essentially based on the existing feedback experience on the SFR
secondary circuits [6] , but also on published results of studies on last projects as ASTRID in France, BN 1200 in
Russia or PGSFR in Korea.

3. GENERAL DESCRIPTION OF THE ESF SMART SECONDARY CIRCUIT

The secondary system transfers the heat from the intermediate heat exchangers to the steam generator units
during power operation and during operational decay heat removal via the tertiary water/steam system after
reactor shutdown. It represents a non-radioactive barrier between the radioactive primary sodium system and the
non-radioactive tertiary water/steam system.

Each of the six loops has one intermediate heat exchanger (IHX), one secondary pump and one steam generator
unit with six modules. Each of the loops has also one Decay Heat Removal System (DHRS-1) able to keep heat
removal from IHX even if the main secondary loop is drained. A quick draining system allows draining the
sodium of the loop to the storage tank.

The sodium flowrate in each loop is 10800 m3/h (2541 kg/s) and the power of each steam generator is 100
MWth.

1 – Intermediate heat exchanger 5– Decay Heat Removal System (DHRS-2)

FIG. 1. General view of the ESFR-SMART secondary loop.

Note that there is a permanent flow of sodium in the loop for purification and monitoring. The sodium leaks at
the secondary pump (hydrostatic bearing and impeller labyrinths / either a few percentage of the flow/ are
collected passively by gravity and sent to the storage tank. There a lift pump returns this flow through a
purification circuit to the loop. Note that permanent circulation of sodium is passively ensured in DHRS 1, which
therefore does not require any sodium monitoring or purification circuit.

4. THE MAIN COMPONENTS

The main components are classical and issued from initial EFR propositions

Heat exchangers: Compared to the exchanger dimensioned during the EFR project, there will be a slight increase
in diameter linked to the existence in connection with DHRS 1 of a sampling tube in the hot collector and a
return tube of the cooled sodium in the central cold sodium arrival tube.

Secondary pump: The REX of the secondary pumps shows a high reliability of these components and few
problems in operation (see [6], chapter 10 and [7], and chapter 10). Tests of high power electromagnetic pumps
were carried out in the frame of the ASTRID project in France in 2017/2018 to try to appreciate the advantages /
disadvantages of these systems in regards of usual mechanical pumps. As of today, there are not yet clear
advantages demonstrated in cost, sizing or reliability of these new systems. At ESFR SMART level, we therefore
kept the mechanical pumps sized for EFR.

Steam generator: The straight-tube module option, which was the EFR initial option, was kept. This type of
modules is used in the Indian PFBR and Russian BN-800 reactors, and selected in several projects as PGSFR
and Chinese fast reactor project. This modular option was preferred to the concept of helical steam generator
selected and used for Superphenix (see [7], chapter 12 ) due to several advantages in terms of industrial
manufacturing, of maintenance (possibility of periodically replace a module) and in terms of safety files (better
detection and mitigation of accidents with sodium/water reactions). Moreover, having six modules on a loop
makes it possible to increase the available heat exchange surface and to be capable to create a possibility of
removing the decay heat by simple natural convection of the atmospheric air around the walls of these modules

4.1. DHRS 1 system

To ensure the highest safety in case of unavailability of the main heat removal route through the secondary
sodium loop ,a special decay heat removal system, DHRS-1, is implemented at each of the six IHXs (see
FIG. 5).

The DHRS-1 loop operates in parallel to the secondary loop using the hot secondary sodium extracted from the
IHX as the working medium. The heat is rejected to the environment using sodium/air heat exchanger located at
the bottom of the air stack the largest part of which is located outside of the reactor building. The cold secondary
sodium comes back to the IHX cold sodium entry. Such a scheme promotes cooling of the primary sodium in the
IHX and therefore enhances the primary sodium natural convection through the core and IHX.

This operation is very passive. The operator has only to open the window of air circuit and passive circulation of
sodium (enhanced by the thermal pumps) will begin in the DHRS-1 circuit.

3
 FR21: IAEA-CN-291/201

FIG. 2. View of the DHRS 1 system attached to his heat exchanger.

To increase passivity a thermal pump can be installed on the hot line of 200mm diameter. . A thermal pump is a
passive electromagnetic pump, which uses thermoelectricity generated by the difference in temperatures (hot
sodium and atmospheric air). This pump does not need external electricity supply and provides a supplementary
flow rate. The concept of the thermal pump is illustrated in figure 3: a magnetic field is created by permanent
magnets. An electric current is created by thermos-elements. The magnetic field and the electric current initiate a
pressure drop and a flowrate

FIG. 3. View of the thermal pump principle.

This component needs some R&D, it means test in sodium. It could be noticed that, if results were not available,
a little electromagnetic pump with secured power supply could replace this component.

5. PIPING

To limit sodium velocity by about 5 m/s in normal operation the following diameters of the secondary pipes
proposed are Ø 850 mm for the main lines and Ø 350 mm for the SG Main lines.

Two options are possible for these pipes

The first option is to have pipes relatively long and flexible, to be able to accommodate thermal expansion. It has
been the option of almost all existing plants and particularly for Phenix and Superphenix. In this case, the main
pipes have a wall thickness of 12.5 mm and a length of about 220 m, including elbows to accommodate thermal
expansion and discharge lines for sodium draining at hot leg and at cold leg. For the steam generators lines, we
have wall thickness of 15 mm; length of about 30 m, including elbows to accommodate thermal expansion and
discharge lines for sodium draining at lower part of each steam generator. The lengths of the lines depend on the
chosen material and its thermal expansion coefficient. The resulting volume of the secondary sodium in each of
the six loops is about 250 m3, including 90 m3 in steam generators, 27 m3 in heat exchanger, 7 m3 in the
secondary pump and about 116 m3 inside the piping.

The second option is to have straight and rigid piping where thermal expansion is taken up by bellows. Indeed
the feedback from these flexible pipes shows some problems. On Superphenix the sodium pipes had lyres to
resume the dilations between cold state and operation. However, the heavy weight of these pipes requires
supports. It is therefore necessary that the pipes can slide in these supports. However, anti-seismic standards
require pipes firmly maintained during an earthquake. This led to the definition of rather complex systems on
Superphenix that did not work well. After each transient, the pipe was found in abnormal positions (see below
part of the chapter of the SPX book in reference 7)

"The long length of the relatively flexible pipes led to numerous support devices allowing their expansion while
blocking them in the event of an earthquake (the DAB self-locking devices).Many of these devices (2400) and
the non-linearity of their behaviour made their monitoring and maintenance very cumbersome. Indeed, DAB in
bad state could induce blockages of pipes, which would cause significant mechanical stresses. "

Another thing that worked poorly was the thermal insulation on these flexible pipes. This led to difficulties in
detecting leaks, risks of corrosion by undetected leaks, and numerous false alarms that were very difficult to
verify [6,7].

Based on this negative feedback in terms of investment ,safety and operation , ESFR SMART propose to define
fixed and non-sliding supports, which plays their support role in normal operation and in case of earthquake.
Between these fixed points, the pipes are straight. A bellow is installed in the middle of this right part, which
support the dilatation effect. A choice of material other than 316L, as for example 9 Cr, would also significantly
reduce this dilatation. The pipes length reduction allows also a significant reduction of dilatations. It should be
noted that the Russians on the BN 1200 project chose this bellows option for the future.

The benefits are as follows:

— Decrease of the costs because decrease of the lengths of pipes, quantities of secondary sodium, volumes
of the tanks of storage, etc.
— Simpler, cheaper and more efficient supports / Safety Gain/ Easier to manage in exploitation,
— Ability to use a removable insulation easy to install on these straight parts and without direct contact /
improved detection possibilities / less false alarms / improved safety,
— Improvement of the circuit's natural convection circulation with shorter lines,
— Reduce the distance between fixed points and so dimensions and cost of secondary building,
— In addition, the mechanical dimensioning of the pipes is simpler because the flexibility option required
small thickness of the pipes and numerous welds for the elbows and expansion lyres. With straight piping
we can minimize the number of welds and take the desired thickness for pipes sizing.
In practical terms, on the drawings, we will draw straight pipes to join the fixed points that are the components:
exchangers, SG, pumps, DHRS 1... In this case, we can propose to reduce all these lengths with a circular
secondary building around the primary vessel. This disposition allows also to have the same chimney for the
casing and the DHRS1 We arrive at the following disposition of the secondary loop in figure 4.

5
 FR21: IAEA-CN-291/201

FIG. 4. View of the compact secondary loop with the same chimney for casing and DHRS1.
The original flexible loop had a length of 195m of 850mm diameter tubing (or 220m if I count the piping
towards the sodium-draining tank) and around 150m of 350mm diameter piping. In contrast, with the new design
we have 67m of 850mm diameter tubing (or around 88.5m if I count the piping towards the sodium draining
tank) and around 12m of 350mm diameter tubing. Based on the above lengths, the available sodium volume in
the piping only (so not taking into account the steam generators, IHX and secondary pump) for one secondary
circuit for the original case is around 116m3, whereas for the simplified circuit this number is around 37m3. In
this sense, the reduction in sodium volume is 79m3.

6. GENERAL LAY OUT OF THE PLANT

There is also a big benefit on the final sizing of the secondary building including all pipes and components with
the possibility to compact by this circular disposition

FIG. 5. Top view of the Circular disposition of the secondary loops round the primary vessel.

This new circular disposition allows a big improvement of the general lay out as shown on figure 6
 J. GUIDEZ et al.

FIG. 6 Comparison of the initial lay out (left picture) and the new circular disposition with common chimneys for DHRS1
and steam generators casings (right piscture).

7. SAFETY ANALYSIS

7.2. Sodium leaks and fires

The option of straight lines without elbows allows the use of protection against leakage by a double wall piping .
This installation would be difficult almost impossible, on large flexible pipes with large movements. The
external wall is internally covered with insulation and it exists a gap between the sodium pipe and this insulation.
This external wall can be easily open to allow easy interventions, for example in case of alarm

Classical sodium fire detection are installed on the sodium pipe to detect any leak. These detections are
particularly installed around the bellows and in the lower part. Therefore, a sodium leak detection is possible
before any chemical interaction of the sodium with the insulation. Complementary detections can be added
between the pipe and the removable insulation, as sodium smoke detection in the partitioned interior zone. This
set of provisions allows quick detection and good containment of any sodium leak inside this double wall.

7.3. Sodium/water reaction

Conventional devices enable to efficiently control the risk of water-sodium reaction by detection of hydrogen in
the sodium at the outlet of each module. The modularity makes it easier to quickly detect a sodium-water
reaction, isolate and drain the failed module. It allows also minimizing the theoretical envelope accidents. Even
with a hypothetical rupture of all the tubes in a module, the accident can be managed in terms of overpressure
and mitigation. In case of water/sodium reaction, the consequences for the plant operations are limited and the
operation can continue with remaining modules, after isolation of the defective module.. Mitigation measures
against risk of sodium-water-air reaction have also to be taken into account in the building concept. In particular,
“water area” and “sodium area” in the secondary system buildings should be strictly separated to avoid any
interaction. The casing is sized to resist any external aggression

7.4. Decay heat removal

In case of unavailability of the secondary loop and even if this loop is drained, the DGRS 1 can provide 100% of
the function, very passively in natural convection of air and sodium [4, 5].

7
 FR21: IAEA-CN-291/201

8. R&D NEEDS

R&D on bellows: Some R&D is necessary for these bellows of diameter 850 mm in terms of dilatation capacity
and time life. However, the use of bellows in sodium is not very new. These bellows exist on many sodium
valves especially in Phenix and Superphenix, and inside the Phenix heat exchangers .A bellow of large diameter
(approximately 800 mm) was installed in Superphenix (figure 8) on the internal part of the hot collector of the
intermediate exchangers to take up the differential expansions with the external part. This device had several
expansion waves and a thickness of 8 mm. It has undergone a cycling test with a large number of cycles for
validation.

FIG. 7. View of the bellow tested on Superphenix heat exchanger.

On the ESFR SMART steam generator module himself, a bellow of large diameter (750 mm) exist to allow
relative dilatation between the external wall of the component, and his internal bundle. The bellow conception
will be helped with the drastic reduction of dilatations due to short lines with adapted materials.

R&D on thermal pumps: Some calculations were provided on the thermal pump for DHRS1. However, a test
scale 1 in sodium would be necessary for final validation and industrial demonstration

9. CONCLUSION

Based on the REX of existing reactors, the project ESFR SMART propose an optimization of the secondary loop
allowing to increase safety in terms of sodium leak detection and mitigation, in terms of possibilities of passive
decay heat removal, and in terms of detection and mitigation of water/sodium reaction.

This proposed design allows also to reduce distances between components and so to reduce drastically sizing
and cost of the loops but also of the secondary building with a circular option. The length of the pipes, for each
loop is reduced from 195 to 67 m and the sodium volume in the piping from 116 to 37 m3. The surface of
secondary building is reduced more than 50%.
 J. GUIDEZ et al.

FIG. 8. General view of the secondary circuits around the primary vessel.

The points that will need R&D on this loop concept are mainly the fabrication and test of bellows and thermal
pump.

ACKNOWLEDGEMENTS

The work has been prepared within EU Project ESFR-SMART, which has received funding from the
EURATOM Research and Training Programme 2014-2018 under the Grant Agreement No. 754501.

9
 FR21: IAEA-CN-291/201

REFERENCES

[1] J. Guidez, A. Rineiski, G. Prêle, E. Girardi, J. Bodi, K. Mikityuk, “Proposal of new safety measures for
European Sodium Fast Reactor to be evaluated in framework of Horizon-2020 ESFR-SMART project”, 2018
International Congress on Advances in Nuclear Power Plants (ICAPP 18), Charlotte, NC The Westin Charlotte,
April 8-11, 2018.
[2] J. Guidez, A. Rineiski, E. Girardi, K. Mikityuk , J. Bodi, “Status of new safety measures considered for
European Sodium Fast Reactor in the ESFR-SMART project”, 2019 International Congress on Advances in
Nuclear Power Plants (ICAPP 19), Juan-les-Pins, France, May 12-15, 2019.
[3] J. Guidez, J. Bodi, K. Mikityuk, A. Rineiski, E. Girardi, “New safety measures proposed for European Sodium
Fast Reactor in Horizon-2020 ESFR-SMART project”, GIF Symposium UIC, Paris, France, October 16-17,
2018.
[4] J Guidez, A Gershenfeld, K Mikitiuk, J Bodi, E Girardi, J Bittan, C Bore, A Grah “Innovative Decay Heat
Removal Systems in European Sodium Fast Reactor” ”, 2020 International Congress on Advances in Nuclear
Power Plants (ICAPP 2020)
[5] J. Bittan, C. Boré, J. Guidez, “Preliminary assessment of decay heat removal systems in the ESFR-SMART
design: the role of natural air convection around steam generator outer shells in accidental conditions”,
International Youth Nuclear Congress (IYNC), March 8-12, 2020, Sydney, Australia.
[6] J. Guidez, “Phenix experience feedback”, Edition EDP Sciences, EAN13: 9791092041057 (2014).
[7] J. Guidez and G. Prêle, “Technical and scientific Superphenix. achievements”, Edition Springer, ISBN 978-94-
6239-245-8 (2017).
[8] J. Guidez ,L Martin , S.C.Chetal, P. Chillapandi; Baldev Raj “ LESSONS LEARNED FROM SODIUM COOLED FAST
REACTOR OPERATION AND THEIR RAMIFICATIONS FOR FUTURE REACTORS WITH RESPECT TO
ENHANCED SAFETY AND RELIABILITY”, Nuclear Technology, Vol 164 Nov 2008