Final Report Summary - LEADER (Lead-cooled European Advanced Demonstration Reactor)
Executive Summary:
Introduction
The EU FP7 LEADER (Lead-cooled European Advanced Demonstration Reactor) Project had as main result the conceptual design of the demonstrator ALFRED (Advanced Lead Fast Reactor European Demonstrator). ALFRED is a 300 MWth pool system aimed at proving the viability of the European LFR technology for use in a future commercial power plant. ALFRED is designed close as possible to the reference ELFR, but proven and already available solutions are adopted to be able to construct it in a short term.
Description of the Nuclear System
The ALFRED primary system configuration is pool-type with all components removable. It presents a simple flow path of the primary coolant allowing an efficient natural circulation. The primary coolant leaving from the core moves upward through the Primary Pump (PP) to the vertical shaft and then, through the Steam Generator (SG) lead inlet holes, flows downwards on the shell out of the steam side and continuing moving downwards in the cold plenum reaches the inlet of the core. The volume between the primary coolant free levels and the reactor roof is filled by Argon. The Reactor Vessel (RV) is cylindrical with a torispherical bottom head and it is anchored to the reactor cavity from the top. A cone frustum, welded to the bottom head, has the function of bottom radial restraint of the Inner Vessel (IV). A steel liner covering the reactor pit constitutes the Safety vessel (SV). The volume between the RV and SV is like that, in case of RV leak, the primary coolant continues to cover the SG inlet maintaining the flow path. The IV has the main functions of core support and hot/cold plena separation. It is fixed to the cover by bolts and it is radially restrained at bottom. The Core Support plate is mechanically connected to the IV with pins for easy removal/replacement. The core, with a total power of 300 MWth, is constituted by 171 wrapped hexagonal Fuel Assemblies (FA), 12 Control Rods (CRs) and 4 Safety Rods (SRs), surrounded by 108 Dummy Elements. It utilizes hollow pellets of MOX fuel with maximum Plutonium enrichment of 30%. Each FA is long about 8 m and consists of 127 fuel pins fixed to the bottom of the wrapper and restrained sideways by spacer grids and has a Tungsten ballast to overcome buoyancy forces. The SG and PP are integrated into a single vertical unit. Eight SG/PP units are located in the annular space between the IV and the RV walls. SG is bayonet type with double walls based on a bayonet vertical tube with external safety tube and internal insulating layer (to assure superheated dry steam). The double walls gap is filled with helium and high thermal conductivities particles. The double walls concept prevents water/lead interaction in case of break of one tube wall and, moreover, a tube wall break event can be easily detected monitoring the Helium gap pressure.
Description of the Safety Concept
ALFRED is equipped with two diverse, redundant and separate shutdown systems: the first system, that has also the control function, is made by absorber rods passively inserted by buoyancy from the bottom of the core; the second one is made of absorbers rods passively inserted by pneumatic system (depressurization) from the top of core. The Decay Heat Removal (DHR) system consists of two passive, redundant and independent systems, each one composed of four Isolation Condenser systems (ICs) connected to four SGs secondary side. Three out of four ICs are sufficient to remove the decay heat power. Both systems are passive, with actively actuated valves, equipped with redundant and diverse locally stored energy sources. 2D seismic isolators are installed below the reactor building to cut the horizontal seismic loads.
Deployment Status and Planned Schedule
The successful demonstration of the LFR technology is reaffirmed by the effective operation of a First-Of-A-Kind of the industrial-scale ELFR. The achievement of this goal by 2040 requires the realization of the ALFRED demonstrator around 2025.
Project Context and Objectives:
Concerns over energy resource availability, climate change, air quality, and energy security suggest an important role for nuclear power in future energy supplies. While the current Generation II and III nuclear power plant designs provide an economically and publicly acceptable electricity supply in many markets, further advances in nuclear energy system design can broaden the opportunities for the use of nuclear energy. To explore these opportunities, worldwide governments, industries, and research centres started a wide-ranging discussion on the development of next-generation nuclear energy systems known as “Generation IV.”in Europe. The EC organized the Sustainable Nuclear Energy Technology Platform (SNETP) that, through its Strategic Research Agenda promoted the development of fast reactors with closed fuel cycle The Roadmap proposed by the European Sustainable Nuclear Industrial Initiative (ESNII), includes the lead-cooled fast reactor as an alternative technology to be developed in parallel with the sodium-cooled fast reactor.
The LFR system features a fast-neutron spectrum and a closed fuel cycle for efficient conversion of fertile uranium and management of actinides. A full actinide recycle fuel cycle with central or regional fuel cycle facilities is envisioned. The LFR can also be used as a burner of actinides from spent fuel by using inert matrix fuel.
During the 1970’s and 80’s, considerable experience was developed in Russia in the use of Lead-Bismuth Eutectic (LBE) for reactors dedicated to submarine propulsion and to develop new reactor designs based on both LBE (i.e. the SVBR reactor) and lead (i.e. the BREST reactor) as primary coolants. In the USA, in the past considerable effort was devoted to investigations of lead corrosion and materials performance issues as well as system design of the SSTAR reactor, while more recently the focus has included the development of the desired characteristics and design of a technology pilot plant or demonstrator reactor.
More recently, an extensive R&D program on Lead technology was initiated in Europe and is still ongoing. In particular, the R&D activities are focused on suitable materials, surfaces coatings (oxide layers, alluminization, tantalum, etc.) and lead chemistry (corrosions inhibitors).
The first step in the development of a Lead Cooled Critical Fast Reactor in Europe started in 2006, when EURATOM decided to fund ELSY (European Lead cooled SYstem). The ELSY project, coordinated by Ansaldo Nucleare, developed a very innovative pre-conceptual design of an industrial plant for electricity production able to close the fuel cycle.
After the end of ELSY project in February 2010, the LFR development continued with the LEADER project (Lead-cooled European Advanced DEmonstration Reactor), started in April 2010 in the frame of EC 7th FP.
The LEADER project is based on the EC call for a conceptual design of a Lead Fast Reactor and it takes into account the indications emerged from SNETP, as well as the main goals of ESNII. The LEADER project is strongly committed to the conceptual design of a scaled/demonstrator reactor (European Lead Fast Reactor Technology Demonstrator Reactor – ETDR) to be constructed in the relatively short term. The ETDR, during the LEADER project kick off meeting, was named ALFRED (Advanced Lead Fast Reactor European Demonstrator).
The focus of the first part of the LEADER project activities has been the resolution of the key issues emerged in the frame of the previous projects (e.g. core not enough constrained, floating of fuel assemblies, risk of core bypass caused by not fixed connection between pump ducts and Inner vessel, too complicated FA upper restrains), to reach a new configuration. With reference to this updated configuration of an industrial size Lead Fast Reactor (ELFR), the design of a scaled, fully representative (ALFRED) has been performed.
The objectives of the project activities for ALFRED are:
- to define the main suitable characteristic and design guidelines for the facility;
- to use components/technologies already available in the short term to be able to proceed in the near future to a detailed design followed by the construction phase;
- to evaluate safety aspects and perform a preliminary safety analysis;
- to minimise the cost of the demonstrator.
Moreover, the demonstrator shall confirm that the newly developed and adopted materials, both structural materials and innovative fuel, are able to sustain high fast neutron fluxes and high temperatures.
The LEADER project involved 17 partners from Industry, research organisations and universities. The total effort was 502 person-months over a period of 36 months, then extended to 42 months.
Project Results:
The LEADER project deals with the development of Lead Cooled Reactor technology through two main goals: the conceptual design of an industrial-size LFR (the so-called European LFR or ELFR) and the conceptual design of a scaled down facility, the demonstration reactor called ALFRED (Advanced Lead Fast Reactor European Demonstrator).
The project concentrated its efforts mainly on ALFRED demonstrator design, since its full operation is expected in 2025, whereas the design and construction of the first of a kind 600 MWe industrial reactor (ELFR) is foreseen for 2040.
The main results of the project are summarized in the following, in terms of mechanical, neutronic and thermalhydraulic design, safety analyses, materials tests and evaluation of R&D needs and roadmap for the deployment of both the design.
ALFRED reference configuration
The configuration of the primary system is pool-type. This concept permits to contain all the primary coolant within the Reactor Vessel, thus eliminating all problems related to out-of vessel circulation of the primary coolant (Figure 1).
The primary coolant is molten lead, which presents several favourable characteristics, such as:
a very high boiling point (1745°C) and very low Partial Pressure. For this reason, there is no need to pressurize the plant primary side, resulting in a easier design of the reactor vessel. Moreover, lead boiling scenario is practically impossible and this, together with proper Reactor vessel design, drastically reduces core voiding risk.
It is chemically inert with air and water: this allows the elimination of any intermediate loop and the installation of the Steam Generator Unit inside the Reactor Vessel.
favorable neutronic characteristics (it is a low moderating medium and has a low absorption cross-section). This gives the possibility to have a fast neutron flux even with large amount of coolant in the core; there is no need of compact Fuel Assemblies and it is possible to have relatively large spacing among the fuel rods, with reduced core pressure losses and thus enhanced natural circulation capability.
density is slightly higher than that of the oxide fuel, thus, in the LFR, fuel dispersion dominates over fuel compaction, reducing considerably the likelihood of the occurrence of severe re-criticality events in the case of core disruption. In fact, the dispersion of fuel and molten clad from core active zone is enhanced resulting in a self-reducing extension of core degradation.
As for all Gen IV advanced systems, the development of the technology is associated with research challenges. In the case of the LFR, these challenges include:
• molten lead interacts with structural materials, mainly with the mechanisms of corrosion at high-temperature and erosion. The provisions that can be adopted to improve the compatibility of lead and steels are:
- Operate at low temperature range (400 °C - 480°C) and maintain a controlled amount of oxygen dissolved in the coolant.
- Select material, such as Austenitic low-carbon steels (e.g. AISI 316L), ferritic-martensitic steels (e.g. T91), 15-15/Ti steel.
- Utilize surface coatings, e .g. by alluminisation of surfaces (with Fe-Cr-Al-Y) and surface treatment by electron beam (GESA treatment).
- Limit coolant flow velocity to a value that cause a negligible erosion (typically 2–3 m/s).
• lead has a high melting point (327.4°C) and this can result in a risk of coolant freezing, although this is not considered to be a safety issue but an investment protection issue. To decrease this risk, LFR is equipped with active systems to keep the lead molten during any normal operational conditions (e.g. during planned shutdown). In case of accidental emergency conditions, a sufficient grace time (more than 30 min) is available for the operator to reduce the heat removed by the Decay Heat Removal System (e.g. shutdown one DHR loop by manually closing one valve). However, promising activities are on going on the DHR design, in order to have the reduction of the removed heat by passive means.
• lead is opaque and this makes the in-service inspection and handling of fuel assemblies more difficult. For this reason, each component inside the Reactor vessel is removable and the fuel assemblies upper end extends beyond the lead free surface in the cover gas for refuelling without the need of in-vessel machines.
PRIMARY SYSTEM ARRANGEMENT
The Reactor assembly (Figure 2) presents a simple flow path of the primary coolant, with a Riser and a Downcomer. The heat source (the Core), located below the Riser, and the heat sink (the Steam Generators) at the top of the Downcomer, allow an efficient natural circulation of the coolant. The primary coolant moves upward through the pump impeller to the vertical shaft, then enters the SG through the lead inlet holes, flows downwards on the shell and exits the steam generator.
The free level of the hot pools inside the Steam Generator & Primary Pump units is higher than the free level inside the Inner Vessel, the different heads depending on the pressure losses across component parts of the primary circuit. The volume between the primary coolant free levels and the reactor roof is filled by a cover gas plenum.
The Reactor Vessel (RV) is cylindrical with a torospherical bottom head. It is anchored to the reactor cavity from the top, by means of a vessel support. The upper part is divided in two branches by a “Y” junction: the conical skirt that supports the whole weight and the cylindrical one, that supports the Reactor Cover. A cone frustum, welded to the bottom head, has the function of bottom radial restraint of Inner Vessel.
A steel layer covering the reactor pit, constitutes the Safety Vessel (SV). The dimensions of gap between the safety vessel and the reactor vessel is sufficient to the space for reactor vessel ISI tools. The safety vessel is cooled by the same system that cools the concrete of cavity walls. This system is inserted inside the concrete and is independent from the reactor cooling systems. This design solution mitigates the consequences of through-wall cracks with leakage of lead: any reactor vessel leakage is discharged into the Safety Vessel. The RV and the SV are arranged in such a manner that in case of a reactor vessel leak, the resulting primary coolant always covers the SG inlet and the lead flow path is indefinitely maintained.
The Inner Vessel (IV), Figure 3, has two main functions: Fuel Assemblies support and separation between hot plenum and cold plenum. It is fixed to the cover by bolts and is radially restrained at bottom. Lead flow is guided from the FAs outlet towards the PP inlet pipes by a toroidal half-ring. Moreover, the pipes that connect the hot zone with the inlet of PP are integrated in the Inner Vessel.
The cylindrical IV has a double wall shell: the outer thick wall has a structural function, while the inner thin wall follows the core section profile. The Core Lower grid is a box structure with two horizontal perforated plates connected by vertical plates. The plates holes are the housing of FAs foots and the plates distances must be sufficient to assure the verticality of FAs. The diagrid is mechanically connected to the inner vessel with pins (possible removal/replacement during reactor lifetime). The Core Upper grid is a box structure like the lower grid but more stiffer. It has the function to push down the FAs during the reactor operation. A series of preloaded disk springs press each FA on its lower housing. A hole is present for each disk to allow the passage of instrumentation (i.e. thermocouples).
CORE
The adopted core configuration is constituted by wrapped Hexagonal Fuel Assemblies. It utilizes MOX as fuel and uses hollow pellets, a low active height and a large fuel rod pitch (thanks to the lead physical properties) in order to improve the natural circulation. The total power is 300 MWth.
The core scheme is made of 171 Fuel Assemblies (FAs), 12 CR (Control Rods) and 4 SR (Safety Rods), surrounded by 108 Dummy Elements (ZrO2-Y2O3) shielding the Inner Vessel.
Each Fuel Assembly (Figure 4) is about 8 m long and consists of 127 fuel pins, fixed to the bottom of the wrapper and restrained sideways by grids (the wires can not be adopted due to the selected fuel pitch). Tungsten deadweight (Ballast) prevents buoyancy forces in lead. Upper elastic elements (cup springs) prevent lifting induced by hydrodynamic loads and accommodate axial thermal expansions. The FAs upper end extends beyond the lead free surface in the cover gas for easy inspection and handling. In this way it is possible to make the refuelling without the need of in-vessel refuelling machines.
ALFRED is equipped with two diverse, redundant and separate shutdown systems (adapted from the one that is under investigation in the frame of the CDT-MYRRHA project):
(1) CR (Control Rod) system, used for both normal control of the reactor (start-up, reactivity control during the fuel cycle and shutdown) and for SCRAM in case of emergency. The Control rods are extracted downward and rise up by buoyancy in case of SCRAM. The control mechanism pushes the assembly down with a ball screw, placed, with its motor and resolver atop the cover (at cold temperature (<70°C)), and protected from radiation by a shielding block. The actuator is coupled to a long rod by an electromagnet. When the coupling electromagnet is switch off (in case of SCRAM), the absorber assembly and the rod are free to rise up. Control rods use a 19 pins absorber bundle, cooled by the primary coolant flow. These pins are fitted with a gas plenum collecting the Helium and Tritium, produced by nuclear reaction of B10
(2) SR (Safety Rod) system, is the redundant and diversified complement to the control rods for SCRAM) only. The absorber bundle stays in the primary coolant. The rod is extracted upward and inserted downward against the buoyancy force. The absorber gets inserted by the actuation of a pneumatic system. In case of loss of this system, a tungsten ballast will force the absorber down by gravity in a slow insertion. For both systems the materials considered are B4C enriched in 10B at 90% as absorber, T91 for the guide tube, 15-15 Ti for the clad and ZrO2 (95%) - Y2O3 (5%) for the insulator and the reflector.
Both systems have been designed in order to be inserted also in case of core deformation (e.g. proper clearance and adaptive rollers to follow the deformation, improved mechanical resistance of guide tubes).
STEAM GENERATOR AND PRIMARY PUMPS UNIT
The steam generator and primary pump are integrated into a single vertical unit. Eight SG/PP units are located in the annular space between the cylindrical inner vessel and the reactor vessel wall. The primary pump is placed in the hot side of the steam generator, having its mechanical suction in the hot pool inside the inner vessel. The primary coolant moves upward through the pump impeller to the vertical shaft, then enters the SG through the lead inlet holes, flows downwards on the shell and exits the steam generator (Figure 5). The thermal stresses on SG tube plates, caused by the large temperature difference between the feedwater and the super heated steam, have been evaluated by means of the FEM Ansys showing large stress margins.
Each Steam Generator (Figure 6) consists of a bundle of 542 bayonet tubes immersed in the lead vessel pool for six metres of their length.
The bayonet tube is a vertical tube with external safety tube and internal insulating layer, composed by 4 concentric tubes (Figure 6): slave tube, inner tube, outer tube and outermost tube. The internal insulating layer (delimited by the slave tube) has been introduced to ensure the production of superheated dry steam: in fact, without an insulating system, the high ∆T (≈115°C) between the rising steam and the descending feedwater would promote steam condensation in the upper part of the steam generator. The gap between the outermost and the outer bayonet tube is filled with pressurized helium and high thermal conductivities particles (such as synthetic diamonds) to enhance the heat exchange capability. In case of an external tube break this arrangement guarantees that primary lead does not interact with the secondary water. Moreover, a tube break can be easily detected monitoring the Helium gap pressure.
The Primary Pump (PP) is located in the hot side of the SG because this strongly simplifies the internals components. However, the temperature difference between the cold and the hot side is not significant (80 °C). It is surrounded by the SG tube bundle and its housing is the Steam Generator casing. The Pump is fixed to the top of SG casing by a bolted joint. This allows an easy removal of the component.
The Primary Pump studies are in progress. Based on analyses performed during previous LFR projects (EUROTRANS, ELSY) an axial pump has been adopted. The PP impeller material is still an issue and test campaigns must be performed to select the proper one. The stainless steel (SS) cannot be used because the high speed achieved at the top of the impeller blades (about 9.8 m/s) induces a very fast material corrosion. Maxthal ceramic material has been proposed, but its reliability must be still demonstrated. An alternative solution can be a SS impeller with a coating, like Tantalum, already available in the chemical industry, to be qualified for nuclear application.
DECAY HEAT REMOVAL SYSTEM (DHR)
The Decay Heat Removal system (DHR) consists of two passive, redundant and independent systems, DHR1 and DHR 2, both composed of four Isolation Condenser systems (ICs) connected to four Steam Generators (SGs) secondary side (i.e. one IC for each SG). The system design considers the single failure criteria, since three out of four ICs are sufficient to remove the decay heat power.
The DHRs are dedicated safety systems, not used for normal operation. The separation is achieved through placing the two DHRs in physically different locations. A physical structural barrier or another means of protection will be placed between adjacent IC to ensure that failure of one of them could not harm another one. The diversity requirement has been relaxed (in any case the two DHR systems will be fabricated by diverse manufacturers) due to the high redundancy and considering that the SG tubes bayonet concept allows a continuous monitoring of the SG status.
Both systems are completely passive, with an active actuation (valves) that are however equipped with redundant and diverse energy sources (batteries or locally stored energy). Each DHR system must be ready to operate after the reactor trip in order to remove the decay heat power, in case of unavailability of the normal path (i.e. the by-pass to the Condenser). The actuation logic shall guarantee the actuation of DHR1 first, and of DHR2 only in case of failure of the first system and that the total number of isolation condensers called to operate never overcomes the four units, in order to avoid an excessive cooling of the primary coolant leading to fluid solidification.
Each of the four independent IC sub-systems consists of (Figure 7):
One heat exchanger (isolation condenser), constituted by a vertical tube bundle with an upper and lower horizontal header.
One water pool, where the isolation condenser is immersed; the amount of water contained in the pool is sufficient to guarantee 3 days of decay heat removal operation (after 3 days, the pool can be refilled with onsite ancillary water or, in case of extreme accident, by fire trucks).
One condensate isolation valve (to meet the single failure criteria this function shall be performed at least by two parallel valves).
Each isolation condenser is connected to a Steam Generator: the upper header of the isolation condenser is connected to the main steam line and the lower header of the Isolation Condenser is connected to the main feed water line. In normal operation the isolation valve below the condenser is closed, the condenser is full of water and no heat exchange takes place. To put in operation the component, the feed water line and steam line must be isolated and the condenser isolation valve opens injecting water into the Steam Generator. The steam produced in the steam generator now can reach the Isolation condenser starting the steam condensation process.
ALFRED CORE DESIGN
A comprehensive core design strategy has been adopted, approaching systemically, from a thermal-hydraulic, thermo-mechanic and neutronic point of view, all the aspects of the reactor core impacting on the safety and the robustness of the plant. As a matter of fact, such an approach allows for the a priori definition of a core which respects, at least in first approximation, all the technological constraints and fulfils all the performance objectives by design. Then the results are checked a posteriori with more detailed code calculations.
According to this, the ALFRED fuel pin, the fuel assembly, the control and safety systems, the reflector and the core have been designed. In particular, preliminary analyses have been performed to check whether the investigated fuel pin and the coolant sub-channel allow for an acceptable thermic of the pin and a sufficient natural circulation in ULOF conditions without exceeding some safety limits, mainly linked to the clad temperature.
Fuel pin and coolant sub-channel design
The first phase of the core design process foresees a preliminary thermal/hydraulic analysis to check the respect of the design limits to what concerns both the fuel and cladding maximum temperatures in the hottest channel. Since the latter – together with the coolant inlet and outlet temperatures – are the same as for the ELFR, the design of the fuel pin has been borrowed by this system, together with the maximum linear power rating (≈ 340 W/cm). The same assumption holds considering the target peak BU (≈ 100 MWd/kg), so that the hollowed fuel concept of the ELFR has been kept to accommodate the irradiation-induced swelling without incurring in Pellet-Cladding Mechanical Interaction (PCMI). The resulting (radial) fuel pin design is shown in Figure 8.
The design of the sub-channel cannot omit preliminary safety analyses to verify the respect of the cladding integrity for the desired grace time (30 minutes before SCRAM is manually actuated) in case of accident. To account for this constraint, the natural circulation regime during a ULOF transient has to be evaluated. Having this in mind, the sub-channel has to be designed so as to offer a hydraulic diameter the pressure losses associated to which can be overtaken with a small prevalence, attained through acceptable coolant temperatures.
On the other hand, the wider the pitch, the lower the fuel volume fraction in the elementary cell, to the detriment of the reactivity of the system. According to this, two values of the nominal coolant flow velocity have been therefore assumed for the present study, 1.4 and 1.5 m/s, supposing these as reasonable for achieving a sufficient natural circulation without affecting excessively the core reactivity. Furthermore the choice of the preliminary core configuration has to cope also with the constraint imposed for the inner vessel radius.
At last, two active heights (60 and 90 cm) have been considered in the present study not only to encourage the natural circulation by increasing the chimney height (since the shorter fuel and plenum length), but also to allow for keeping in the width of the lattice pitch, gaining in reactivity. The natural circulation regime in case of ULOF has been then evaluated by means of the SIMMER code for the four different configurations resulting by combining the two active height options with the two coolant velocities. According to the results of this investigation, the configuration corresponding to a coolant velocity of 1.4 m/s and 60 cm active core height has been chosen as reference, to which corresponds a value of the maximum cladding temperature during ULOF below 740 °C.
For the final dimensioning of the coolant sub-channel, the radial and axial power form factors had to be evaluated. Concerning the axial form factor, a value of 1.15 has been borrowed from the analyses performed on a similar system. The radial power form factor has been set equal to 1.2 assuming a proper power/FA distribution flattening is achievable through suitable Pu enrichment zoning of the core. According to these assumptions – which automatically become recommendations for the neutronic design of the core – the corresponding pitch for the fuel pins lattice is set to 13.86 mm.
Core design
Starting from the maximum linear power rating and the axial and radial power form factors, the average linear power rating is computed and used to define the number of fuel pins to be arranged to sum the aimed 300 MWth. The corresponding number of FAs (171) can be arranged to form a cylindrical core, together with 16 positions for Control and Safety Rods (CRs and SRs respectively), surrounded by two rings of dummy elements shielding the inner vessel within the imposed radius. Extensive neutronic calculations – by means of both MCNPX and ERANOScodes with the JEFF 3.1 cross-sections library– followed, to define the Pu enrichments and zoning, along with the positioning of control and safety elements, ensuring the aimed criticality level, a power/FA distribution flattening complying with the assumptions made for the design of the fuel pin and coolant sub-channel, and anti-reactivity worth for control and safety systems ensuring the respect of safety constraints. The underlying hypothesis is a fuel management according to a 5-batches cycle without reshuffling, with 5 y FA residence time in pile. The latter comes directly from the assumption on the maximum linear power rating and the chosen pellet geometry, to target the aimed peak burn-up of 100 MWd/kg.
The requirements for the control/safety systems are:
- a sufficient worth of the CRs to compensate the reactivity swing during the 1-year sub-cycle;
- the worth of each single CR when inserted to compensate the criticality swing not higher than 1 $;
- 5000 pcm for the refuelling status (keff = 0.95);
o about 2500 pcm for the SCRAM. This has been set according to:
o some 500 pcm for the transition from hot to cold state;
o 1000 pcm of sub-criticality, keff = 0.99 (value to be verified by kinetics analyses);
o pcm of uncertainty (assumed).
The strategy to accomplish these requirements foresees:
- two redundant systems for SCRAM (2500 pcm each), one of which also devoted to the compensation of the reactivity swing during the cycle;
- the simultaneous use of both CR and SR systems during refuelling to reach keff = 0.95.
The resulting core configuration is characterized by 57 FAs in the inner region, with a Pu enrichment of 21.7 wt.%, and 114 FAs in the outer region, with a Pu enrichment of 27.8 wt.%. 12 CRs are positioned within the outer region with the twofold goal of lowering the flux insisting on the most enriched FAs, while compensating the over-criticality along the cycle through a progressive withdrawal from their maximum-inserted position at BoC. 4 SRs are also positioned close to the core centre to minimize the shadow effect with the CRs. The ALFRED core, and the power/FA distribution at Beginning and End of Cycle (BoC and EoC) are shown in Figure 9.
For this configuration, a criticality swing of about 2600 pcm is found during the 1 year cycle. The worth of the control and safety systems is found to be –8500 pcm and –3300 pcm, respectively. Even in case of one SR stuck in withdrawn position, the anti-reactivity provided by the safety system is still about 2600 pcm. These values are found sufficient for the achievement – in every condition – of the required anti-reactivity margins for control, SCRAM and refuelling. To compensate the reactivity swing during the sub-cycle, a partial insertion of the CRs by some 17 cm is sufficient. This is the main responsible for the radial and axial distortion of the power distribution in the core. Thanks to the high number of CRs, the limit for the maximum worth of 1 CR at BoC is also respected, being 220 pcm the maximum accidental reactivity insertion against 335 pcm of delayed neutrons fraction.
For the most powerful FAs in the two zones the detailed power/pin distribution has been computed by means of MCNPX. A complete thermal analysis of the fuel element followed, in order to verify the respect of the maximum fuel and cladding temperatures.
This analysis has been performed by means of the ANTEO-LFR sub-channel code originally developed in ENEA for fast reactors fuel assembly thermal/hydraulic studies, and adapted to lead coolant.
The distribution of the coolant temperature at core outlet among the sub-channels of both these FAs is shown in Figure 10. For the FA in the outer region, since the more unfavourable power distribution within the FA itself, a complete analysis of the fuel pins followed, showing the respect of the design limits even for the hottest pins (Figure 11).
RESULTS OF THE SAFETY ANALYSES TRANSIENTS PERFORMED FOR ALFRED
Safety of ALFRED is extensively based on the use of the defense in depth criteria, enhanced by the use of passive safety systems (actively actuated through local stored energy sources, always available, and fully passively actuated). Safety features of the LFR system have been designed since the beginning of LFR related activities to face challenging plant conditions and events, taking into account the forgiving and advantageous physical characteristics of the coolant. As an example, there is no need for off-site or emergency AC electrical power supply to manage design basis accident conditions, as the only manual action needed is the supply of additional water to maintain the water level in the decay heat removal (DHR) pools that are sized to guarantee at least three days of unassisted, fully passive operation. These pools can be easily re-supplied with water in the subsequent days.
Design Basis Conditions Transients
Within the LEADER project all of the most important DBC transient initiators were analyzed for the ALFRED reactor. A selected set of the analyzed DBC transients are briefly presented hereafter
Spurious withdrawal of the most reactive control rod (PTOP)
For this simulation it was assumed that: 1) at t = 0 sec, the event of spurious withdrawal of the most reactive control rod (CR worth: 250 pcm) starts; 2) CR withdrawal velocity by mechanical means is limited to 0.1 mm/s, or 0.15 pcm/s; 3) as all CRs at EOC (End-Of-Cycle) condition are withdrawn from the active core region, the BOC (Beginning-Of-Cycle) condition is used for this simulation as all CRs are inserted into the active core region by 16 cm, 4) spurious withdrawal of the most reactive CR proceeds, and at t = 524 sec, a reactor scram signal is being generated based on high neutron flux > 120 % nominal, but this signal is being neglected; 5) at t = 548 sec, the second reactor scram signal is generated based on peak power FA temperature difference being > 1.2; 6) based on this reactor scram signal, the reactor is being shutdown at t = 549 sec into the transient; 7) main steam lines and main feedwater lines are then closed in ~2 sec time interval, while condensate isolation valves of four ICs (Isolation Condensers) of the DHR-1 system fully open in ~ 20 sec, removing maximal 7 MW of heat from the primary cooling circuit.
As can be observed in Fig. 12, CR withdrawal out of the active core region continues for 549 sec until the reactor trips. During this time interval a total positive reactivity of ~83 pcm is being inserted into the core. Insertion of ~83 pcm in 549 sec time interval at BOC conditions leads to an increase in reactor power ~ 1.2 nominal. The maximum fuel and clad temperatures increase from 2028°C and 516°C to 2313 °C and 555 °C respectively. Minimum clad failure time of the peak pin during the transient always remains above 1E+11 sec and thus no fuel pin failure (creep rupture) is expected.
As related to the above study, one can observe that for reactivity insertion of ~83 pcm in 549 sec time interval of the most reactive control rod (250 pcm total worth at BOC, HFP: Hot Full Power conditions), the ALFRED peak fuel pin cladding accommodates this transient; in addition no fuel melting is expected.
Loss of AC power (PLOOP)
This transient is initiated by assuming the total loss of offsite power supply (station blackout). As a consequence, the forced circulation in the primary system is lost (pump coastdown) with simultaneous turbine and feedwater trip on the secondary side and reactor scram. The secondary circuits are isolated by the main steam isolation valve closure, and the decay heat removal by the DHR-1 system (four IC loops are supposed in service) is promptly activated through the opening of the triggering valve positioned below the isolation condenser.
In spite of the very low primary pump inertia (speed halving time < 1 s), there is a smooth core flow rate reduction in the initial phase of the transient (see Fig. 13), due to lead free levels equalization in the primary system. As a result, the initial clad peak temperature increase is not significant (max T-clad = 564 °C) from the safety point of view.
The core decay power is first removed by water evaporation inside the MHX and then by steam condensation inside the isolation condenser immersed in the boiling water pool, with final decay power release to the atmosphere. The whole system functions passively in natural circulation; the pool water storage is guaranteed for at least three days of DHR-1 operation. Additional water can be easily supplied manually, thereby assuring practically unlimited operation.
The Fig. 14 shows that the core decay power is efficiently removed by the DHR-1 during the whole transient phase. The MHX power in excess to IC power in the initial phase is removed by steam release through the safety relief valves of the over-pressurized secondary circuits. After about 15 minutes, the DHR-1 power removed (about 7 MW for 4 IC loops) surpasses the core decay heat and the lead temperatures in the primary circuit start to decrease (see Fig. 14). After 3 hours the minimum lead temperature is reached at the MHX outlet (350 °C), still sufficiently distanced from the lead solidification point (327 °C). Manual or automatic control of the IC unit operation is needed in the long term to avoid potential lead freezing. Passive control of the operation of the IC units is under investigation in order to exclude the needs for manual intervention in the control systems.
Based on the above results one can conclude that, in case of total loss of offsite power supply, the core decay heat can be removed safely indefinitely in a passive manner without the need of active energy sources.
Protected partial flow blockage in the hottest fuel assembly
Transient analysis of the unprotected partial blockage of the hottest SA w/o radial heat transfer for ALFRED can be summarized as follows:
(1) ALFRED will not experience any fuel pin failure for flow blockage less than 75%, even under unprotected conditions;
(2) For flow blockages above 75%, clad failures (creep rupture) should be expected;
(3) Fuel melting is not expected to be a safety issue for the ALFRED as fuel melting temperatures are not reached, even in 97.5% SA flow blockage case.
Using these results, the influence of delaying reactor trip by 1, 2, 3, 5 and 10 sec assuming a 70% SA flow blockage was analysed. Blockage was assumed activated at t = 1 sec transient time, followed by reactor trip (signal based on the peak SA temperature difference value exceeding 1.2 nominal) assuming different time delays in order to study the impact of the trip signal delay on the maximum fuel and clad temperatures. Reactor trip thus occurring at time t = 4, 5, 6, 8 and 13 sec into the transient. Main steam lines and main feedwater lines are being closed in 20 sec time period, at the same time opening the condensate isolation valves of four ICs of the DHR-1 system, removing maximal 7 MW of heat from the primary cooling circuit.
Simulations showed that the maximum clad and coolant temperatures are achieved when the reactor trip is being delayed for 10 sec. The maximum increase of the peak clad temperature in this case is 172 oC for BOC (from initially 516 oC) conditions, and 185 oC for EOC (from initially 514 oC) conditions. The maximum increase of the peak coolant core outlet temperature is 184 oC for BOC conditions (from the initial value of 487 oC) and 195 oC for EOC conditions (from initially 487 oC). Later, following reactor trip, all core and primary cooling circuit temperatures decrease.
The 70% SA flow blockage case for 10 sec reactor trip delay at EOC is shown below in Fig. 15.
The maximum clad temperature of the peak pin increases from 514 oC up to 699 oC (Fig. 15) at transient time t ~ 14 sec. At the same time point minimum clad failure time of the peak pin drops down to 5E+6 sec (Fig. 5), later recovering to 5E+16 sec, indicating that the clad material is not challanged. The fuel temperature of the peak pin is the highest at the beginning of the transient (2064 oC), decreases after reactor trip. All other temperatures around the primary cooling circuit also decrease following reactor trip.
The performed analysis demonstrated that the clad of the peak pins of the ALFRED reactor has a very large margin to clad failure (rupture) during the simulated peak power SA for the 70% flow blockage transient. Reactor safety is ensured by the reactor protection system, shutting down the reactor, in conjunction with the DHRS, efficiently removing the decay heat from the reactor primary cooling circuit. Even a 10s delay in reactor trip does not challenge the integrity of cladding material in ALFRED.
Reduction of FW (feed water) temperature from 335 oC to 300 oC
The over-cooling (OVC) transient was initiated by the reduction of feedwater temperature from 335 °C to 300 °C within 1 second. When feedwater temperature was lower than 327 °C, the SCRAM signal was triggered to shutdown the reactor with 1 second delay. Later, main steam lines and main feedwater lines in SG system were closed in 2 seconds. At the same time, condensate isolation valves are fully opened, leading to the functioning of DHR systems, that are able to provide 7 MW heat removal capability. The opening process of condensate isolation valves takes 20 seconds. During the whole transient, safety valves maintain the pressure in isolation condensers under 195 bar and temperature of liquid water leaving isolation condensers is kept at 300 °C.
Simulation results of the OVC transient are presented in both short-term and long-term time scales.
When condensate isolation valves start to open, liquid water stored in isolation condensers will flow into steam lines and evaporate instantly, leading to a significant increase of MHX power up to 140% nominal. Thereafter, MHX power will continuously decrease with the core power, but it is still sufficient to reduce the temperature difference between the core outlet and core inlet, until only negligible temperature difference between core inlet and outlet remains, as it is shown in Fig. 16.
As a design basis, the rated power of one DHR system is 7 MW, which will be higher than core power after 900 seconds decay. Hence, primary coolant temperature will decrease slowly as it is shown in Fig. 17. After 14000 seconds, the lowest coolant temperature in the primary circuit (at the MHX outlet) will be lower than the freezing point of liquid lead (~ 327 °C), lead at that location will freeze, unless manual intervention will reduce the DHRS heat removal capability by shutting down appropriate sub-systems of the DHRS.
Increase of FW flowrate by 20 %
This transient is initiated by a control malfunction on the secondary side leading to a an increase of the feedwater flowrate (+20% in 25 s) at the inlet of all MHXs.
As a consequence of the feedwater flowrate variation, the power removal through the MHX increases with corresponding reduction of the lead temperature on the primary side at the MHX outlet as shown in Fig. 18. The maximum MHX power increase is about 10% nominal. The temperature perturbation from the MHX outlet propagates through the core and then back to the MHX inlet in about one minute, starting the decrease of the MHX removed power. The core temperature decrease leads to a total positive reactivity insertion (see Fig. 19), mainly driven by radial core, coolant and control rod drive expansion, with consequent progressive core power increase that comes to equilibrium with the MHX power in about 5 minutes. After this time a new steady-state conditions is reached in both primary and secondary systems, without exceeding any reactor trip threshold set-point. The maximum temperature decrease calculated at the core inlet is of 13 °C, while the core power and the core ΔT increase is below 6%. Furthermore, the fuel peak temperature increase is not very significant (about 70 °C, see Fig. 19). From the performed analysis one can see that this transient has no safety implications for the ALFRED reactor.
Loss of all primary pumps (PLOF)
This transient is initiated by the loss (with subsequent coastdown) of all primary pumps at t = 0 s. The reactor scram on low primary pump speed signal should be effective just after 1 s, because of the low pump speed halving time (< 1 s), but for conservative analysis this first scram signal is neglected. Therefore, the automatic reactor shutdown is actuated at t = 3 s, with one 1 s delay after the second scram signal, based on the ΔT increase through each individual FA (ΔT > 1.2 nominal value), is detected (t = 2 s). Following the reactor scram the secondary circuits are promptly isolated (main steam isolation valve closure and feedwater trip) and the DHR-1 system is actuated to remove the core decay power. For conservative analysis against maximum core temperature reached, only three out of four IC units of the DHR-1 system are supposed to be in service.
After the loss of primary pumps, the transition from forced to natural circulation in the primary circuit is similar to the one already observed in sub-section above for the loss of AC power transient. However, since in this case the reactor scram is delayed, the peak clad temperature increases reaching a maximum at t = 4 s slightly below 650 °C (see Fig. 10). After reactor scram the lead and clad temperatures at the core outlet progressively reduces, according to the residual core flowrate and the core decay power level.
Despite the reduced availability of the DHR-1 system (3 IC unit), which is now able to remove a total of about 5 MW power during the transient, the DHR-1 heat removal surpasses the core decay heat after about one hour, and then the primary temperature tends to reduce (see Fig. 20). The core temperature increase at the core outlet after the initial transient phase is not significant from the safety point of view. In fact, the maximum core temperature rises up to about 475 °C, that is below the normal operating value.
The risk of lead freezing at the MHX outlet after DHR-1 start-up is excluded, since the minimum lead temperature (350 °C) is far enough from the freezing point of 327 °C. After 3 hours the lead temperature at the MHX outlet is still above the lead solidification point. As in case of loss of AC power, manual or automatic control of the IC unit operation is needed in the long term to avoid potential lead freezing.
In spite of the conservative assumptions regarding the failure of the first scram signal and of one IC unit of the DHR-1 system, one can see that, also in case of total loss of forced circulation in the primary circuit, the protection and safety systems are able to bring and maintain the plant in safe conditions in the short and long term.
Conclusions
The above results show, that ALFRED has very large grace times (~hours). Calculated peak pin temperatures are not exceeding 700 °C, thus no clad failures (creep rupture) are expected, even during the loss of AC power (PLOOP) and loss of all primary pumps (PLOF) transients. This is mainly due to the large thermal inertia of the Pb-cooled primary system and the relatively high Pb natural convection core mass flow rate observed at the initial stage of the transient. In the long-term, in case of PLOOP or PLOF transients, following reactor trip, core decay heat can be passively and safely removed indefinitely by the DHR-1 system without the need of active energy sources.
Large margin to clad failure (rupture) for the clad of the peak pins of the ALFRED reactor was confirmed also during the simulated peak power SA 70% flow blockage transient. Reactor safety in this case is ensured by the reactor protection system, shutting down the reactor, as well as by DHRS, efficiently removing the decay heat from the reactor primary cooling circuit. Reactor trip delay even by 10 sec does not lead to clad failure for the ALFRED reactor.
During the protected FW temperature decrease to 300 oC transient, apart from a safety issue of Pb-freezing several hours into the transient at the outlet of the MHX for the ALFRED reactor, no other safety related issues are observed. During this time period the operator has sufficient time to deactivate sub-systems of the DHR-1 system, in order to prevent Pb-coolant freezing at the outlet of the MHX. In case of FW temperature reduction to 300 oC, it is actually recommended not to trip the reactor, thereby assuring that the primary coolant will not freeze at the MHX outlet.
As a general conclusion, no relevant, immediate safety issues have been identified during the performance of the simulated transients aside of potential Pb-freezing several hours into the transient in case of all protected DBC transients. Reactor safety is assured by the reactor protection system, shutting down the reactor, in conjunction with a heat removal system (DHRS), removing tightly controlled amounts of decay heat from the reactor primary cooling circuit to assure prevention of freezing of the Pb-coolant at any location of the primary cooling circuit.
As related to the DBC transients, all selected transients examined proved that the ALFRED plant can accommodate a rather wide range of accidental events. The ALFRED plant has proved to be able to enter a safe shutdown phase after every DBC accident analyzed.
The analysis indicated, that ALFRED is a very forgiving plant design, and there is an extended time margin (grace time) of several hours for possible manual operator intervention even under worst accidental conditions (potential of Pb-freezing in the long term, in case of uncontrolled decay heat removal by the DHRS).
Design Extended Conditions Transients
The DEC events that are representative for the safety analysis of ALFRED have been identified by means of common safety approach adopted for liquid metal fast reactors and on the basis of engineering judgment, taking into account the specific features of the ALFRED design. These events of very low frequency are characterized by the failure of reactor scram (so called unprotected accidents) involving:
- Unprotected reactivity insertion transient (UTOP),
- Unprotected loss of flow transient (ULOF),
- Unprotected loss of heat sink transient (ULOHS),
- Unprotected loss of flow and heat sink transient (ULOF+ULOHS),
- Unprotected partial FA blockage,
or by the simultaneous occurrence of multiple failures in some protected transients:
- Loss of all secondary circuits with total unavailability of DHR system,
- Loss of all primary pumps with reduction of feedwater temperature (loss of one preheater),
- Loss of all primary pumps with increase of feedwater flowrate by 20%.
The analysis of the representative DEC transients for ALFRED has highlighted very good intrinsic safety features of the reactor design. In particular, the results of the unprotected transients underline that the reactor can be maintained in a safe and controlled state even under the unlikely accident conditions, which include the failure of the reactor scram, thanks to:
- Establishment of enhanced and stable naturalconvection in the primary cooling circuit following the loss of the primary pumps,
- Dominant negative reactivity feedback effects obtained by optimizing the neutronic core design,
- Large thermal inertia of the primary cooling system,
- Passive and efficient operation of the DHR system for decay heat removal.
As a general conclusion, no relevant safety issues have been identified for ALFRED in case of representative DEC events. In particular, the ULOF transient can be accommodated without the need of corrective operator actions. Finally, the analysis of DEC transients for the lead-cooled ALFRED design has demonstrated the extremely forgiving nature of this plant when compared to other similar plant designs.
MATERIAL SELECTION
The operating temperatures of the different reactor components are identical for both of the discussed nuclear reactors, ALFRED and the ELFR (Table 1). Therefore, most of the materials selected to manufacture the components like reactor and inner vessel and refuelling equipment are identical for both concepts. One major difference between ALFRED and the ELFR is the expected maximum irradiation damage of the fuel assemblies. The ALFRED design foresees 100 dpa, what allows the use of the 15-15Ti stabilized austentic steel also employed in the French Phenix reactor. This steel is already licensed for application in nuclear reactors, which is an additional important aspect for material selection for ALFRED. The fuel assembly of the ELFR needs to withstand a maximum radiation damage of 200 dpa that is too high for the best available 15-15Ti stabilized steel. The class of steels that can tolerate such high radiation damage are the ferritic/martensitic (f/m) steels, like T91 or specific ODS types.
Besides improved irradiation stability, the heat conductivity of the f/m steels is almost a factor of two higher compared to austenitic steels like 316L and 15-15Ti stabilized. Therefore, T91 is also discussed as structural material for heat exchangers and steam generator (SG) components. The thermally most loaded components like fuel claddings and steam generators or heat exchangers, where coolant temperatures of 550°C can be reached, with cladding wall temperatures even 20K higher, might require surface aluminized steels to be operated. For most components, beside the pump impeller, the maximum Pb velocity of 2m/s does not require specific attention.
Compatibility of selected materials with Pb
Compatibility of materials with Pb is mainly driven by the temperature dependant solubility of the alloying elements in the liquid Pb. Ni, Mn, Al and Si have substantial solubility of up to some wt% at 550°C. The main elements of steel, Fe and Cr, are less soluble but still some 10-5wt% can not be neglected. Practically insoluble are refractory alloys like W, Mo and Ta and most ceramics. Therefore Ni containing metals like austenitic steels suffer more from dissolution attack than Ni free f/m steels like T91. To mitigate dissolution any direct contact between the steel and the liquid Pb should be avoided. The most appropriate method is the in-situ oxidation by dissolved oxygen. The oxygen concentration range is determined by the operating temperature range of the reactor. The lower operating temperature of 380°C, to consider some safety margin 325°C (Tmelt of Pb), ascertains the upper limit of 10-4wt% oxygen to avoid PbO formation and any related coolant channel blockage. The highest achievable temperature at normal operation (550°C fuel clad) determines the lower limit of dissolved oxygen of 10-7wt% required to prevent or at least mitigate dissolution attack. Such formed oxide scales act as diffusion barrier and can prevent the dissolution. However, especially in case of f/m steels the growth of oxide scales can, at least after longer exposure times, result in severe reduction of the heat conductivity.
Austenitic steels – 316L, 15-15Ti (1.4970):
The operating range for the austenitic steel 316L is, as depicted in table 1, between 380 and 480 °C. The 15-15Ti that should be employed as fuel assembly material can reach temperatures up to 550°C or even higher due to reduced heat transfer properties by the formation and growth of oxide scales. The upper limits are due to limitations by the material compatibility and might be altered if suitable solutions are available. In previous national and international projects Ni containing austenitic steels were mainly tested in Pb Bi eutectic (LBE). Due to the higher solubility of alloying elements in LBE than in Pb, a direct transfer of the attained knowledge is not simple. The general mechanism, the dissolution of Ni accompanied by penetration of Pb into the steel matrix, is equivalent for both coolants. However, the rate of dissolution and the resulting temperature and oxygen concentration limits might differ. Below 430°C, the maximum temperature of the reactor vessel, even in LBE no dissolution attack was observed as long as the oxygen content is sufficient high for the formation of oxide scales. In Pb a better behaviour of the steels can be expected. However, if the oxygen concentration drops to low values even at this temperature dissolution attack will become an issue. Tests with 316L steel showed that at 450 °C in LBE with low oxygen content dissolution attack already starts after 2000h (Fig 20a). In LBE at a temperature of 500°C dissolution attack is observed after about 10000h. In Pb at the same conditions regarding temperature and oxygen content oxide scales still protect the steel (Fig. 20b). The usability of the steel 316L up to the selected upper temperature limit of 480°C at “normal” operating conditions seems obvious if the oxygen is controlled within the given limits.
The 15-15 Ti stabilized steel (1.4970) foreseen as fuel cladding material, has to be stable up to 550°C or even 570°C. In LBE this temperature is clearly above the operating range for this steel. Local deep dissolution attack starting with leaching of Ni and later Cr will not allow the use of the 15- 15Ti in LBE. But, in Pb this steel showed in experiments at same temperature and oxygen activity the formation of protective oxide scales at least in the first 3000h of exposure. Based on the experience with extrapolation of corrosion data from short to long term experiments, it is unlikely or at least uncertain whether this steel can be operated in Pb at 550°C for longer times. One difficulty that arose during the experimental work is the so called incubation time for the dissolution attack. At 500°C e.g at 10-8wt% oxygen a typical oxide scale develops during exposure for 2000h [Fig 21a]. However, after 5000h dissolution attack wa111s observed. On possible mechanism is the enrichment of the high soluble elements like Ni and Mn beneath the outer oxide sale [Fig. 21b]. These enriched phases are located in the spinel layer and can reach almost the interface to the outwards growing magnetite scale. The magnetite scale is rarely a dense and compact layer and quite frequently penetrated by the liquid metal. Therefore, the Ni and Mn enriched phases might act during prolonged exposure as a starting point for the dissolution attack if they reach the original metal surface and come in interaction with the liquid metal. Therefore all short term tests (<5000h exposure) must be carefully analyzed on any sign of local enrichment that can possibly predict the onset or starting of a dissolution attack in future exposure times.
f/m Steel T91:
The operating temperature range for the f/m steel T91 is similar to that of the austenitic steel between 380 and 550°C. Again, the upper limit is not design but material driven and might be increased if adequate solutions are available. The f/m steel T91 does not contain Ni. Therefore, the susceptibility for dissolution attack is reduced compared to the austenitic steels. For the entire foreseen range of temperatures no dissolution attack of T91 steel is expected also for long exposure times if oxygen concentration is kept under control at nominal values. The formation of oxide scales, which is the prevailing mechanism for Ni free f/m steels, is mainly independent of whether the medium in contact is pure Pb or LBE. The oxidation potential as the driving force and the diffusion constants in the steel, both determine the oxidation behaviour, are anyway similar for both coolants. Above 480°C typically three layered oxide scales that grow significantly with time are formed (Fig. 22a). On average 30μm thick scales grow after 6700h. Assuming parabolic growth, oxide scales of about 40μm thickness are formed after 15,000h. At 550°C oxidation is accelerated and the scatter of the measured data becomes significant (Fig. 22b). The average thickness of the spinel scale grows up to almost 60 μm after 15,000h assuming parabolic oxidation. The large scatter makes long term prediction difficult.
Besides the lack of precise and reliable data for long term, the observed oxidation becomes an issue regarding heat removal through fuel claddings and SG and HX tubes. Temperature increase of cladding tubes due to the reduced heat conductivity of the oxide scale and the reduced efficiency of the SG and HX must be considered during design or answered by a different solution like aluminizing of the steel. Besides, oxygen supply has to consider this effect too.
Alternative materials
For the impeller of the pump the temperature range is between 380 and 480 °C. Regarding solely corrosion, both classes of steels, 316 type and T91 type, can be used. But, the expected local velocities of up to 10m/s or even higher might be problematic for the use of the materials described above.
Localized high velocity flow pattern can severely damage steels as observed in corrosion experiments at IPPE and CIEMAT. 316L steel was heavily eroded in such flow field at temperature and oxygen conditions that are suitable to form protective oxide scales at nominal flow velocities.
The materials discussed at the time are Maxthal, SiSiC or coatings (Ta).
Maxthal: Tests with Maxthal in Pb up to 750°C revealed its excellent compatibility with the liquid metal. The protective oxide scale, depicted in Fig. 23a, is quite thin and composed of TiO2, SiO2 and mixed oxides. This material is discussed as pump impeller material and has to withstand high local velocities.
SiSiC: SiSiC was also tested in Pb up to 600°C and did not show any incompatibility with the liquid metal.
Ta Coating: Ta as refractory metal has no solubility in Pb. Therefore coating of an impeller was discussed as an other alternative solution. Basic exposure tests in Pb showed the expected incompatibility of the Ta with oxygen containing Pb at higher temperatures. At temperatures above 400°C the oxygen diffuses into the Ta and embrittle the material, as seen in the hardness increase of Ta exposed at 500 and 550°C in Fig. 23b. For longer times and higher oxygen contents even entire oxidization of Ta sheets (200μm thick) were observed. The simultaneous use of Ta and oxygen in Pb seems not to be a suitable option for material protection in a Pb cooled nuclear reactor. Tests with a Ta coated steel impeller pump in oxygen reduced Pb are at the time under preparation. If successful the same impeller will be tested also in oxygen containing Pb at 480 °C to monitor whether the oxygen uptake and through that the embrittlement is significantly slowed down at this lower temperature.
At KIT the CORELLA facility was re-build to achieve rotation speeds higher than 1000 rpm and Pb velocities large than 10m/s. This objective was reached and two corrosion/erosion tests with candidate materials were performed at 480°C. The 1st test was stopped after 100h, because the target oxygen content of 10-6 wt% oxygen could not be achieved, the measured value was always far below 10-8wt% oxygen. However, the samples were investigated and only at the sides facing the flowing Pb some damage was observed. For Maxthal it was very difficult to detect erosion due to the non-perfect sample preparation. From the other tested material Noriloy with higher Cr shows the best behaviour. The worst behaviour was observed for T91 material in original status, at which up to 50µm deep erosion was detected. The SiSiC specimen got lost during the 100h. The free Si might have been dissolved at these conditions or the mounting might broke. The 2nd test run for 780h in Pb with expected oxygen content or a slightly higher one. At this condition and after 780h Noriloy and Norihard behave quite similar showing up to 30µm deep erosion at the front side facing the Pb. All the rest of both specimens stay entirely intact. The Maxthal was not tested and the SiSiC was attacked both by the Pb and the etching solution applied after the experiment. Again the non-surface alloyed front side of the GESA specimen, the non-modified T91 part was heavily attacked. Erosion up to 150µm deep at a distance of 1.5mm was observed. All the parts of the specimen that were surface alloyed using GESA do not show any attack.
ENEA designed and constructed a new loop that allows testing of real pump impellers at relevant conditions regarding temperature and oxygen content. The facility will work in pure molten lead and a prototypical mechanical pump is installed and will be tested for the corrosion at the impeller up to 500°C. The components and the process of the facility have been described in the respective deliverable in details. All experiments had to be postponed for the time after the LEADER project.
The scheduling of the experimental activity on the HELENA loop is to fill the facility with molten lead in December 2013 and to start the continuous test on the pump in January 2014. The test will be isothermal at 400°C with low oxygen content <10-8 % in weight; the oxygen content will be monitored continuously. The pump will be gradually driven to the reference mass flow rate 35 kg/s and this mass flow rate will be maintained for 1500 h, i.e. 2 months about. After this, the pump is stopped and the loop is drained. Then, the pump impeller is disassembled by the body and it is analyzed for corrosion.
Surface protection
Fuel claddings are operated up to 550°C. The use of 15-15Ti stabilized steel for the ALFRED cladding is doubtful or at least critical at this elevated temperature due to expected dissolution. The cladding material of the commercial ELFR, the f/m steel T91, will not suffer from dissolution, but oxidation becomes a severe problem. Therefore, surface protection barriers are required to ensure safe operation of Pb cooled fast reactors at the envisaged temperatures. Dense, and stable and slowly growing oxide scales that are formed in-situ are the preferential protection barrier. Alumina, which is employed for protection of stationary gas turbines, but at higher temperatures, is one option. Surface aluminizing can be used to enrich a base steel (either 316 or T91 type steel) with aluminium resulting in the potential to form alumina scales during use. One applicable method is the surface alloying of deposited coatings using pulsed electron beams. By that a surface graded material having the target Al concentration in the outer surface region can be manufactured [Fig. 24a]. Such scales showed their potential to protect steels up to 600°C even in LBE for more than 10,000h. To ensure the formation of such protective scales over the temperature range of 380 to 550°C an optimized FeCrAl composition was explored experimentally. An Al content of 6wt% with 16wt% Cr was effective to form thin Al-rich scales between 400 and 600°C in Pb [Fig. 24b].
Mechanical property degradation
Besides simple corrosion also mechanical properties like creep strength and wear resistance might be affected by the contact with the liquid metal coolant. It was shown that liquid Pb has a deteriorating effect on mechanical properties if it comes in direct contact with the steel. Oxide scales that prevent the direct interaction will, as long as they stay intact, mitigate any effect of the Pb. The creep rupture strength e.g. of f/m steel T91 is reduced remarkably due to the contact with the liquid metal. The influence of the coolant was negligible in case of surface aluminized T91. There, the thin alumina scale was able to catch the deformations of the steel without cracking. Detailed evaluation allowed to define a threshold stress of about 120 MPa, below which, mainly due to the reduced strain that does not result in oxide scale cracking, the influence of Pb becomes negligible [Fig. 25a]. 316L type steels in contrast does not show such pronounced effects at all.
Another interesting phenomena is fretting in Pb, which is a specific type of wear occurring at fuel claddings and heat exchanger tubes. Small amplitudes (<200μm) can lead to severe damage. Tests were done at harsh, accelerated conditions to evaluate the possible influence on the life time of the related components. Steel T91 and the 15-15Ti stabilized steel and surface aluminized T91 were explored. Fretting wear can lead to fuel clad penetration at the T91 and 15-15Ti steel of more than 10% of wall thickness depending on the relevant parameters like amplitude load and temperature already after 150 h at a frequency of 10 Hz (Fig. 25b). Both steels show in addition a strongly increased wear with increasing temperature. The austenitic steel exhibits at 450°C dissolution attack already after 600h, which was never observed even for more than 10,000h at 500°C in Pb. The surface aluminized T91 steel has the highest resistance against fretting wear. Post examination based on the construction of fretting maps allowed to predict the parameters that allow safe operation of fuel cladding regarding fretting damage. Loads above 75N and amplitudes below 15μm combined with surface aluminizing are safe parameters.
R&D NEEDS
Technology breakthroughs and innovations are needed for all Generation IV fast reactors. Innovative design and technology features are needed to achieve high safety and security standards, to minimise waste and enhance non-proliferation through advanced fuel cycles, as well as to improve economic competitiveness, especially with a high availability factor. In particular, structural materials and innovative fuels are needed. to sustain high fast neutron fluxes and high temperatures, as well as to comply with innovative reactor coolants. The development and qualification of new fuels also require significant R&D effort in terms of resources and time.
Key R&D topics
To achieve commercial availability of all Fast Reactors technologies by 2040, it is vital to continue and accelerate R&D on key topics, primarily: primary system design simplification; improved materials and innovative fuels; improved safety (safety standards at least comparable to Gen III standards); innovative heat exchangers and power conversion systems; advanced instrumentation and in-service inspection capabilities; reduction of the plant costs; partitioning and recycling processes; plant availability improvement. In particular the work in the area of partitioning and transmutation with innovative fuels, core performance and associated fuel cycle studies will lay the groundwork for assessment of future sustainable nuclear fuel cycle strategies. Whether and to what extent it should involve transmutation in a dedicated waste-burning Accelerator Driven System or in future Generation IV power plants needs to be assessed. Specifically focusing on LFR, the main challenges for their feasibility are: corrosion and erosion of structural materials; erosion in primary pump; seismic risk due to large mass of lead; in-service inspection of core support structures and possibility of replacing internal components; refueling at high temperature in lead; spent fuel management by remote handling; managing of the Steam Generator Tube rupture inside the primary system; prevention of flow blockage and mitigation of core consequences; radiation damage (at high temperatures); development of measuring devices.
Corrosion and erosion - The use of existing industrial materials for primary system is possible by limiting the core outlet temperature but new materials need to be used for special components (e.g. pump impellers). The present technological solutions are the use of low carbon grade austenitic steels for components at relatively low temperatures and low irradiation fluence, e.g. the reactor vessel, and the use of ferritic-martensitic steels for fuel cladding and core structures. Techniques for coating (i.e. aluminization) of steam generators tubes, fuel cladding and core structures operating at high temperature need further development. R&D needs come from the qualification of: i) an austenitic steel for the reactor vessel, ii) a lead corrosion and erosion resistant material for the in vessel structure, iii) a protective coating for fuel cladding and steam generator tubes, iv) special materials corrosion and erosion resistant for the impeller of the mechanical pumps.
Seismic risk - The mass of lead can be kept low by means of an innovative layout (e.g. Steam Generating Units and Primary Pumps installed in the reactor vessel). In such a configuration of the reactor vessel, seismic loads defined by the European Utility Requirements can be accommodated by means of 2D seismic isolators of the reactor building. To study seismic risk, a specific FP7 project (SILER) has started on 1st of October 2011.
Refueling - The fuel assemblies can be fitted with an extended stem to permit fuel handling, using a simple handling machine that operates in the cover gas at ambient temperature under full visibility. This eliminates in-vessel fuel transfer equipment which has never been designed or tested in lead. The problem of refueling can be considered solved by design, so no specific research activities are needed.
Purification and inspection - The use of molten lead as the coolant implies also research focused on the development and validation of a technique for lead purification before reactor vessel filling and with reactor in operation to prevent/control slag/aerosol formation, on the development and calibration of instrumentation operating in lead and under irradiation, on the development of techniques and instrumentations for in-service inspection of the steam generator tubes and the reactor vessel and on the development and validation of oxygen control techniques for regulating oxidation kinetics.
Fuel cycle - Significant R&D efforts should be focused on innovative fuels and fuel cycle (minor actinides bearing fuels, dense fuels such as nitride or metallic fuels). In the near term an essential goal is to confirm that ready-to-use technical solutions exist, so that fuel can be provided in timing with the ETPP operation. In the mid-term, it is necessary to confirm the possibility of using advanced MA (Minor Actinide)-bearing fuels and to confirm the possibility of achieving high fuel burn-ups. In the long term, it is important to confirm the potential for industrial deployment of advanced MA- bearing fuels and the possibility of using fuels that can withstand high temperatures to exploit the advantage of the high boiling point of lead.
Research and testing facilities - To meet the R&D challenges presented in the previous section, several facilities are required. Therefore the following actions are required: i) design, construction and operation of the necessary irradiation tools and devices to test materials and fuels; ii) design, construction and operation of the necessary fuel fabrication workshops, dedicated to uranium-plutonium fuels, and minor actinide bearing fuels; iii) design, construction, upgrading and operation of a consistent set of experimental facilities for component design, system development, code qualification and validation, that are essential to perform design and safety analyses of the demonstration program (e.g. hot cells, gas loops, liquid metal loops).
Experimental irradiation capacities - There is a clear need to update European irradiation facilities, given that existing facilities are close to end of life. New facilities are currently in construction or design phase in Europe: i) JHR, the Jules Horowitz Reactor (Cadarache, France) dedicated to materials testing for nuclear fission; its construction has started in March 2007, the start of operation is foreseen in 2014. ii) PALLAS (Petten, The Netherlands), a 'tank-in pool' reactor with a high neutron flux, a power of 45-80 MW and an annual degree of utilisation of more than 300 full-power days, equipped to meet the demand for both nuclear knowledge and services and the production of essential medical isotopes. PALLAS will substitute the existing HFR. According to the present plans, PALLAS could be in full operation in 2020. iii) MYRRHA (Mol, Belgium), a flexible fast neutron irradiation facility, able to test both lead coolant systems and accelerator-driven sub-critical systems (ADS) for transmutation. Beyond the availability of these irradiation capacities, it is necessary to develop new experimental devices taking into account cutting-edge progresses in modelling, instrumentation and modern safety standards. Europe has a worldwide leading position in this field and has to keep it through intra-European synergistic developments, to overcome shortage of resources.
Experimental facilities for reactor physics - Development of dedicated experimental facilities for reactor physics in fast reactor systems for component design, system development and code qualification and validation is necessary. GUINEVERE, already in operation, and ELECTRA, in its design phase, are such kind of machines.
Other experimental facilities - R&D demands further experimental facilities for: i) LFR material, coolant physics-chemistry and corrosion/erosion studies; ii) LFR safety experimental studies; iii) LFR studies of moving mechanisms, instrumentation, maintenance, in-service inspection and repair; iv) LFR studies in thermal hydraulics and heat transfer.
Fuel reprocessing capacity - Existing large research facilities (ATALANTE at CEA in France, ITU at JRC in Germany, the Central Laboratory at NNL in the United Kingdom) offer effective potentialities at lab-scale in the short term. They should be used in the future to develop suitable processes and to perform demonstration runs on samples of spent fuel, or on irradiated targets at up to pin-scale. For oxide fuel processing, minor actinide recovery processes, under development at lab-scale, mainly rely on well-known and industrially mature solvent extraction technologies. The important background coming from industrial plant feedback, or from the very important work carried out over past decades to design modern reprocessing plants, make extraction a well-mastered technology. Therefore, considering that there are no important issues for scaling-up hydro-metallurgical processes, the requirements in this field could be postponed. The present reprocessing capacity may represent a bottleneck in the future, due to the overlapping of the time schedules for the development of the SFR and LFR (MYRRHA and ALFRED).
Fuel manufacturing capacities - Beside existing facilities (ATALANTE, ITU, Central Laboratory facility), it is important to improve capability in the field of experimental fuel fabrication. A Prototype Core Facility (PCF) will be needed in a short time, for the production of the MOX fuel to be loaded into the core of the fast reactors prototypes, demonstrators and experimental. An industrial facility is under preliminary design in France by AREVA and CEA. A pin-scale facility will be also needed, able to provide in an efficient manner the (very diverse) experimental pins to be irradiated in experimental facilities during the early phases of the design of possible future fuels (MA-bearing fuel, other than oxide fuel). Such a facility could be located in existing hot labs. The goal is an efficient, modern and flexible tool with the capacity to produce from a few pellets up to a few pins per year, to address the many and diverse experimental needs expected from the R&D fuel research community.
Road Map
The main drawback facing the industrial deployment of a LFR fleet in Europe is the lack of operational experience gathered so far. Whilst a satisfactory technological readiness has been achieved in Russia in the past century, driven by military research, the aim at filling the technology gap in Europe requires the setting up of a complete R&D roadmap, in complete analogy with what has been done in France for the development of the SFR technology chain, whose readiness is almost proven.
The LFR development roadmap (see Fig. 26) towards the industrial deployment of a fleet of European LFRs, requires at first a small Demonstrator reactor (ALFRED) for proving the viability of reliable electricity production for LFR systems. A Prototype reactor is then envisaged, for testing the scaling laws at an intermediate step – according to a common approach focused both on plant size and representativeness of the target reference system – before moving to the First-Of-A-Kind (FOAK) representative of a commercial ELFR fleet.
Besides the reactors planned for the demonstration of the LFR technology chains, further facilities – either existing or under construction – are required (see Fig. 27) to provide the overall frame for LFR technology development, like all the existing and planned EU lead labs focused on thermal/hydraulics, materials development and corrosion testing, like the zero-power facility Guinevere, already operating in Mol, Belgium, like the irradiation facility MYRRHA, to be built and operated in Mol and like the European training facility ELECTRA, planned for realization in Sweden.
The road map of LFR is based on the progressive upscaling from the zero power facility GUINEVERE, in operation since 2010, to the commercial deployment of the first-of-a-kind European industrial size LFR (ELFR) by 2045. The intermediate steps planned are the European Technology Pilot Plant (ETPP), MYRRHA, to be in operation in 2023, the European LFR Technology Demonstrator Reactor (ETDR), ALFRED in 2025 and an LFR prototype of the industrial plant (PROLFR) in 2035.
The construction and operation of MYRRHA and ALFRED are the key steps towards industrial deployment, as they would enable important experience to be acquired for the design and operation of industrial power plants. They would not only establish a specific technology but also act as multipurpose research infrastructures. In particular ALFRED could be an experimental platform to address innovative fuel development, for testing specific minor actinide-bearing fuel to demonstrate and optimise transmutation and also for addressing the development of high temperature applications for industrial end-users. Key issues to be solved to allow the correct proceeding of the road map are the following: i) selection of materials for demonstrators and prototypes; ii) fuel development and qualification (MOX fuel, and in a later phase advanced minor actinide bearing fuel, and lead-fuel interaction); iii) Lead technology (lead purification/filtering techniques, oxygen & chemical control); iv) components development (safety & control rods, pumps, heat exchangers, in service inspection and repair technologies); v) development of models & tools (to study the nuclear/thermal-hydraulic feedback, the reactor stability, as well as the reactivity margin for not reaching prompt-critical conditions, response and resistance of structure to lead sloshing); and vi) conduction of large scale integral tests to characterise the behaviour of the main systems, especially for licensing procedures (key component performance and endurance demonstration, benchmarking of thermal-hydraulics in a rod bundle).
LFR Zero Power Facility (GUINEVERE) - GUINEVERE is a test fast reactor coupled with an accelerator, already in operation since 2010 and constructed by SCK•CEN in cooperation with European partners. It is a zero power facility designed for core design qualification and reduction of design uncertainties (critical mass, power distribution as well as reactivity coefficient).
LFR ETPP (MYRRHA) - MYRRHA is a research FR cooled by Lead- Bismuth Eutectic (LBE) with a power of 70 MWth when operated as a sub-critical mode (as an accelerator driven system) and a power of 100 MWth when operated in critical mode. It employs a MOX fueled core. The main mission of MYRRHA is to test both technologies of fuel and heavy liquid metals, and the endurance of materials, in-service inspection and repair, components and systems to control industrial risks (obtain reactivity feedback at power) for ALFRED, PROLFR and ELFR. Comparing the scope and specifications, the calendar and the current status of the project MYRRHA will fulfill the role of the LFR ETPP (with no need for electricity production). The realisation of MYRRHA will include several phases (2010-2024): Front End Engineering design, tendering and procurement, construction of component and civil engineering, on site assembly, commissioning, progressive start-up and finally full exploitation. The completion of the design phase is planned for 2015. The construction and commissioning is expected in 2022. In parallel, the support R&D program will provide during the 2010-2014 period the necessary answers to the remaining technical challenges. After 2014, the fuel qualification and to a lesser extent the material qualification will remain the main topics of the R&D support program. The completion of fuel and material qualification processes is expected in 2020.
LFR ETDR (ALFRED) - ALFRED is a pool type LFR of 125 MWe ~300 MWth. The primary coolant is pure lead in forced circulation. It can be cooled by natural circulation in emergency conditions. The core inlet temperature is 400 °C and the average core outlet temperature is 480 °C. The secondary coolant is water-superheated steam with the feed-water temperature of 335 °C, the steam temperature of 450°C and the pressure of 18 MPa. The secondary system efficiency is about 41%. The reactor vessel is in austenitic SS. The safety vessel is anchored to the reactor pit. The core barrel – cylindrical and integral with the core support grid and the hot collectors – is designed so as to be removed. The steam generators are bayonet type with double walls and they are integrated in the reactor vessel and removable. Primary pumps are mechanical, installed in the hot collector and removable as well. The fuel assembly is closed hexagonal, weighed down when primary pumps are off but forced in position by springs when primary pumps are on. The fuel is MOX (max Pu enrichment 30%). The design maximum discharged burn-up is 90-100 MWd/kg-HM. Fuel clad material is coated 15-15/Ti. Control/Shutdown Systems are two diverse and redundant systems of the same concept derived from CDT: the 1st system is made of pneumatically inserted absorber rods to be passively actuated by depressurization from the top of core; the 2nd system is made of buoyancy-driven absorbers rods passively inserted by buoyancy from the bottom of the core. As LFR ETDR, ALFRED will have to demonstrate the viability of the LFR technology for use in a future commercial power plant, construction and operation. It is expected to significantly reduce uncertainties in construction and licensing.
The demonstration of economical and reliable electricity generation is a key goal, thus the reactor core and primary coolant path should be as prototypical as possible, as well as the secondary circuit and the Balance of Plant (BoP). It will also allow global evaluations on the plant responses to the electrical grid needs. The objectives regarding neutronics are mainly related to the need of demonstrating the feasibility of an adiabatic core, which is a very crucial issue for an actual sustainability with respect to both natural resources and environmental impact. As to thermal–hydraulics, a key feature to be tested in ALFRED is the coolant ability to remove the decay heat in natural circulation under accidental conditions. Materials extensive testing will not be among the main goals of the demonstration project. Anyway, core materials (fuel, cladding and structures, coolant) should be the same of the ELFR ones so that ALFRED may be as much representative as possible. Within the LEADER project the design of the ALFRED is undergoing but the main suitable characteristic and design guidelines for the facility have been identified. A scaled demonstrator fully representative of the industrial size reactor is being designed, using components/technologies already available in the short term, to be able to proceed in the near future to a detailed design followed by the construction phase. The realisation of ALFRED will include several phases (2010-2025) see Fig. 28: the first step is identified in the set-up of an international consortium (2013) having already chosen a dedicated site for construction (Pitesti in Romania), in parallel several design activities should be running (basic design, siting and pre-licensing in the period 2013-2016 and detailed design and licensing in the period 2016-2019), finally construction of components and civil engineering, on site assembly and commissioning (2019-2025). The completion of the design phase is presently planned for 2019 to be able to take into account the feedback from design and experience from MYRRHA. In parallel, the support R&D program will provide during the 2011-2018 period the necessary answers to the remaining technical challenges. The ALFRED full operation is expected in 2025.
LFR prototype (PROLFR) - The PROLFR needs to be designed. The need of this step emerged in the LEADER project from the consideration that a direct power upscaling of factor 5 from ALFRED to ELFR was considered risky. Its power, of the order of 300-400 MWe, and its characteristics will be therefore between the ones of ALFRED and ELFR. PROLFR realisation is foreseen at the horizon of 2035.
First of a kind European industrial size LFR (ELFR) - Design and construction of the first of a kind 600 MWe industrial reactor is foreseen in 2035 to start with LFR Industrial plants building up at the horizon of 2040-2050.The reactor design parameters and characteristics are the same as in ALFRED with few exceptions. Electric power is 600 MW. Fuel is MOX with equilibrium MA content, maximum Pu enrichment will be lower than in ALFRED. The target value for maximum discharge burnup is 100 MWd/kg-HM. The target secondary system efficiency is above 40%.
LFR Education and Training facility (ELECTRA) - ELECTRA is a low power fast neutron reactor cooled by liquid lead whose design has been presented by KTH. Application of (Pu, Zr)N fuel permits to design a core with very small volume and fuel column height, resulting in highly negative coolant, fuel and structure temperature coefficients and very low channel pressure drop. Full design power of 0.5 MWth can be completely removed by natural circulation in the primary circuit, thus eliminating the need for pumps. Analysis of flow stability and performance under unprotected transients show that the suggested design is highly safe. ELECTRA's mission is in training and education. In particular it could be employed for training operators and staff of the other LFR in the roadmap, such as MYRRHA and ALFRED. In addition it could be employed for research on reactor dynamics. The main milestones for ELECTRA completion include the construction and operation of an electrical mockup at KTH in scale 1:1 in 2013-2014; the completion of the engineering design in 2016; the carrying out of irradiation experiments for licensing of cladding and fuel in 2012-2020 and the licensing for operation of ELECTRA in 2022.
ELFR CONFIGURATION AND SAFETY ANALYSES
Compared to ALFRED, the ELFR is a larger system (600 MWe), since it is intended primarily for central station electricity generation.
The design of most of its primary system components, such as reactor vessel, fuel assembly, inner vessel and core supports, is the same already described for ALFRED, with larger dimensions. Some other components are different, since, in order to simplify the design and licensing phases for an early construction of ALFRED, proven and qualified design solutions have been adopted for the demonstrator. In particular:
- Steam Generator: in the current ELFR configuration a once through SG with spiral tubes bundle is considered, whereas ALFRED SG is constituted by double wall straight bayonet tubes, allowing an easy coating and/or surface treatment to speed-up the construction;
- DHRs: in ELFR, the DHRs consistof two diverse systems: the Isolation Condenser and the DHR-2 system, constituted of dip coolers immersed in the reactor pool. In ALFRED, considering that the bayonet concept allows a continuous monitoring of the SG status, the diversity requirement has been relaxed: the two DHRs are both constituted by Isolation condensers;
A list of the main parameters concerning the design of the ALFRED and ELFR is provided in Table 2.
Safety analysis of the ELFR was performed testing all the “weak” points in the LFR design that were determined during the previous ELSY project. The most important DBC and DEC transients were repeatedly reanalyzed for the re-designed ELFR configuration. The full list of the analyzed transients for ELFR is asfollows:
- Protected loss of flow transient (PLOF),
- Unprotected loss of flow transients (ULOF),
- Unprotected loss of heat sink transient (ULOHS),
- Unprotected reactivity insertion transient (UTOP),
- Unprotected loos of flow and loss of heat sink transient (ULOF+ULOHS),
- Protected overcooling transient (OVC),
- Protected steam line break transient (SLB),
- Unprotected sub-assembly (SA) blockage transient, and
- Steam generator tube rupture (SGTR) accident.
The results of the safety analyses can be summarized as follows:
- Protected transients (PLOF, OVC and SLB): the automatic reactor shutdown activated by different scram signals is able to rapidly bring the ELFR plant to safe plant condition. The consequent isolation of the secondary circuits and start up of decay heat removal system is able to maintain the plant in safe condition in the medium and long term. In all transients, the potential of lead freezing in the coldest points of the primary system is reached after several hours within the transient, assuring sufficient grace time for manual, corrective operator action.
- Unprotected transients (ULOF; ULOHS and ULOF + ULOHS): due to the enhanced natural convection capability in the primary circuit, in case of ULOF the maximum temperatures reached in the primary system are low enough to assure the integrity of the clad and the vessel in the short term, providing sufficient grace time for corrective operator action. The main potential safety issue is the maximum reactor vessel wall temperature, which might exceed 700 °C within ~12 min. The integrity of the clad and the vessel seems not guaranteed in the medium/long term, because of the high temperatures reached in the primary system. An optimization of the neutronic core design, in order to reduce the positive coolant expansion reactivity feedback, could provide additional grace time.
- Reactivity insertion: for reactivity insertion of 200 pcm in 10 sec time interval at EOC conditions, peak fuel pin cladding survives and fuel melting is not observed, even in the center of the peak fuel pins (pellets). For reactivity insertion of 260 pcm in 10 sec time interval at EOC conditions, peak fuel pin cladding survives, however fuel melting should be expected in the center of the peak fuel pins (pellets). These transients envelope positive reactivity insertions of the design basis events such as fuel handling errors, control rods withdrawal or seismic core compaction.
- FA flow blockage: for blockages less than 75% blockage area, it is not expected any pin failures nor fuel melting, even under unprotected conditions. For blockage above 75%, peak power pins clad failure shall be expected, but fuel melting is not expected even for blockage over 97.5%. However, there is enough time (several hundreds seconds) to detect the flow blockage occurrence, by means of temperature measuring devices installed at each FA outlet.
- SGTR accident: several limiting mechanisms and potentially important effects have been analyzed and suggest that: (i) the initial pressure shock wave poses no likely threat to in-vessel structures, except very few adjacent heat-exchange tubes; (ii) the sloshing-related fluid motion is well bounded in a domain beyond the heat exchanger; and yet (iii) the steam/water entrainment is expected to be comparatively limited due to the verylarge difference of density between steam and lead. The potential gradual pressurization of the vessel after SGTR due to inflow of steam is limited by rupture disks to relief the resulting over-pressure. Moreover, a Venturi nozzle placed inside each spiral tube, mitigate the severity of SGTR interaction and reduce the potential effects on the entire reactor system. Anyway, a dedicated scaled facility should be foreseen to analyze in depth the SGTR phenomena further as part of the future R&D activities.
In order to assure prevention of freezing of the lead coolant at the coldest location of the primary loop, a tight and continuous operational control of the secondary coolant conditions is needed. Under certain adverse transient conditions it is conceivable that the primary lead HX outlet temperature (nominally 400°C, well above the freezing point of lead (327°C)) decreases to the feedwater inlet temperature 335°C, (only 8 °C margins to freezing). In addition, any malfunction in the FW temperature control could progressively bring the coolant to its freezing point. In general, the safety analysis performed for the lead-cooled ELFR design demonstrated the extremely robust nature of this plant design when compared to other similar plant designs, ascribable to the inherently, large thermal inertia of the lead-cooled primary system and optimization of safety relevant control, safety systems and components.
Potential Impact:
Electrical energy generation by means of nuclear power plants is at present an important fraction of the electric energy generation and continues to be an essential option for the future in Europe and worldwide.
The crucial issue of nuclear power, however, is that it should simultaneously be competitive, reliable, sustainable, safe, and perceived safe by the public and manage adequately the long-lived waste that it produces.
Modern water-cooled reactors are competitive and safe, but their energy generation is far from being sustainable, owing mainly to their ineffective use of the fuel and the production of long-lived highly radioactive waste, and to their high operating pressure which makes the high level of safety more difficult to achieve.
The LEADER project can have an important societal impact because of the emphasis on sustainability. The Public confidence in nuclear power depends to a large extent on the ability to demonstrate that it represents an unlimited resource of energy without creating critical issues for waste management. If LEADER confirms LFR as a reactor able to substantially reduce the long-term burden of a geological disposal and increase the extent of the present uranium resources, it will contribute to increase confidence and acceptance of nuclear power as an essential component of the energy mix within the EU.
A huge potential increase in the sustainability of nuclear energy will be achieved through demonstrating the technical, industrial and economic viability of Generation IV fast neutron reactors, thereby ensuring that nuclear energy can remain a long-term contributor to a low carbon economy. This demonstration program will play a key role by involving European industry and maintaining and developing European leadership in nuclear technologies worldwide. It will also make possible the further commercial deployment by the European industry of these technologies by 2040 and beyond. This is the prime goal for industry, which in the mean time will seek to maintain at least a 30% share of EU electricity from currently available reactors for the benefit of the European economy (the industrial needs for nuclear energy could be enhanced with an expansion towards co-generation of process heat for industrial applications when such markets develop). With the construction and operation of a LFR ETPP (MYRRHA) and ETDR (ALFRED), Europe will be in an excellent position to secure the development of a safe, sustainable and competitive fast spectrum technology. The program will allow to investigate and address the main technological issues that can then be implemented in the LFR prototype around 2020-2035. This LFR prototype, in turn, will pave the way for industrial deployment of LFR by 2050, and hence contribute significantly to the development of a sustainable and secure energy supply for Europe from the second half of this century onwards.
List of Websites:
http://www.leader-fp7.eu
Project coordinator: Alessandro Alemberti, Ansaldo Nucleare, Corso Perrone 25, 16152 Genova Italy (alessandro.alemberti@ann.ansaldo.it)
Introduction
The EU FP7 LEADER (Lead-cooled European Advanced Demonstration Reactor) Project had as main result the conceptual design of the demonstrator ALFRED (Advanced Lead Fast Reactor European Demonstrator). ALFRED is a 300 MWth pool system aimed at proving the viability of the European LFR technology for use in a future commercial power plant. ALFRED is designed close as possible to the reference ELFR, but proven and already available solutions are adopted to be able to construct it in a short term.
Description of the Nuclear System
The ALFRED primary system configuration is pool-type with all components removable. It presents a simple flow path of the primary coolant allowing an efficient natural circulation. The primary coolant leaving from the core moves upward through the Primary Pump (PP) to the vertical shaft and then, through the Steam Generator (SG) lead inlet holes, flows downwards on the shell out of the steam side and continuing moving downwards in the cold plenum reaches the inlet of the core. The volume between the primary coolant free levels and the reactor roof is filled by Argon. The Reactor Vessel (RV) is cylindrical with a torispherical bottom head and it is anchored to the reactor cavity from the top. A cone frustum, welded to the bottom head, has the function of bottom radial restraint of the Inner Vessel (IV). A steel liner covering the reactor pit constitutes the Safety vessel (SV). The volume between the RV and SV is like that, in case of RV leak, the primary coolant continues to cover the SG inlet maintaining the flow path. The IV has the main functions of core support and hot/cold plena separation. It is fixed to the cover by bolts and it is radially restrained at bottom. The Core Support plate is mechanically connected to the IV with pins for easy removal/replacement. The core, with a total power of 300 MWth, is constituted by 171 wrapped hexagonal Fuel Assemblies (FA), 12 Control Rods (CRs) and 4 Safety Rods (SRs), surrounded by 108 Dummy Elements. It utilizes hollow pellets of MOX fuel with maximum Plutonium enrichment of 30%. Each FA is long about 8 m and consists of 127 fuel pins fixed to the bottom of the wrapper and restrained sideways by spacer grids and has a Tungsten ballast to overcome buoyancy forces. The SG and PP are integrated into a single vertical unit. Eight SG/PP units are located in the annular space between the IV and the RV walls. SG is bayonet type with double walls based on a bayonet vertical tube with external safety tube and internal insulating layer (to assure superheated dry steam). The double walls gap is filled with helium and high thermal conductivities particles. The double walls concept prevents water/lead interaction in case of break of one tube wall and, moreover, a tube wall break event can be easily detected monitoring the Helium gap pressure.
Description of the Safety Concept
ALFRED is equipped with two diverse, redundant and separate shutdown systems: the first system, that has also the control function, is made by absorber rods passively inserted by buoyancy from the bottom of the core; the second one is made of absorbers rods passively inserted by pneumatic system (depressurization) from the top of core. The Decay Heat Removal (DHR) system consists of two passive, redundant and independent systems, each one composed of four Isolation Condenser systems (ICs) connected to four SGs secondary side. Three out of four ICs are sufficient to remove the decay heat power. Both systems are passive, with actively actuated valves, equipped with redundant and diverse locally stored energy sources. 2D seismic isolators are installed below the reactor building to cut the horizontal seismic loads.
Deployment Status and Planned Schedule
The successful demonstration of the LFR technology is reaffirmed by the effective operation of a First-Of-A-Kind of the industrial-scale ELFR. The achievement of this goal by 2040 requires the realization of the ALFRED demonstrator around 2025.
Project Context and Objectives:
Concerns over energy resource availability, climate change, air quality, and energy security suggest an important role for nuclear power in future energy supplies. While the current Generation II and III nuclear power plant designs provide an economically and publicly acceptable electricity supply in many markets, further advances in nuclear energy system design can broaden the opportunities for the use of nuclear energy. To explore these opportunities, worldwide governments, industries, and research centres started a wide-ranging discussion on the development of next-generation nuclear energy systems known as “Generation IV.”in Europe. The EC organized the Sustainable Nuclear Energy Technology Platform (SNETP) that, through its Strategic Research Agenda promoted the development of fast reactors with closed fuel cycle The Roadmap proposed by the European Sustainable Nuclear Industrial Initiative (ESNII), includes the lead-cooled fast reactor as an alternative technology to be developed in parallel with the sodium-cooled fast reactor.
The LFR system features a fast-neutron spectrum and a closed fuel cycle for efficient conversion of fertile uranium and management of actinides. A full actinide recycle fuel cycle with central or regional fuel cycle facilities is envisioned. The LFR can also be used as a burner of actinides from spent fuel by using inert matrix fuel.
During the 1970’s and 80’s, considerable experience was developed in Russia in the use of Lead-Bismuth Eutectic (LBE) for reactors dedicated to submarine propulsion and to develop new reactor designs based on both LBE (i.e. the SVBR reactor) and lead (i.e. the BREST reactor) as primary coolants. In the USA, in the past considerable effort was devoted to investigations of lead corrosion and materials performance issues as well as system design of the SSTAR reactor, while more recently the focus has included the development of the desired characteristics and design of a technology pilot plant or demonstrator reactor.
More recently, an extensive R&D program on Lead technology was initiated in Europe and is still ongoing. In particular, the R&D activities are focused on suitable materials, surfaces coatings (oxide layers, alluminization, tantalum, etc.) and lead chemistry (corrosions inhibitors).
The first step in the development of a Lead Cooled Critical Fast Reactor in Europe started in 2006, when EURATOM decided to fund ELSY (European Lead cooled SYstem). The ELSY project, coordinated by Ansaldo Nucleare, developed a very innovative pre-conceptual design of an industrial plant for electricity production able to close the fuel cycle.
After the end of ELSY project in February 2010, the LFR development continued with the LEADER project (Lead-cooled European Advanced DEmonstration Reactor), started in April 2010 in the frame of EC 7th FP.
The LEADER project is based on the EC call for a conceptual design of a Lead Fast Reactor and it takes into account the indications emerged from SNETP, as well as the main goals of ESNII. The LEADER project is strongly committed to the conceptual design of a scaled/demonstrator reactor (European Lead Fast Reactor Technology Demonstrator Reactor – ETDR) to be constructed in the relatively short term. The ETDR, during the LEADER project kick off meeting, was named ALFRED (Advanced Lead Fast Reactor European Demonstrator).
The focus of the first part of the LEADER project activities has been the resolution of the key issues emerged in the frame of the previous projects (e.g. core not enough constrained, floating of fuel assemblies, risk of core bypass caused by not fixed connection between pump ducts and Inner vessel, too complicated FA upper restrains), to reach a new configuration. With reference to this updated configuration of an industrial size Lead Fast Reactor (ELFR), the design of a scaled, fully representative (ALFRED) has been performed.
The objectives of the project activities for ALFRED are:
- to define the main suitable characteristic and design guidelines for the facility;
- to use components/technologies already available in the short term to be able to proceed in the near future to a detailed design followed by the construction phase;
- to evaluate safety aspects and perform a preliminary safety analysis;
- to minimise the cost of the demonstrator.
Moreover, the demonstrator shall confirm that the newly developed and adopted materials, both structural materials and innovative fuel, are able to sustain high fast neutron fluxes and high temperatures.
The LEADER project involved 17 partners from Industry, research organisations and universities. The total effort was 502 person-months over a period of 36 months, then extended to 42 months.
Project Results:
The LEADER project deals with the development of Lead Cooled Reactor technology through two main goals: the conceptual design of an industrial-size LFR (the so-called European LFR or ELFR) and the conceptual design of a scaled down facility, the demonstration reactor called ALFRED (Advanced Lead Fast Reactor European Demonstrator).
The project concentrated its efforts mainly on ALFRED demonstrator design, since its full operation is expected in 2025, whereas the design and construction of the first of a kind 600 MWe industrial reactor (ELFR) is foreseen for 2040.
The main results of the project are summarized in the following, in terms of mechanical, neutronic and thermalhydraulic design, safety analyses, materials tests and evaluation of R&D needs and roadmap for the deployment of both the design.
ALFRED reference configuration
The configuration of the primary system is pool-type. This concept permits to contain all the primary coolant within the Reactor Vessel, thus eliminating all problems related to out-of vessel circulation of the primary coolant (Figure 1).
The primary coolant is molten lead, which presents several favourable characteristics, such as:
a very high boiling point (1745°C) and very low Partial Pressure. For this reason, there is no need to pressurize the plant primary side, resulting in a easier design of the reactor vessel. Moreover, lead boiling scenario is practically impossible and this, together with proper Reactor vessel design, drastically reduces core voiding risk.
It is chemically inert with air and water: this allows the elimination of any intermediate loop and the installation of the Steam Generator Unit inside the Reactor Vessel.
favorable neutronic characteristics (it is a low moderating medium and has a low absorption cross-section). This gives the possibility to have a fast neutron flux even with large amount of coolant in the core; there is no need of compact Fuel Assemblies and it is possible to have relatively large spacing among the fuel rods, with reduced core pressure losses and thus enhanced natural circulation capability.
density is slightly higher than that of the oxide fuel, thus, in the LFR, fuel dispersion dominates over fuel compaction, reducing considerably the likelihood of the occurrence of severe re-criticality events in the case of core disruption. In fact, the dispersion of fuel and molten clad from core active zone is enhanced resulting in a self-reducing extension of core degradation.
As for all Gen IV advanced systems, the development of the technology is associated with research challenges. In the case of the LFR, these challenges include:
• molten lead interacts with structural materials, mainly with the mechanisms of corrosion at high-temperature and erosion. The provisions that can be adopted to improve the compatibility of lead and steels are:
- Operate at low temperature range (400 °C - 480°C) and maintain a controlled amount of oxygen dissolved in the coolant.
- Select material, such as Austenitic low-carbon steels (e.g. AISI 316L), ferritic-martensitic steels (e.g. T91), 15-15/Ti steel.
- Utilize surface coatings, e .g. by alluminisation of surfaces (with Fe-Cr-Al-Y) and surface treatment by electron beam (GESA treatment).
- Limit coolant flow velocity to a value that cause a negligible erosion (typically 2–3 m/s).
• lead has a high melting point (327.4°C) and this can result in a risk of coolant freezing, although this is not considered to be a safety issue but an investment protection issue. To decrease this risk, LFR is equipped with active systems to keep the lead molten during any normal operational conditions (e.g. during planned shutdown). In case of accidental emergency conditions, a sufficient grace time (more than 30 min) is available for the operator to reduce the heat removed by the Decay Heat Removal System (e.g. shutdown one DHR loop by manually closing one valve). However, promising activities are on going on the DHR design, in order to have the reduction of the removed heat by passive means.
• lead is opaque and this makes the in-service inspection and handling of fuel assemblies more difficult. For this reason, each component inside the Reactor vessel is removable and the fuel assemblies upper end extends beyond the lead free surface in the cover gas for refuelling without the need of in-vessel machines.
PRIMARY SYSTEM ARRANGEMENT
The Reactor assembly (Figure 2) presents a simple flow path of the primary coolant, with a Riser and a Downcomer. The heat source (the Core), located below the Riser, and the heat sink (the Steam Generators) at the top of the Downcomer, allow an efficient natural circulation of the coolant. The primary coolant moves upward through the pump impeller to the vertical shaft, then enters the SG through the lead inlet holes, flows downwards on the shell and exits the steam generator.
The free level of the hot pools inside the Steam Generator & Primary Pump units is higher than the free level inside the Inner Vessel, the different heads depending on the pressure losses across component parts of the primary circuit. The volume between the primary coolant free levels and the reactor roof is filled by a cover gas plenum.
The Reactor Vessel (RV) is cylindrical with a torospherical bottom head. It is anchored to the reactor cavity from the top, by means of a vessel support. The upper part is divided in two branches by a “Y” junction: the conical skirt that supports the whole weight and the cylindrical one, that supports the Reactor Cover. A cone frustum, welded to the bottom head, has the function of bottom radial restraint of Inner Vessel.
A steel layer covering the reactor pit, constitutes the Safety Vessel (SV). The dimensions of gap between the safety vessel and the reactor vessel is sufficient to the space for reactor vessel ISI tools. The safety vessel is cooled by the same system that cools the concrete of cavity walls. This system is inserted inside the concrete and is independent from the reactor cooling systems. This design solution mitigates the consequences of through-wall cracks with leakage of lead: any reactor vessel leakage is discharged into the Safety Vessel. The RV and the SV are arranged in such a manner that in case of a reactor vessel leak, the resulting primary coolant always covers the SG inlet and the lead flow path is indefinitely maintained.
The Inner Vessel (IV), Figure 3, has two main functions: Fuel Assemblies support and separation between hot plenum and cold plenum. It is fixed to the cover by bolts and is radially restrained at bottom. Lead flow is guided from the FAs outlet towards the PP inlet pipes by a toroidal half-ring. Moreover, the pipes that connect the hot zone with the inlet of PP are integrated in the Inner Vessel.
The cylindrical IV has a double wall shell: the outer thick wall has a structural function, while the inner thin wall follows the core section profile. The Core Lower grid is a box structure with two horizontal perforated plates connected by vertical plates. The plates holes are the housing of FAs foots and the plates distances must be sufficient to assure the verticality of FAs. The diagrid is mechanically connected to the inner vessel with pins (possible removal/replacement during reactor lifetime). The Core Upper grid is a box structure like the lower grid but more stiffer. It has the function to push down the FAs during the reactor operation. A series of preloaded disk springs press each FA on its lower housing. A hole is present for each disk to allow the passage of instrumentation (i.e. thermocouples).
CORE
The adopted core configuration is constituted by wrapped Hexagonal Fuel Assemblies. It utilizes MOX as fuel and uses hollow pellets, a low active height and a large fuel rod pitch (thanks to the lead physical properties) in order to improve the natural circulation. The total power is 300 MWth.
The core scheme is made of 171 Fuel Assemblies (FAs), 12 CR (Control Rods) and 4 SR (Safety Rods), surrounded by 108 Dummy Elements (ZrO2-Y2O3) shielding the Inner Vessel.
Each Fuel Assembly (Figure 4) is about 8 m long and consists of 127 fuel pins, fixed to the bottom of the wrapper and restrained sideways by grids (the wires can not be adopted due to the selected fuel pitch). Tungsten deadweight (Ballast) prevents buoyancy forces in lead. Upper elastic elements (cup springs) prevent lifting induced by hydrodynamic loads and accommodate axial thermal expansions. The FAs upper end extends beyond the lead free surface in the cover gas for easy inspection and handling. In this way it is possible to make the refuelling without the need of in-vessel refuelling machines.
ALFRED is equipped with two diverse, redundant and separate shutdown systems (adapted from the one that is under investigation in the frame of the CDT-MYRRHA project):
(1) CR (Control Rod) system, used for both normal control of the reactor (start-up, reactivity control during the fuel cycle and shutdown) and for SCRAM in case of emergency. The Control rods are extracted downward and rise up by buoyancy in case of SCRAM. The control mechanism pushes the assembly down with a ball screw, placed, with its motor and resolver atop the cover (at cold temperature (<70°C)), and protected from radiation by a shielding block. The actuator is coupled to a long rod by an electromagnet. When the coupling electromagnet is switch off (in case of SCRAM), the absorber assembly and the rod are free to rise up. Control rods use a 19 pins absorber bundle, cooled by the primary coolant flow. These pins are fitted with a gas plenum collecting the Helium and Tritium, produced by nuclear reaction of B10
(2) SR (Safety Rod) system, is the redundant and diversified complement to the control rods for SCRAM) only. The absorber bundle stays in the primary coolant. The rod is extracted upward and inserted downward against the buoyancy force. The absorber gets inserted by the actuation of a pneumatic system. In case of loss of this system, a tungsten ballast will force the absorber down by gravity in a slow insertion. For both systems the materials considered are B4C enriched in 10B at 90% as absorber, T91 for the guide tube, 15-15 Ti for the clad and ZrO2 (95%) - Y2O3 (5%) for the insulator and the reflector.
Both systems have been designed in order to be inserted also in case of core deformation (e.g. proper clearance and adaptive rollers to follow the deformation, improved mechanical resistance of guide tubes).
STEAM GENERATOR AND PRIMARY PUMPS UNIT
The steam generator and primary pump are integrated into a single vertical unit. Eight SG/PP units are located in the annular space between the cylindrical inner vessel and the reactor vessel wall. The primary pump is placed in the hot side of the steam generator, having its mechanical suction in the hot pool inside the inner vessel. The primary coolant moves upward through the pump impeller to the vertical shaft, then enters the SG through the lead inlet holes, flows downwards on the shell and exits the steam generator (Figure 5). The thermal stresses on SG tube plates, caused by the large temperature difference between the feedwater and the super heated steam, have been evaluated by means of the FEM Ansys showing large stress margins.
Each Steam Generator (Figure 6) consists of a bundle of 542 bayonet tubes immersed in the lead vessel pool for six metres of their length.
The bayonet tube is a vertical tube with external safety tube and internal insulating layer, composed by 4 concentric tubes (Figure 6): slave tube, inner tube, outer tube and outermost tube. The internal insulating layer (delimited by the slave tube) has been introduced to ensure the production of superheated dry steam: in fact, without an insulating system, the high ∆T (≈115°C) between the rising steam and the descending feedwater would promote steam condensation in the upper part of the steam generator. The gap between the outermost and the outer bayonet tube is filled with pressurized helium and high thermal conductivities particles (such as synthetic diamonds) to enhance the heat exchange capability. In case of an external tube break this arrangement guarantees that primary lead does not interact with the secondary water. Moreover, a tube break can be easily detected monitoring the Helium gap pressure.
The Primary Pump (PP) is located in the hot side of the SG because this strongly simplifies the internals components. However, the temperature difference between the cold and the hot side is not significant (80 °C). It is surrounded by the SG tube bundle and its housing is the Steam Generator casing. The Pump is fixed to the top of SG casing by a bolted joint. This allows an easy removal of the component.
The Primary Pump studies are in progress. Based on analyses performed during previous LFR projects (EUROTRANS, ELSY) an axial pump has been adopted. The PP impeller material is still an issue and test campaigns must be performed to select the proper one. The stainless steel (SS) cannot be used because the high speed achieved at the top of the impeller blades (about 9.8 m/s) induces a very fast material corrosion. Maxthal ceramic material has been proposed, but its reliability must be still demonstrated. An alternative solution can be a SS impeller with a coating, like Tantalum, already available in the chemical industry, to be qualified for nuclear application.
DECAY HEAT REMOVAL SYSTEM (DHR)
The Decay Heat Removal system (DHR) consists of two passive, redundant and independent systems, DHR1 and DHR 2, both composed of four Isolation Condenser systems (ICs) connected to four Steam Generators (SGs) secondary side (i.e. one IC for each SG). The system design considers the single failure criteria, since three out of four ICs are sufficient to remove the decay heat power.
The DHRs are dedicated safety systems, not used for normal operation. The separation is achieved through placing the two DHRs in physically different locations. A physical structural barrier or another means of protection will be placed between adjacent IC to ensure that failure of one of them could not harm another one. The diversity requirement has been relaxed (in any case the two DHR systems will be fabricated by diverse manufacturers) due to the high redundancy and considering that the SG tubes bayonet concept allows a continuous monitoring of the SG status.
Both systems are completely passive, with an active actuation (valves) that are however equipped with redundant and diverse energy sources (batteries or locally stored energy). Each DHR system must be ready to operate after the reactor trip in order to remove the decay heat power, in case of unavailability of the normal path (i.e. the by-pass to the Condenser). The actuation logic shall guarantee the actuation of DHR1 first, and of DHR2 only in case of failure of the first system and that the total number of isolation condensers called to operate never overcomes the four units, in order to avoid an excessive cooling of the primary coolant leading to fluid solidification.
Each of the four independent IC sub-systems consists of (Figure 7):
One heat exchanger (isolation condenser), constituted by a vertical tube bundle with an upper and lower horizontal header.
One water pool, where the isolation condenser is immersed; the amount of water contained in the pool is sufficient to guarantee 3 days of decay heat removal operation (after 3 days, the pool can be refilled with onsite ancillary water or, in case of extreme accident, by fire trucks).
One condensate isolation valve (to meet the single failure criteria this function shall be performed at least by two parallel valves).
Each isolation condenser is connected to a Steam Generator: the upper header of the isolation condenser is connected to the main steam line and the lower header of the Isolation Condenser is connected to the main feed water line. In normal operation the isolation valve below the condenser is closed, the condenser is full of water and no heat exchange takes place. To put in operation the component, the feed water line and steam line must be isolated and the condenser isolation valve opens injecting water into the Steam Generator. The steam produced in the steam generator now can reach the Isolation condenser starting the steam condensation process.
ALFRED CORE DESIGN
A comprehensive core design strategy has been adopted, approaching systemically, from a thermal-hydraulic, thermo-mechanic and neutronic point of view, all the aspects of the reactor core impacting on the safety and the robustness of the plant. As a matter of fact, such an approach allows for the a priori definition of a core which respects, at least in first approximation, all the technological constraints and fulfils all the performance objectives by design. Then the results are checked a posteriori with more detailed code calculations.
According to this, the ALFRED fuel pin, the fuel assembly, the control and safety systems, the reflector and the core have been designed. In particular, preliminary analyses have been performed to check whether the investigated fuel pin and the coolant sub-channel allow for an acceptable thermic of the pin and a sufficient natural circulation in ULOF conditions without exceeding some safety limits, mainly linked to the clad temperature.
Fuel pin and coolant sub-channel design
The first phase of the core design process foresees a preliminary thermal/hydraulic analysis to check the respect of the design limits to what concerns both the fuel and cladding maximum temperatures in the hottest channel. Since the latter – together with the coolant inlet and outlet temperatures – are the same as for the ELFR, the design of the fuel pin has been borrowed by this system, together with the maximum linear power rating (≈ 340 W/cm). The same assumption holds considering the target peak BU (≈ 100 MWd/kg), so that the hollowed fuel concept of the ELFR has been kept to accommodate the irradiation-induced swelling without incurring in Pellet-Cladding Mechanical Interaction (PCMI). The resulting (radial) fuel pin design is shown in Figure 8.
The design of the sub-channel cannot omit preliminary safety analyses to verify the respect of the cladding integrity for the desired grace time (30 minutes before SCRAM is manually actuated) in case of accident. To account for this constraint, the natural circulation regime during a ULOF transient has to be evaluated. Having this in mind, the sub-channel has to be designed so as to offer a hydraulic diameter the pressure losses associated to which can be overtaken with a small prevalence, attained through acceptable coolant temperatures.
On the other hand, the wider the pitch, the lower the fuel volume fraction in the elementary cell, to the detriment of the reactivity of the system. According to this, two values of the nominal coolant flow velocity have been therefore assumed for the present study, 1.4 and 1.5 m/s, supposing these as reasonable for achieving a sufficient natural circulation without affecting excessively the core reactivity. Furthermore the choice of the preliminary core configuration has to cope also with the constraint imposed for the inner vessel radius.
At last, two active heights (60 and 90 cm) have been considered in the present study not only to encourage the natural circulation by increasing the chimney height (since the shorter fuel and plenum length), but also to allow for keeping in the width of the lattice pitch, gaining in reactivity. The natural circulation regime in case of ULOF has been then evaluated by means of the SIMMER code for the four different configurations resulting by combining the two active height options with the two coolant velocities. According to the results of this investigation, the configuration corresponding to a coolant velocity of 1.4 m/s and 60 cm active core height has been chosen as reference, to which corresponds a value of the maximum cladding temperature during ULOF below 740 °C.
For the final dimensioning of the coolant sub-channel, the radial and axial power form factors had to be evaluated. Concerning the axial form factor, a value of 1.15 has been borrowed from the analyses performed on a similar system. The radial power form factor has been set equal to 1.2 assuming a proper power/FA distribution flattening is achievable through suitable Pu enrichment zoning of the core. According to these assumptions – which automatically become recommendations for the neutronic design of the core – the corresponding pitch for the fuel pins lattice is set to 13.86 mm.
Core design
Starting from the maximum linear power rating and the axial and radial power form factors, the average linear power rating is computed and used to define the number of fuel pins to be arranged to sum the aimed 300 MWth. The corresponding number of FAs (171) can be arranged to form a cylindrical core, together with 16 positions for Control and Safety Rods (CRs and SRs respectively), surrounded by two rings of dummy elements shielding the inner vessel within the imposed radius. Extensive neutronic calculations – by means of both MCNPX and ERANOScodes with the JEFF 3.1 cross-sections library– followed, to define the Pu enrichments and zoning, along with the positioning of control and safety elements, ensuring the aimed criticality level, a power/FA distribution flattening complying with the assumptions made for the design of the fuel pin and coolant sub-channel, and anti-reactivity worth for control and safety systems ensuring the respect of safety constraints. The underlying hypothesis is a fuel management according to a 5-batches cycle without reshuffling, with 5 y FA residence time in pile. The latter comes directly from the assumption on the maximum linear power rating and the chosen pellet geometry, to target the aimed peak burn-up of 100 MWd/kg.
The requirements for the control/safety systems are:
- a sufficient worth of the CRs to compensate the reactivity swing during the 1-year sub-cycle;
- the worth of each single CR when inserted to compensate the criticality swing not higher than 1 $;
- 5000 pcm for the refuelling status (keff = 0.95);
o about 2500 pcm for the SCRAM. This has been set according to:
o some 500 pcm for the transition from hot to cold state;
o 1000 pcm of sub-criticality, keff = 0.99 (value to be verified by kinetics analyses);
o pcm of uncertainty (assumed).
The strategy to accomplish these requirements foresees:
- two redundant systems for SCRAM (2500 pcm each), one of which also devoted to the compensation of the reactivity swing during the cycle;
- the simultaneous use of both CR and SR systems during refuelling to reach keff = 0.95.
The resulting core configuration is characterized by 57 FAs in the inner region, with a Pu enrichment of 21.7 wt.%, and 114 FAs in the outer region, with a Pu enrichment of 27.8 wt.%. 12 CRs are positioned within the outer region with the twofold goal of lowering the flux insisting on the most enriched FAs, while compensating the over-criticality along the cycle through a progressive withdrawal from their maximum-inserted position at BoC. 4 SRs are also positioned close to the core centre to minimize the shadow effect with the CRs. The ALFRED core, and the power/FA distribution at Beginning and End of Cycle (BoC and EoC) are shown in Figure 9.
For this configuration, a criticality swing of about 2600 pcm is found during the 1 year cycle. The worth of the control and safety systems is found to be –8500 pcm and –3300 pcm, respectively. Even in case of one SR stuck in withdrawn position, the anti-reactivity provided by the safety system is still about 2600 pcm. These values are found sufficient for the achievement – in every condition – of the required anti-reactivity margins for control, SCRAM and refuelling. To compensate the reactivity swing during the sub-cycle, a partial insertion of the CRs by some 17 cm is sufficient. This is the main responsible for the radial and axial distortion of the power distribution in the core. Thanks to the high number of CRs, the limit for the maximum worth of 1 CR at BoC is also respected, being 220 pcm the maximum accidental reactivity insertion against 335 pcm of delayed neutrons fraction.
For the most powerful FAs in the two zones the detailed power/pin distribution has been computed by means of MCNPX. A complete thermal analysis of the fuel element followed, in order to verify the respect of the maximum fuel and cladding temperatures.
This analysis has been performed by means of the ANTEO-LFR sub-channel code originally developed in ENEA for fast reactors fuel assembly thermal/hydraulic studies, and adapted to lead coolant.
The distribution of the coolant temperature at core outlet among the sub-channels of both these FAs is shown in Figure 10. For the FA in the outer region, since the more unfavourable power distribution within the FA itself, a complete analysis of the fuel pins followed, showing the respect of the design limits even for the hottest pins (Figure 11).
RESULTS OF THE SAFETY ANALYSES TRANSIENTS PERFORMED FOR ALFRED
Safety of ALFRED is extensively based on the use of the defense in depth criteria, enhanced by the use of passive safety systems (actively actuated through local stored energy sources, always available, and fully passively actuated). Safety features of the LFR system have been designed since the beginning of LFR related activities to face challenging plant conditions and events, taking into account the forgiving and advantageous physical characteristics of the coolant. As an example, there is no need for off-site or emergency AC electrical power supply to manage design basis accident conditions, as the only manual action needed is the supply of additional water to maintain the water level in the decay heat removal (DHR) pools that are sized to guarantee at least three days of unassisted, fully passive operation. These pools can be easily re-supplied with water in the subsequent days.
Design Basis Conditions Transients
Within the LEADER project all of the most important DBC transient initiators were analyzed for the ALFRED reactor. A selected set of the analyzed DBC transients are briefly presented hereafter
Spurious withdrawal of the most reactive control rod (PTOP)
For this simulation it was assumed that: 1) at t = 0 sec, the event of spurious withdrawal of the most reactive control rod (CR worth: 250 pcm) starts; 2) CR withdrawal velocity by mechanical means is limited to 0.1 mm/s, or 0.15 pcm/s; 3) as all CRs at EOC (End-Of-Cycle) condition are withdrawn from the active core region, the BOC (Beginning-Of-Cycle) condition is used for this simulation as all CRs are inserted into the active core region by 16 cm, 4) spurious withdrawal of the most reactive CR proceeds, and at t = 524 sec, a reactor scram signal is being generated based on high neutron flux > 120 % nominal, but this signal is being neglected; 5) at t = 548 sec, the second reactor scram signal is generated based on peak power FA temperature difference being > 1.2; 6) based on this reactor scram signal, the reactor is being shutdown at t = 549 sec into the transient; 7) main steam lines and main feedwater lines are then closed in ~2 sec time interval, while condensate isolation valves of four ICs (Isolation Condensers) of the DHR-1 system fully open in ~ 20 sec, removing maximal 7 MW of heat from the primary cooling circuit.
As can be observed in Fig. 12, CR withdrawal out of the active core region continues for 549 sec until the reactor trips. During this time interval a total positive reactivity of ~83 pcm is being inserted into the core. Insertion of ~83 pcm in 549 sec time interval at BOC conditions leads to an increase in reactor power ~ 1.2 nominal. The maximum fuel and clad temperatures increase from 2028°C and 516°C to 2313 °C and 555 °C respectively. Minimum clad failure time of the peak pin during the transient always remains above 1E+11 sec and thus no fuel pin failure (creep rupture) is expected.
As related to the above study, one can observe that for reactivity insertion of ~83 pcm in 549 sec time interval of the most reactive control rod (250 pcm total worth at BOC, HFP: Hot Full Power conditions), the ALFRED peak fuel pin cladding accommodates this transient; in addition no fuel melting is expected.
Loss of AC power (PLOOP)
This transient is initiated by assuming the total loss of offsite power supply (station blackout). As a consequence, the forced circulation in the primary system is lost (pump coastdown) with simultaneous turbine and feedwater trip on the secondary side and reactor scram. The secondary circuits are isolated by the main steam isolation valve closure, and the decay heat removal by the DHR-1 system (four IC loops are supposed in service) is promptly activated through the opening of the triggering valve positioned below the isolation condenser.
In spite of the very low primary pump inertia (speed halving time < 1 s), there is a smooth core flow rate reduction in the initial phase of the transient (see Fig. 13), due to lead free levels equalization in the primary system. As a result, the initial clad peak temperature increase is not significant (max T-clad = 564 °C) from the safety point of view.
The core decay power is first removed by water evaporation inside the MHX and then by steam condensation inside the isolation condenser immersed in the boiling water pool, with final decay power release to the atmosphere. The whole system functions passively in natural circulation; the pool water storage is guaranteed for at least three days of DHR-1 operation. Additional water can be easily supplied manually, thereby assuring practically unlimited operation.
The Fig. 14 shows that the core decay power is efficiently removed by the DHR-1 during the whole transient phase. The MHX power in excess to IC power in the initial phase is removed by steam release through the safety relief valves of the over-pressurized secondary circuits. After about 15 minutes, the DHR-1 power removed (about 7 MW for 4 IC loops) surpasses the core decay heat and the lead temperatures in the primary circuit start to decrease (see Fig. 14). After 3 hours the minimum lead temperature is reached at the MHX outlet (350 °C), still sufficiently distanced from the lead solidification point (327 °C). Manual or automatic control of the IC unit operation is needed in the long term to avoid potential lead freezing. Passive control of the operation of the IC units is under investigation in order to exclude the needs for manual intervention in the control systems.
Based on the above results one can conclude that, in case of total loss of offsite power supply, the core decay heat can be removed safely indefinitely in a passive manner without the need of active energy sources.
Protected partial flow blockage in the hottest fuel assembly
Transient analysis of the unprotected partial blockage of the hottest SA w/o radial heat transfer for ALFRED can be summarized as follows:
(1) ALFRED will not experience any fuel pin failure for flow blockage less than 75%, even under unprotected conditions;
(2) For flow blockages above 75%, clad failures (creep rupture) should be expected;
(3) Fuel melting is not expected to be a safety issue for the ALFRED as fuel melting temperatures are not reached, even in 97.5% SA flow blockage case.
Using these results, the influence of delaying reactor trip by 1, 2, 3, 5 and 10 sec assuming a 70% SA flow blockage was analysed. Blockage was assumed activated at t = 1 sec transient time, followed by reactor trip (signal based on the peak SA temperature difference value exceeding 1.2 nominal) assuming different time delays in order to study the impact of the trip signal delay on the maximum fuel and clad temperatures. Reactor trip thus occurring at time t = 4, 5, 6, 8 and 13 sec into the transient. Main steam lines and main feedwater lines are being closed in 20 sec time period, at the same time opening the condensate isolation valves of four ICs of the DHR-1 system, removing maximal 7 MW of heat from the primary cooling circuit.
Simulations showed that the maximum clad and coolant temperatures are achieved when the reactor trip is being delayed for 10 sec. The maximum increase of the peak clad temperature in this case is 172 oC for BOC (from initially 516 oC) conditions, and 185 oC for EOC (from initially 514 oC) conditions. The maximum increase of the peak coolant core outlet temperature is 184 oC for BOC conditions (from the initial value of 487 oC) and 195 oC for EOC conditions (from initially 487 oC). Later, following reactor trip, all core and primary cooling circuit temperatures decrease.
The 70% SA flow blockage case for 10 sec reactor trip delay at EOC is shown below in Fig. 15.
The maximum clad temperature of the peak pin increases from 514 oC up to 699 oC (Fig. 15) at transient time t ~ 14 sec. At the same time point minimum clad failure time of the peak pin drops down to 5E+6 sec (Fig. 5), later recovering to 5E+16 sec, indicating that the clad material is not challanged. The fuel temperature of the peak pin is the highest at the beginning of the transient (2064 oC), decreases after reactor trip. All other temperatures around the primary cooling circuit also decrease following reactor trip.
The performed analysis demonstrated that the clad of the peak pins of the ALFRED reactor has a very large margin to clad failure (rupture) during the simulated peak power SA for the 70% flow blockage transient. Reactor safety is ensured by the reactor protection system, shutting down the reactor, in conjunction with the DHRS, efficiently removing the decay heat from the reactor primary cooling circuit. Even a 10s delay in reactor trip does not challenge the integrity of cladding material in ALFRED.
Reduction of FW (feed water) temperature from 335 oC to 300 oC
The over-cooling (OVC) transient was initiated by the reduction of feedwater temperature from 335 °C to 300 °C within 1 second. When feedwater temperature was lower than 327 °C, the SCRAM signal was triggered to shutdown the reactor with 1 second delay. Later, main steam lines and main feedwater lines in SG system were closed in 2 seconds. At the same time, condensate isolation valves are fully opened, leading to the functioning of DHR systems, that are able to provide 7 MW heat removal capability. The opening process of condensate isolation valves takes 20 seconds. During the whole transient, safety valves maintain the pressure in isolation condensers under 195 bar and temperature of liquid water leaving isolation condensers is kept at 300 °C.
Simulation results of the OVC transient are presented in both short-term and long-term time scales.
When condensate isolation valves start to open, liquid water stored in isolation condensers will flow into steam lines and evaporate instantly, leading to a significant increase of MHX power up to 140% nominal. Thereafter, MHX power will continuously decrease with the core power, but it is still sufficient to reduce the temperature difference between the core outlet and core inlet, until only negligible temperature difference between core inlet and outlet remains, as it is shown in Fig. 16.
As a design basis, the rated power of one DHR system is 7 MW, which will be higher than core power after 900 seconds decay. Hence, primary coolant temperature will decrease slowly as it is shown in Fig. 17. After 14000 seconds, the lowest coolant temperature in the primary circuit (at the MHX outlet) will be lower than the freezing point of liquid lead (~ 327 °C), lead at that location will freeze, unless manual intervention will reduce the DHRS heat removal capability by shutting down appropriate sub-systems of the DHRS.
Increase of FW flowrate by 20 %
This transient is initiated by a control malfunction on the secondary side leading to a an increase of the feedwater flowrate (+20% in 25 s) at the inlet of all MHXs.
As a consequence of the feedwater flowrate variation, the power removal through the MHX increases with corresponding reduction of the lead temperature on the primary side at the MHX outlet as shown in Fig. 18. The maximum MHX power increase is about 10% nominal. The temperature perturbation from the MHX outlet propagates through the core and then back to the MHX inlet in about one minute, starting the decrease of the MHX removed power. The core temperature decrease leads to a total positive reactivity insertion (see Fig. 19), mainly driven by radial core, coolant and control rod drive expansion, with consequent progressive core power increase that comes to equilibrium with the MHX power in about 5 minutes. After this time a new steady-state conditions is reached in both primary and secondary systems, without exceeding any reactor trip threshold set-point. The maximum temperature decrease calculated at the core inlet is of 13 °C, while the core power and the core ΔT increase is below 6%. Furthermore, the fuel peak temperature increase is not very significant (about 70 °C, see Fig. 19). From the performed analysis one can see that this transient has no safety implications for the ALFRED reactor.
Loss of all primary pumps (PLOF)
This transient is initiated by the loss (with subsequent coastdown) of all primary pumps at t = 0 s. The reactor scram on low primary pump speed signal should be effective just after 1 s, because of the low pump speed halving time (< 1 s), but for conservative analysis this first scram signal is neglected. Therefore, the automatic reactor shutdown is actuated at t = 3 s, with one 1 s delay after the second scram signal, based on the ΔT increase through each individual FA (ΔT > 1.2 nominal value), is detected (t = 2 s). Following the reactor scram the secondary circuits are promptly isolated (main steam isolation valve closure and feedwater trip) and the DHR-1 system is actuated to remove the core decay power. For conservative analysis against maximum core temperature reached, only three out of four IC units of the DHR-1 system are supposed to be in service.
After the loss of primary pumps, the transition from forced to natural circulation in the primary circuit is similar to the one already observed in sub-section above for the loss of AC power transient. However, since in this case the reactor scram is delayed, the peak clad temperature increases reaching a maximum at t = 4 s slightly below 650 °C (see Fig. 10). After reactor scram the lead and clad temperatures at the core outlet progressively reduces, according to the residual core flowrate and the core decay power level.
Despite the reduced availability of the DHR-1 system (3 IC unit), which is now able to remove a total of about 5 MW power during the transient, the DHR-1 heat removal surpasses the core decay heat after about one hour, and then the primary temperature tends to reduce (see Fig. 20). The core temperature increase at the core outlet after the initial transient phase is not significant from the safety point of view. In fact, the maximum core temperature rises up to about 475 °C, that is below the normal operating value.
The risk of lead freezing at the MHX outlet after DHR-1 start-up is excluded, since the minimum lead temperature (350 °C) is far enough from the freezing point of 327 °C. After 3 hours the lead temperature at the MHX outlet is still above the lead solidification point. As in case of loss of AC power, manual or automatic control of the IC unit operation is needed in the long term to avoid potential lead freezing.
In spite of the conservative assumptions regarding the failure of the first scram signal and of one IC unit of the DHR-1 system, one can see that, also in case of total loss of forced circulation in the primary circuit, the protection and safety systems are able to bring and maintain the plant in safe conditions in the short and long term.
Conclusions
The above results show, that ALFRED has very large grace times (~hours). Calculated peak pin temperatures are not exceeding 700 °C, thus no clad failures (creep rupture) are expected, even during the loss of AC power (PLOOP) and loss of all primary pumps (PLOF) transients. This is mainly due to the large thermal inertia of the Pb-cooled primary system and the relatively high Pb natural convection core mass flow rate observed at the initial stage of the transient. In the long-term, in case of PLOOP or PLOF transients, following reactor trip, core decay heat can be passively and safely removed indefinitely by the DHR-1 system without the need of active energy sources.
Large margin to clad failure (rupture) for the clad of the peak pins of the ALFRED reactor was confirmed also during the simulated peak power SA 70% flow blockage transient. Reactor safety in this case is ensured by the reactor protection system, shutting down the reactor, as well as by DHRS, efficiently removing the decay heat from the reactor primary cooling circuit. Reactor trip delay even by 10 sec does not lead to clad failure for the ALFRED reactor.
During the protected FW temperature decrease to 300 oC transient, apart from a safety issue of Pb-freezing several hours into the transient at the outlet of the MHX for the ALFRED reactor, no other safety related issues are observed. During this time period the operator has sufficient time to deactivate sub-systems of the DHR-1 system, in order to prevent Pb-coolant freezing at the outlet of the MHX. In case of FW temperature reduction to 300 oC, it is actually recommended not to trip the reactor, thereby assuring that the primary coolant will not freeze at the MHX outlet.
As a general conclusion, no relevant, immediate safety issues have been identified during the performance of the simulated transients aside of potential Pb-freezing several hours into the transient in case of all protected DBC transients. Reactor safety is assured by the reactor protection system, shutting down the reactor, in conjunction with a heat removal system (DHRS), removing tightly controlled amounts of decay heat from the reactor primary cooling circuit to assure prevention of freezing of the Pb-coolant at any location of the primary cooling circuit.
As related to the DBC transients, all selected transients examined proved that the ALFRED plant can accommodate a rather wide range of accidental events. The ALFRED plant has proved to be able to enter a safe shutdown phase after every DBC accident analyzed.
The analysis indicated, that ALFRED is a very forgiving plant design, and there is an extended time margin (grace time) of several hours for possible manual operator intervention even under worst accidental conditions (potential of Pb-freezing in the long term, in case of uncontrolled decay heat removal by the DHRS).
Design Extended Conditions Transients
The DEC events that are representative for the safety analysis of ALFRED have been identified by means of common safety approach adopted for liquid metal fast reactors and on the basis of engineering judgment, taking into account the specific features of the ALFRED design. These events of very low frequency are characterized by the failure of reactor scram (so called unprotected accidents) involving:
- Unprotected reactivity insertion transient (UTOP),
- Unprotected loss of flow transient (ULOF),
- Unprotected loss of heat sink transient (ULOHS),
- Unprotected loss of flow and heat sink transient (ULOF+ULOHS),
- Unprotected partial FA blockage,
or by the simultaneous occurrence of multiple failures in some protected transients:
- Loss of all secondary circuits with total unavailability of DHR system,
- Loss of all primary pumps with reduction of feedwater temperature (loss of one preheater),
- Loss of all primary pumps with increase of feedwater flowrate by 20%.
The analysis of the representative DEC transients for ALFRED has highlighted very good intrinsic safety features of the reactor design. In particular, the results of the unprotected transients underline that the reactor can be maintained in a safe and controlled state even under the unlikely accident conditions, which include the failure of the reactor scram, thanks to:
- Establishment of enhanced and stable naturalconvection in the primary cooling circuit following the loss of the primary pumps,
- Dominant negative reactivity feedback effects obtained by optimizing the neutronic core design,
- Large thermal inertia of the primary cooling system,
- Passive and efficient operation of the DHR system for decay heat removal.
As a general conclusion, no relevant safety issues have been identified for ALFRED in case of representative DEC events. In particular, the ULOF transient can be accommodated without the need of corrective operator actions. Finally, the analysis of DEC transients for the lead-cooled ALFRED design has demonstrated the extremely forgiving nature of this plant when compared to other similar plant designs.
MATERIAL SELECTION
The operating temperatures of the different reactor components are identical for both of the discussed nuclear reactors, ALFRED and the ELFR (Table 1). Therefore, most of the materials selected to manufacture the components like reactor and inner vessel and refuelling equipment are identical for both concepts. One major difference between ALFRED and the ELFR is the expected maximum irradiation damage of the fuel assemblies. The ALFRED design foresees 100 dpa, what allows the use of the 15-15Ti stabilized austentic steel also employed in the French Phenix reactor. This steel is already licensed for application in nuclear reactors, which is an additional important aspect for material selection for ALFRED. The fuel assembly of the ELFR needs to withstand a maximum radiation damage of 200 dpa that is too high for the best available 15-15Ti stabilized steel. The class of steels that can tolerate such high radiation damage are the ferritic/martensitic (f/m) steels, like T91 or specific ODS types.
Besides improved irradiation stability, the heat conductivity of the f/m steels is almost a factor of two higher compared to austenitic steels like 316L and 15-15Ti stabilized. Therefore, T91 is also discussed as structural material for heat exchangers and steam generator (SG) components. The thermally most loaded components like fuel claddings and steam generators or heat exchangers, where coolant temperatures of 550°C can be reached, with cladding wall temperatures even 20K higher, might require surface aluminized steels to be operated. For most components, beside the pump impeller, the maximum Pb velocity of 2m/s does not require specific attention.
Compatibility of selected materials with Pb
Compatibility of materials with Pb is mainly driven by the temperature dependant solubility of the alloying elements in the liquid Pb. Ni, Mn, Al and Si have substantial solubility of up to some wt% at 550°C. The main elements of steel, Fe and Cr, are less soluble but still some 10-5wt% can not be neglected. Practically insoluble are refractory alloys like W, Mo and Ta and most ceramics. Therefore Ni containing metals like austenitic steels suffer more from dissolution attack than Ni free f/m steels like T91. To mitigate dissolution any direct contact between the steel and the liquid Pb should be avoided. The most appropriate method is the in-situ oxidation by dissolved oxygen. The oxygen concentration range is determined by the operating temperature range of the reactor. The lower operating temperature of 380°C, to consider some safety margin 325°C (Tmelt of Pb), ascertains the upper limit of 10-4wt% oxygen to avoid PbO formation and any related coolant channel blockage. The highest achievable temperature at normal operation (550°C fuel clad) determines the lower limit of dissolved oxygen of 10-7wt% required to prevent or at least mitigate dissolution attack. Such formed oxide scales act as diffusion barrier and can prevent the dissolution. However, especially in case of f/m steels the growth of oxide scales can, at least after longer exposure times, result in severe reduction of the heat conductivity.
Austenitic steels – 316L, 15-15Ti (1.4970):
The operating range for the austenitic steel 316L is, as depicted in table 1, between 380 and 480 °C. The 15-15Ti that should be employed as fuel assembly material can reach temperatures up to 550°C or even higher due to reduced heat transfer properties by the formation and growth of oxide scales. The upper limits are due to limitations by the material compatibility and might be altered if suitable solutions are available. In previous national and international projects Ni containing austenitic steels were mainly tested in Pb Bi eutectic (LBE). Due to the higher solubility of alloying elements in LBE than in Pb, a direct transfer of the attained knowledge is not simple. The general mechanism, the dissolution of Ni accompanied by penetration of Pb into the steel matrix, is equivalent for both coolants. However, the rate of dissolution and the resulting temperature and oxygen concentration limits might differ. Below 430°C, the maximum temperature of the reactor vessel, even in LBE no dissolution attack was observed as long as the oxygen content is sufficient high for the formation of oxide scales. In Pb a better behaviour of the steels can be expected. However, if the oxygen concentration drops to low values even at this temperature dissolution attack will become an issue. Tests with 316L steel showed that at 450 °C in LBE with low oxygen content dissolution attack already starts after 2000h (Fig 20a). In LBE at a temperature of 500°C dissolution attack is observed after about 10000h. In Pb at the same conditions regarding temperature and oxygen content oxide scales still protect the steel (Fig. 20b). The usability of the steel 316L up to the selected upper temperature limit of 480°C at “normal” operating conditions seems obvious if the oxygen is controlled within the given limits.
The 15-15 Ti stabilized steel (1.4970) foreseen as fuel cladding material, has to be stable up to 550°C or even 570°C. In LBE this temperature is clearly above the operating range for this steel. Local deep dissolution attack starting with leaching of Ni and later Cr will not allow the use of the 15- 15Ti in LBE. But, in Pb this steel showed in experiments at same temperature and oxygen activity the formation of protective oxide scales at least in the first 3000h of exposure. Based on the experience with extrapolation of corrosion data from short to long term experiments, it is unlikely or at least uncertain whether this steel can be operated in Pb at 550°C for longer times. One difficulty that arose during the experimental work is the so called incubation time for the dissolution attack. At 500°C e.g at 10-8wt% oxygen a typical oxide scale develops during exposure for 2000h [Fig 21a]. However, after 5000h dissolution attack wa111s observed. On possible mechanism is the enrichment of the high soluble elements like Ni and Mn beneath the outer oxide sale [Fig. 21b]. These enriched phases are located in the spinel layer and can reach almost the interface to the outwards growing magnetite scale. The magnetite scale is rarely a dense and compact layer and quite frequently penetrated by the liquid metal. Therefore, the Ni and Mn enriched phases might act during prolonged exposure as a starting point for the dissolution attack if they reach the original metal surface and come in interaction with the liquid metal. Therefore all short term tests (<5000h exposure) must be carefully analyzed on any sign of local enrichment that can possibly predict the onset or starting of a dissolution attack in future exposure times.
f/m Steel T91:
The operating temperature range for the f/m steel T91 is similar to that of the austenitic steel between 380 and 550°C. Again, the upper limit is not design but material driven and might be increased if adequate solutions are available. The f/m steel T91 does not contain Ni. Therefore, the susceptibility for dissolution attack is reduced compared to the austenitic steels. For the entire foreseen range of temperatures no dissolution attack of T91 steel is expected also for long exposure times if oxygen concentration is kept under control at nominal values. The formation of oxide scales, which is the prevailing mechanism for Ni free f/m steels, is mainly independent of whether the medium in contact is pure Pb or LBE. The oxidation potential as the driving force and the diffusion constants in the steel, both determine the oxidation behaviour, are anyway similar for both coolants. Above 480°C typically three layered oxide scales that grow significantly with time are formed (Fig. 22a). On average 30μm thick scales grow after 6700h. Assuming parabolic growth, oxide scales of about 40μm thickness are formed after 15,000h. At 550°C oxidation is accelerated and the scatter of the measured data becomes significant (Fig. 22b). The average thickness of the spinel scale grows up to almost 60 μm after 15,000h assuming parabolic oxidation. The large scatter makes long term prediction difficult.
Besides the lack of precise and reliable data for long term, the observed oxidation becomes an issue regarding heat removal through fuel claddings and SG and HX tubes. Temperature increase of cladding tubes due to the reduced heat conductivity of the oxide scale and the reduced efficiency of the SG and HX must be considered during design or answered by a different solution like aluminizing of the steel. Besides, oxygen supply has to consider this effect too.
Alternative materials
For the impeller of the pump the temperature range is between 380 and 480 °C. Regarding solely corrosion, both classes of steels, 316 type and T91 type, can be used. But, the expected local velocities of up to 10m/s or even higher might be problematic for the use of the materials described above.
Localized high velocity flow pattern can severely damage steels as observed in corrosion experiments at IPPE and CIEMAT. 316L steel was heavily eroded in such flow field at temperature and oxygen conditions that are suitable to form protective oxide scales at nominal flow velocities.
The materials discussed at the time are Maxthal, SiSiC or coatings (Ta).
Maxthal: Tests with Maxthal in Pb up to 750°C revealed its excellent compatibility with the liquid metal. The protective oxide scale, depicted in Fig. 23a, is quite thin and composed of TiO2, SiO2 and mixed oxides. This material is discussed as pump impeller material and has to withstand high local velocities.
SiSiC: SiSiC was also tested in Pb up to 600°C and did not show any incompatibility with the liquid metal.
Ta Coating: Ta as refractory metal has no solubility in Pb. Therefore coating of an impeller was discussed as an other alternative solution. Basic exposure tests in Pb showed the expected incompatibility of the Ta with oxygen containing Pb at higher temperatures. At temperatures above 400°C the oxygen diffuses into the Ta and embrittle the material, as seen in the hardness increase of Ta exposed at 500 and 550°C in Fig. 23b. For longer times and higher oxygen contents even entire oxidization of Ta sheets (200μm thick) were observed. The simultaneous use of Ta and oxygen in Pb seems not to be a suitable option for material protection in a Pb cooled nuclear reactor. Tests with a Ta coated steel impeller pump in oxygen reduced Pb are at the time under preparation. If successful the same impeller will be tested also in oxygen containing Pb at 480 °C to monitor whether the oxygen uptake and through that the embrittlement is significantly slowed down at this lower temperature.
At KIT the CORELLA facility was re-build to achieve rotation speeds higher than 1000 rpm and Pb velocities large than 10m/s. This objective was reached and two corrosion/erosion tests with candidate materials were performed at 480°C. The 1st test was stopped after 100h, because the target oxygen content of 10-6 wt% oxygen could not be achieved, the measured value was always far below 10-8wt% oxygen. However, the samples were investigated and only at the sides facing the flowing Pb some damage was observed. For Maxthal it was very difficult to detect erosion due to the non-perfect sample preparation. From the other tested material Noriloy with higher Cr shows the best behaviour. The worst behaviour was observed for T91 material in original status, at which up to 50µm deep erosion was detected. The SiSiC specimen got lost during the 100h. The free Si might have been dissolved at these conditions or the mounting might broke. The 2nd test run for 780h in Pb with expected oxygen content or a slightly higher one. At this condition and after 780h Noriloy and Norihard behave quite similar showing up to 30µm deep erosion at the front side facing the Pb. All the rest of both specimens stay entirely intact. The Maxthal was not tested and the SiSiC was attacked both by the Pb and the etching solution applied after the experiment. Again the non-surface alloyed front side of the GESA specimen, the non-modified T91 part was heavily attacked. Erosion up to 150µm deep at a distance of 1.5mm was observed. All the parts of the specimen that were surface alloyed using GESA do not show any attack.
ENEA designed and constructed a new loop that allows testing of real pump impellers at relevant conditions regarding temperature and oxygen content. The facility will work in pure molten lead and a prototypical mechanical pump is installed and will be tested for the corrosion at the impeller up to 500°C. The components and the process of the facility have been described in the respective deliverable in details. All experiments had to be postponed for the time after the LEADER project.
The scheduling of the experimental activity on the HELENA loop is to fill the facility with molten lead in December 2013 and to start the continuous test on the pump in January 2014. The test will be isothermal at 400°C with low oxygen content <10-8 % in weight; the oxygen content will be monitored continuously. The pump will be gradually driven to the reference mass flow rate 35 kg/s and this mass flow rate will be maintained for 1500 h, i.e. 2 months about. After this, the pump is stopped and the loop is drained. Then, the pump impeller is disassembled by the body and it is analyzed for corrosion.
Surface protection
Fuel claddings are operated up to 550°C. The use of 15-15Ti stabilized steel for the ALFRED cladding is doubtful or at least critical at this elevated temperature due to expected dissolution. The cladding material of the commercial ELFR, the f/m steel T91, will not suffer from dissolution, but oxidation becomes a severe problem. Therefore, surface protection barriers are required to ensure safe operation of Pb cooled fast reactors at the envisaged temperatures. Dense, and stable and slowly growing oxide scales that are formed in-situ are the preferential protection barrier. Alumina, which is employed for protection of stationary gas turbines, but at higher temperatures, is one option. Surface aluminizing can be used to enrich a base steel (either 316 or T91 type steel) with aluminium resulting in the potential to form alumina scales during use. One applicable method is the surface alloying of deposited coatings using pulsed electron beams. By that a surface graded material having the target Al concentration in the outer surface region can be manufactured [Fig. 24a]. Such scales showed their potential to protect steels up to 600°C even in LBE for more than 10,000h. To ensure the formation of such protective scales over the temperature range of 380 to 550°C an optimized FeCrAl composition was explored experimentally. An Al content of 6wt% with 16wt% Cr was effective to form thin Al-rich scales between 400 and 600°C in Pb [Fig. 24b].
Mechanical property degradation
Besides simple corrosion also mechanical properties like creep strength and wear resistance might be affected by the contact with the liquid metal coolant. It was shown that liquid Pb has a deteriorating effect on mechanical properties if it comes in direct contact with the steel. Oxide scales that prevent the direct interaction will, as long as they stay intact, mitigate any effect of the Pb. The creep rupture strength e.g. of f/m steel T91 is reduced remarkably due to the contact with the liquid metal. The influence of the coolant was negligible in case of surface aluminized T91. There, the thin alumina scale was able to catch the deformations of the steel without cracking. Detailed evaluation allowed to define a threshold stress of about 120 MPa, below which, mainly due to the reduced strain that does not result in oxide scale cracking, the influence of Pb becomes negligible [Fig. 25a]. 316L type steels in contrast does not show such pronounced effects at all.
Another interesting phenomena is fretting in Pb, which is a specific type of wear occurring at fuel claddings and heat exchanger tubes. Small amplitudes (<200μm) can lead to severe damage. Tests were done at harsh, accelerated conditions to evaluate the possible influence on the life time of the related components. Steel T91 and the 15-15Ti stabilized steel and surface aluminized T91 were explored. Fretting wear can lead to fuel clad penetration at the T91 and 15-15Ti steel of more than 10% of wall thickness depending on the relevant parameters like amplitude load and temperature already after 150 h at a frequency of 10 Hz (Fig. 25b). Both steels show in addition a strongly increased wear with increasing temperature. The austenitic steel exhibits at 450°C dissolution attack already after 600h, which was never observed even for more than 10,000h at 500°C in Pb. The surface aluminized T91 steel has the highest resistance against fretting wear. Post examination based on the construction of fretting maps allowed to predict the parameters that allow safe operation of fuel cladding regarding fretting damage. Loads above 75N and amplitudes below 15μm combined with surface aluminizing are safe parameters.
R&D NEEDS
Technology breakthroughs and innovations are needed for all Generation IV fast reactors. Innovative design and technology features are needed to achieve high safety and security standards, to minimise waste and enhance non-proliferation through advanced fuel cycles, as well as to improve economic competitiveness, especially with a high availability factor. In particular, structural materials and innovative fuels are needed. to sustain high fast neutron fluxes and high temperatures, as well as to comply with innovative reactor coolants. The development and qualification of new fuels also require significant R&D effort in terms of resources and time.
Key R&D topics
To achieve commercial availability of all Fast Reactors technologies by 2040, it is vital to continue and accelerate R&D on key topics, primarily: primary system design simplification; improved materials and innovative fuels; improved safety (safety standards at least comparable to Gen III standards); innovative heat exchangers and power conversion systems; advanced instrumentation and in-service inspection capabilities; reduction of the plant costs; partitioning and recycling processes; plant availability improvement. In particular the work in the area of partitioning and transmutation with innovative fuels, core performance and associated fuel cycle studies will lay the groundwork for assessment of future sustainable nuclear fuel cycle strategies. Whether and to what extent it should involve transmutation in a dedicated waste-burning Accelerator Driven System or in future Generation IV power plants needs to be assessed. Specifically focusing on LFR, the main challenges for their feasibility are: corrosion and erosion of structural materials; erosion in primary pump; seismic risk due to large mass of lead; in-service inspection of core support structures and possibility of replacing internal components; refueling at high temperature in lead; spent fuel management by remote handling; managing of the Steam Generator Tube rupture inside the primary system; prevention of flow blockage and mitigation of core consequences; radiation damage (at high temperatures); development of measuring devices.
Corrosion and erosion - The use of existing industrial materials for primary system is possible by limiting the core outlet temperature but new materials need to be used for special components (e.g. pump impellers). The present technological solutions are the use of low carbon grade austenitic steels for components at relatively low temperatures and low irradiation fluence, e.g. the reactor vessel, and the use of ferritic-martensitic steels for fuel cladding and core structures. Techniques for coating (i.e. aluminization) of steam generators tubes, fuel cladding and core structures operating at high temperature need further development. R&D needs come from the qualification of: i) an austenitic steel for the reactor vessel, ii) a lead corrosion and erosion resistant material for the in vessel structure, iii) a protective coating for fuel cladding and steam generator tubes, iv) special materials corrosion and erosion resistant for the impeller of the mechanical pumps.
Seismic risk - The mass of lead can be kept low by means of an innovative layout (e.g. Steam Generating Units and Primary Pumps installed in the reactor vessel). In such a configuration of the reactor vessel, seismic loads defined by the European Utility Requirements can be accommodated by means of 2D seismic isolators of the reactor building. To study seismic risk, a specific FP7 project (SILER) has started on 1st of October 2011.
Refueling - The fuel assemblies can be fitted with an extended stem to permit fuel handling, using a simple handling machine that operates in the cover gas at ambient temperature under full visibility. This eliminates in-vessel fuel transfer equipment which has never been designed or tested in lead. The problem of refueling can be considered solved by design, so no specific research activities are needed.
Purification and inspection - The use of molten lead as the coolant implies also research focused on the development and validation of a technique for lead purification before reactor vessel filling and with reactor in operation to prevent/control slag/aerosol formation, on the development and calibration of instrumentation operating in lead and under irradiation, on the development of techniques and instrumentations for in-service inspection of the steam generator tubes and the reactor vessel and on the development and validation of oxygen control techniques for regulating oxidation kinetics.
Fuel cycle - Significant R&D efforts should be focused on innovative fuels and fuel cycle (minor actinides bearing fuels, dense fuels such as nitride or metallic fuels). In the near term an essential goal is to confirm that ready-to-use technical solutions exist, so that fuel can be provided in timing with the ETPP operation. In the mid-term, it is necessary to confirm the possibility of using advanced MA (Minor Actinide)-bearing fuels and to confirm the possibility of achieving high fuel burn-ups. In the long term, it is important to confirm the potential for industrial deployment of advanced MA- bearing fuels and the possibility of using fuels that can withstand high temperatures to exploit the advantage of the high boiling point of lead.
Research and testing facilities - To meet the R&D challenges presented in the previous section, several facilities are required. Therefore the following actions are required: i) design, construction and operation of the necessary irradiation tools and devices to test materials and fuels; ii) design, construction and operation of the necessary fuel fabrication workshops, dedicated to uranium-plutonium fuels, and minor actinide bearing fuels; iii) design, construction, upgrading and operation of a consistent set of experimental facilities for component design, system development, code qualification and validation, that are essential to perform design and safety analyses of the demonstration program (e.g. hot cells, gas loops, liquid metal loops).
Experimental irradiation capacities - There is a clear need to update European irradiation facilities, given that existing facilities are close to end of life. New facilities are currently in construction or design phase in Europe: i) JHR, the Jules Horowitz Reactor (Cadarache, France) dedicated to materials testing for nuclear fission; its construction has started in March 2007, the start of operation is foreseen in 2014. ii) PALLAS (Petten, The Netherlands), a 'tank-in pool' reactor with a high neutron flux, a power of 45-80 MW and an annual degree of utilisation of more than 300 full-power days, equipped to meet the demand for both nuclear knowledge and services and the production of essential medical isotopes. PALLAS will substitute the existing HFR. According to the present plans, PALLAS could be in full operation in 2020. iii) MYRRHA (Mol, Belgium), a flexible fast neutron irradiation facility, able to test both lead coolant systems and accelerator-driven sub-critical systems (ADS) for transmutation. Beyond the availability of these irradiation capacities, it is necessary to develop new experimental devices taking into account cutting-edge progresses in modelling, instrumentation and modern safety standards. Europe has a worldwide leading position in this field and has to keep it through intra-European synergistic developments, to overcome shortage of resources.
Experimental facilities for reactor physics - Development of dedicated experimental facilities for reactor physics in fast reactor systems for component design, system development and code qualification and validation is necessary. GUINEVERE, already in operation, and ELECTRA, in its design phase, are such kind of machines.
Other experimental facilities - R&D demands further experimental facilities for: i) LFR material, coolant physics-chemistry and corrosion/erosion studies; ii) LFR safety experimental studies; iii) LFR studies of moving mechanisms, instrumentation, maintenance, in-service inspection and repair; iv) LFR studies in thermal hydraulics and heat transfer.
Fuel reprocessing capacity - Existing large research facilities (ATALANTE at CEA in France, ITU at JRC in Germany, the Central Laboratory at NNL in the United Kingdom) offer effective potentialities at lab-scale in the short term. They should be used in the future to develop suitable processes and to perform demonstration runs on samples of spent fuel, or on irradiated targets at up to pin-scale. For oxide fuel processing, minor actinide recovery processes, under development at lab-scale, mainly rely on well-known and industrially mature solvent extraction technologies. The important background coming from industrial plant feedback, or from the very important work carried out over past decades to design modern reprocessing plants, make extraction a well-mastered technology. Therefore, considering that there are no important issues for scaling-up hydro-metallurgical processes, the requirements in this field could be postponed. The present reprocessing capacity may represent a bottleneck in the future, due to the overlapping of the time schedules for the development of the SFR and LFR (MYRRHA and ALFRED).
Fuel manufacturing capacities - Beside existing facilities (ATALANTE, ITU, Central Laboratory facility), it is important to improve capability in the field of experimental fuel fabrication. A Prototype Core Facility (PCF) will be needed in a short time, for the production of the MOX fuel to be loaded into the core of the fast reactors prototypes, demonstrators and experimental. An industrial facility is under preliminary design in France by AREVA and CEA. A pin-scale facility will be also needed, able to provide in an efficient manner the (very diverse) experimental pins to be irradiated in experimental facilities during the early phases of the design of possible future fuels (MA-bearing fuel, other than oxide fuel). Such a facility could be located in existing hot labs. The goal is an efficient, modern and flexible tool with the capacity to produce from a few pellets up to a few pins per year, to address the many and diverse experimental needs expected from the R&D fuel research community.
Road Map
The main drawback facing the industrial deployment of a LFR fleet in Europe is the lack of operational experience gathered so far. Whilst a satisfactory technological readiness has been achieved in Russia in the past century, driven by military research, the aim at filling the technology gap in Europe requires the setting up of a complete R&D roadmap, in complete analogy with what has been done in France for the development of the SFR technology chain, whose readiness is almost proven.
The LFR development roadmap (see Fig. 26) towards the industrial deployment of a fleet of European LFRs, requires at first a small Demonstrator reactor (ALFRED) for proving the viability of reliable electricity production for LFR systems. A Prototype reactor is then envisaged, for testing the scaling laws at an intermediate step – according to a common approach focused both on plant size and representativeness of the target reference system – before moving to the First-Of-A-Kind (FOAK) representative of a commercial ELFR fleet.
Besides the reactors planned for the demonstration of the LFR technology chains, further facilities – either existing or under construction – are required (see Fig. 27) to provide the overall frame for LFR technology development, like all the existing and planned EU lead labs focused on thermal/hydraulics, materials development and corrosion testing, like the zero-power facility Guinevere, already operating in Mol, Belgium, like the irradiation facility MYRRHA, to be built and operated in Mol and like the European training facility ELECTRA, planned for realization in Sweden.
The road map of LFR is based on the progressive upscaling from the zero power facility GUINEVERE, in operation since 2010, to the commercial deployment of the first-of-a-kind European industrial size LFR (ELFR) by 2045. The intermediate steps planned are the European Technology Pilot Plant (ETPP), MYRRHA, to be in operation in 2023, the European LFR Technology Demonstrator Reactor (ETDR), ALFRED in 2025 and an LFR prototype of the industrial plant (PROLFR) in 2035.
The construction and operation of MYRRHA and ALFRED are the key steps towards industrial deployment, as they would enable important experience to be acquired for the design and operation of industrial power plants. They would not only establish a specific technology but also act as multipurpose research infrastructures. In particular ALFRED could be an experimental platform to address innovative fuel development, for testing specific minor actinide-bearing fuel to demonstrate and optimise transmutation and also for addressing the development of high temperature applications for industrial end-users. Key issues to be solved to allow the correct proceeding of the road map are the following: i) selection of materials for demonstrators and prototypes; ii) fuel development and qualification (MOX fuel, and in a later phase advanced minor actinide bearing fuel, and lead-fuel interaction); iii) Lead technology (lead purification/filtering techniques, oxygen & chemical control); iv) components development (safety & control rods, pumps, heat exchangers, in service inspection and repair technologies); v) development of models & tools (to study the nuclear/thermal-hydraulic feedback, the reactor stability, as well as the reactivity margin for not reaching prompt-critical conditions, response and resistance of structure to lead sloshing); and vi) conduction of large scale integral tests to characterise the behaviour of the main systems, especially for licensing procedures (key component performance and endurance demonstration, benchmarking of thermal-hydraulics in a rod bundle).
LFR Zero Power Facility (GUINEVERE) - GUINEVERE is a test fast reactor coupled with an accelerator, already in operation since 2010 and constructed by SCK•CEN in cooperation with European partners. It is a zero power facility designed for core design qualification and reduction of design uncertainties (critical mass, power distribution as well as reactivity coefficient).
LFR ETPP (MYRRHA) - MYRRHA is a research FR cooled by Lead- Bismuth Eutectic (LBE) with a power of 70 MWth when operated as a sub-critical mode (as an accelerator driven system) and a power of 100 MWth when operated in critical mode. It employs a MOX fueled core. The main mission of MYRRHA is to test both technologies of fuel and heavy liquid metals, and the endurance of materials, in-service inspection and repair, components and systems to control industrial risks (obtain reactivity feedback at power) for ALFRED, PROLFR and ELFR. Comparing the scope and specifications, the calendar and the current status of the project MYRRHA will fulfill the role of the LFR ETPP (with no need for electricity production). The realisation of MYRRHA will include several phases (2010-2024): Front End Engineering design, tendering and procurement, construction of component and civil engineering, on site assembly, commissioning, progressive start-up and finally full exploitation. The completion of the design phase is planned for 2015. The construction and commissioning is expected in 2022. In parallel, the support R&D program will provide during the 2010-2014 period the necessary answers to the remaining technical challenges. After 2014, the fuel qualification and to a lesser extent the material qualification will remain the main topics of the R&D support program. The completion of fuel and material qualification processes is expected in 2020.
LFR ETDR (ALFRED) - ALFRED is a pool type LFR of 125 MWe ~300 MWth. The primary coolant is pure lead in forced circulation. It can be cooled by natural circulation in emergency conditions. The core inlet temperature is 400 °C and the average core outlet temperature is 480 °C. The secondary coolant is water-superheated steam with the feed-water temperature of 335 °C, the steam temperature of 450°C and the pressure of 18 MPa. The secondary system efficiency is about 41%. The reactor vessel is in austenitic SS. The safety vessel is anchored to the reactor pit. The core barrel – cylindrical and integral with the core support grid and the hot collectors – is designed so as to be removed. The steam generators are bayonet type with double walls and they are integrated in the reactor vessel and removable. Primary pumps are mechanical, installed in the hot collector and removable as well. The fuel assembly is closed hexagonal, weighed down when primary pumps are off but forced in position by springs when primary pumps are on. The fuel is MOX (max Pu enrichment 30%). The design maximum discharged burn-up is 90-100 MWd/kg-HM. Fuel clad material is coated 15-15/Ti. Control/Shutdown Systems are two diverse and redundant systems of the same concept derived from CDT: the 1st system is made of pneumatically inserted absorber rods to be passively actuated by depressurization from the top of core; the 2nd system is made of buoyancy-driven absorbers rods passively inserted by buoyancy from the bottom of the core. As LFR ETDR, ALFRED will have to demonstrate the viability of the LFR technology for use in a future commercial power plant, construction and operation. It is expected to significantly reduce uncertainties in construction and licensing.
The demonstration of economical and reliable electricity generation is a key goal, thus the reactor core and primary coolant path should be as prototypical as possible, as well as the secondary circuit and the Balance of Plant (BoP). It will also allow global evaluations on the plant responses to the electrical grid needs. The objectives regarding neutronics are mainly related to the need of demonstrating the feasibility of an adiabatic core, which is a very crucial issue for an actual sustainability with respect to both natural resources and environmental impact. As to thermal–hydraulics, a key feature to be tested in ALFRED is the coolant ability to remove the decay heat in natural circulation under accidental conditions. Materials extensive testing will not be among the main goals of the demonstration project. Anyway, core materials (fuel, cladding and structures, coolant) should be the same of the ELFR ones so that ALFRED may be as much representative as possible. Within the LEADER project the design of the ALFRED is undergoing but the main suitable characteristic and design guidelines for the facility have been identified. A scaled demonstrator fully representative of the industrial size reactor is being designed, using components/technologies already available in the short term, to be able to proceed in the near future to a detailed design followed by the construction phase. The realisation of ALFRED will include several phases (2010-2025) see Fig. 28: the first step is identified in the set-up of an international consortium (2013) having already chosen a dedicated site for construction (Pitesti in Romania), in parallel several design activities should be running (basic design, siting and pre-licensing in the period 2013-2016 and detailed design and licensing in the period 2016-2019), finally construction of components and civil engineering, on site assembly and commissioning (2019-2025). The completion of the design phase is presently planned for 2019 to be able to take into account the feedback from design and experience from MYRRHA. In parallel, the support R&D program will provide during the 2011-2018 period the necessary answers to the remaining technical challenges. The ALFRED full operation is expected in 2025.
LFR prototype (PROLFR) - The PROLFR needs to be designed. The need of this step emerged in the LEADER project from the consideration that a direct power upscaling of factor 5 from ALFRED to ELFR was considered risky. Its power, of the order of 300-400 MWe, and its characteristics will be therefore between the ones of ALFRED and ELFR. PROLFR realisation is foreseen at the horizon of 2035.
First of a kind European industrial size LFR (ELFR) - Design and construction of the first of a kind 600 MWe industrial reactor is foreseen in 2035 to start with LFR Industrial plants building up at the horizon of 2040-2050.The reactor design parameters and characteristics are the same as in ALFRED with few exceptions. Electric power is 600 MW. Fuel is MOX with equilibrium MA content, maximum Pu enrichment will be lower than in ALFRED. The target value for maximum discharge burnup is 100 MWd/kg-HM. The target secondary system efficiency is above 40%.
LFR Education and Training facility (ELECTRA) - ELECTRA is a low power fast neutron reactor cooled by liquid lead whose design has been presented by KTH. Application of (Pu, Zr)N fuel permits to design a core with very small volume and fuel column height, resulting in highly negative coolant, fuel and structure temperature coefficients and very low channel pressure drop. Full design power of 0.5 MWth can be completely removed by natural circulation in the primary circuit, thus eliminating the need for pumps. Analysis of flow stability and performance under unprotected transients show that the suggested design is highly safe. ELECTRA's mission is in training and education. In particular it could be employed for training operators and staff of the other LFR in the roadmap, such as MYRRHA and ALFRED. In addition it could be employed for research on reactor dynamics. The main milestones for ELECTRA completion include the construction and operation of an electrical mockup at KTH in scale 1:1 in 2013-2014; the completion of the engineering design in 2016; the carrying out of irradiation experiments for licensing of cladding and fuel in 2012-2020 and the licensing for operation of ELECTRA in 2022.
ELFR CONFIGURATION AND SAFETY ANALYSES
Compared to ALFRED, the ELFR is a larger system (600 MWe), since it is intended primarily for central station electricity generation.
The design of most of its primary system components, such as reactor vessel, fuel assembly, inner vessel and core supports, is the same already described for ALFRED, with larger dimensions. Some other components are different, since, in order to simplify the design and licensing phases for an early construction of ALFRED, proven and qualified design solutions have been adopted for the demonstrator. In particular:
- Steam Generator: in the current ELFR configuration a once through SG with spiral tubes bundle is considered, whereas ALFRED SG is constituted by double wall straight bayonet tubes, allowing an easy coating and/or surface treatment to speed-up the construction;
- DHRs: in ELFR, the DHRs consistof two diverse systems: the Isolation Condenser and the DHR-2 system, constituted of dip coolers immersed in the reactor pool. In ALFRED, considering that the bayonet concept allows a continuous monitoring of the SG status, the diversity requirement has been relaxed: the two DHRs are both constituted by Isolation condensers;
A list of the main parameters concerning the design of the ALFRED and ELFR is provided in Table 2.
Safety analysis of the ELFR was performed testing all the “weak” points in the LFR design that were determined during the previous ELSY project. The most important DBC and DEC transients were repeatedly reanalyzed for the re-designed ELFR configuration. The full list of the analyzed transients for ELFR is asfollows:
- Protected loss of flow transient (PLOF),
- Unprotected loss of flow transients (ULOF),
- Unprotected loss of heat sink transient (ULOHS),
- Unprotected reactivity insertion transient (UTOP),
- Unprotected loos of flow and loss of heat sink transient (ULOF+ULOHS),
- Protected overcooling transient (OVC),
- Protected steam line break transient (SLB),
- Unprotected sub-assembly (SA) blockage transient, and
- Steam generator tube rupture (SGTR) accident.
The results of the safety analyses can be summarized as follows:
- Protected transients (PLOF, OVC and SLB): the automatic reactor shutdown activated by different scram signals is able to rapidly bring the ELFR plant to safe plant condition. The consequent isolation of the secondary circuits and start up of decay heat removal system is able to maintain the plant in safe condition in the medium and long term. In all transients, the potential of lead freezing in the coldest points of the primary system is reached after several hours within the transient, assuring sufficient grace time for manual, corrective operator action.
- Unprotected transients (ULOF; ULOHS and ULOF + ULOHS): due to the enhanced natural convection capability in the primary circuit, in case of ULOF the maximum temperatures reached in the primary system are low enough to assure the integrity of the clad and the vessel in the short term, providing sufficient grace time for corrective operator action. The main potential safety issue is the maximum reactor vessel wall temperature, which might exceed 700 °C within ~12 min. The integrity of the clad and the vessel seems not guaranteed in the medium/long term, because of the high temperatures reached in the primary system. An optimization of the neutronic core design, in order to reduce the positive coolant expansion reactivity feedback, could provide additional grace time.
- Reactivity insertion: for reactivity insertion of 200 pcm in 10 sec time interval at EOC conditions, peak fuel pin cladding survives and fuel melting is not observed, even in the center of the peak fuel pins (pellets). For reactivity insertion of 260 pcm in 10 sec time interval at EOC conditions, peak fuel pin cladding survives, however fuel melting should be expected in the center of the peak fuel pins (pellets). These transients envelope positive reactivity insertions of the design basis events such as fuel handling errors, control rods withdrawal or seismic core compaction.
- FA flow blockage: for blockages less than 75% blockage area, it is not expected any pin failures nor fuel melting, even under unprotected conditions. For blockage above 75%, peak power pins clad failure shall be expected, but fuel melting is not expected even for blockage over 97.5%. However, there is enough time (several hundreds seconds) to detect the flow blockage occurrence, by means of temperature measuring devices installed at each FA outlet.
- SGTR accident: several limiting mechanisms and potentially important effects have been analyzed and suggest that: (i) the initial pressure shock wave poses no likely threat to in-vessel structures, except very few adjacent heat-exchange tubes; (ii) the sloshing-related fluid motion is well bounded in a domain beyond the heat exchanger; and yet (iii) the steam/water entrainment is expected to be comparatively limited due to the verylarge difference of density between steam and lead. The potential gradual pressurization of the vessel after SGTR due to inflow of steam is limited by rupture disks to relief the resulting over-pressure. Moreover, a Venturi nozzle placed inside each spiral tube, mitigate the severity of SGTR interaction and reduce the potential effects on the entire reactor system. Anyway, a dedicated scaled facility should be foreseen to analyze in depth the SGTR phenomena further as part of the future R&D activities.
In order to assure prevention of freezing of the lead coolant at the coldest location of the primary loop, a tight and continuous operational control of the secondary coolant conditions is needed. Under certain adverse transient conditions it is conceivable that the primary lead HX outlet temperature (nominally 400°C, well above the freezing point of lead (327°C)) decreases to the feedwater inlet temperature 335°C, (only 8 °C margins to freezing). In addition, any malfunction in the FW temperature control could progressively bring the coolant to its freezing point. In general, the safety analysis performed for the lead-cooled ELFR design demonstrated the extremely robust nature of this plant design when compared to other similar plant designs, ascribable to the inherently, large thermal inertia of the lead-cooled primary system and optimization of safety relevant control, safety systems and components.
Potential Impact:
Electrical energy generation by means of nuclear power plants is at present an important fraction of the electric energy generation and continues to be an essential option for the future in Europe and worldwide.
The crucial issue of nuclear power, however, is that it should simultaneously be competitive, reliable, sustainable, safe, and perceived safe by the public and manage adequately the long-lived waste that it produces.
Modern water-cooled reactors are competitive and safe, but their energy generation is far from being sustainable, owing mainly to their ineffective use of the fuel and the production of long-lived highly radioactive waste, and to their high operating pressure which makes the high level of safety more difficult to achieve.
The LEADER project can have an important societal impact because of the emphasis on sustainability. The Public confidence in nuclear power depends to a large extent on the ability to demonstrate that it represents an unlimited resource of energy without creating critical issues for waste management. If LEADER confirms LFR as a reactor able to substantially reduce the long-term burden of a geological disposal and increase the extent of the present uranium resources, it will contribute to increase confidence and acceptance of nuclear power as an essential component of the energy mix within the EU.
A huge potential increase in the sustainability of nuclear energy will be achieved through demonstrating the technical, industrial and economic viability of Generation IV fast neutron reactors, thereby ensuring that nuclear energy can remain a long-term contributor to a low carbon economy. This demonstration program will play a key role by involving European industry and maintaining and developing European leadership in nuclear technologies worldwide. It will also make possible the further commercial deployment by the European industry of these technologies by 2040 and beyond. This is the prime goal for industry, which in the mean time will seek to maintain at least a 30% share of EU electricity from currently available reactors for the benefit of the European economy (the industrial needs for nuclear energy could be enhanced with an expansion towards co-generation of process heat for industrial applications when such markets develop). With the construction and operation of a LFR ETPP (MYRRHA) and ETDR (ALFRED), Europe will be in an excellent position to secure the development of a safe, sustainable and competitive fast spectrum technology. The program will allow to investigate and address the main technological issues that can then be implemented in the LFR prototype around 2020-2035. This LFR prototype, in turn, will pave the way for industrial deployment of LFR by 2050, and hence contribute significantly to the development of a sustainable and secure energy supply for Europe from the second half of this century onwards.
List of Websites:
http://www.leader-fp7.eu
Project coordinator: Alessandro Alemberti, Ansaldo Nucleare, Corso Perrone 25, 16152 Genova Italy (alessandro.alemberti@ann.ansaldo.it)
