LMFRs

Distribution coefficients may be further modified and operating temperatures reduced by dissolving uranium fuel in a low-melting metal such as bismuth or zinc. Separation of uranium from fission products by liquid extraction between molten bismuth and fused chlorides was extensively studied at Brookhaven National Laboratory [D5] in connection with the liquid-metal fuel reactor (LMFR), which used a dilute solution of in bismuth as fuel. Extraction of fission products from molten plutonium by fused chlorides was studied at Los Alamos [L2] in connection with the LAMPRE reactor. 

The Technical Committee Meeting (TCM) on Unusual Occurrences During LMFR Operation Review of Experience and Consequences for Reactor Systems was held on the recommendation of the International Working Group on Fast Reactors (IWGFR) at the IAEA Headquarters in Vienna from 9 to 13 November 1998. Participants from nine countries (China, France, India, Japan, Kazakhstan, the Republic of Korea, the Russian Federation, the United Kingdom and the United States of America) were in attendance. 

The discussion was focused on those aspects of LMFR technology which are unique and distinctive to plant design and operation. 

Stable operation of the demonstration reactor BN-600 in Russia with a nominal power output of 600MW(e) for 20 years and an average load factor of 72%, successful operation of the prototype reactors BN-350 in Kazakshstan and Phenix in France as well as the reliable operation of MOX fuel at high bumup (20% witti an irradiation dose in excess of 160 displacement per atom (dpa) in the cladding) in PFR (UK) and Phenix, are milestone in the implementation of LMFR technology. 

There is only one disadvantage inherent in the LMFR coolant, liquid-sodium, namely that it interacts chemically with water/steam and air. Protection against sodium fires was an important theme of the discussion at the meeting. 

Steam generators (SGs) are generally regarded as the most critical of all sodium system components. Design, manufacture and experimental testing should be carried out with special case. It seems that all was done to install reliable SGs in prototype, demonstration and semicommercial LMFRs. Three prototype fast reactors (BN-350, Phenix and PFR) were commissioned in the 1960s and two of them (BN-350 and PFR) had unforeseen occurrences with SG. 

It was concluded that the type of direct tube-to-tube plate weld adopted initially at PFR, which could not be heat-treated after manufacture, should be avoided in future reactors. The UK specialists consider that austenitic steels are unsuitable for LMFR steam generators because of the high risk of caustic stress corrosion damage following even small leaks. 

For the complete realisation of potentially feasible LMFRs from the point of view of cost, metal content and reliability new design solutions for SGs and other equipment may be required, differing from those being used at the present time. 

It was concluded that the use of ferritic steels of this type for structures in permanent contact with sodium should be banned for the future LMFR, austenitic steels being preferred. 

The solution of the problem of the reliable elimination of coolant leaks is determined by the application of experimentally and analytically proved design, structural materials, manufacture and installation, as well as quality control at all stages of LMFR components manufacture. Technological procedures and approaches, as well as quality criteria, should strictly correspond to the related regulatoiy documents. 

The comprehensive operational experience with LMFRs BN-350, Phenix, PFR, BN-600, Superphenix and Monju has shown that, if plant components have been designed and manufactured without errors and representative specimens or models have been tested prior to installation, reliable operation can be ensured during the whole operational life. 

The very low corrosion of sodium, the near atmospheric operating pressure, the use of ductile structural materials, and the reliable heat removal by a coolant having no phase change, imply that there should be nothing to provoke loss of the sodium system integrity in a LMFR. 

Each of these is srnnmarised below, with diagrams. The causes and the steps taken to rectify the problem are explained, and the general lessons learnt are set out in the context of the future development of LMFR technology. References are given in the cases for which more detailed information has been published. 

Oil ingress into the primary circuits of an LMFR is undesirable because of the potential release of methane gas through the core causing reactivity effects, and possible blockage of the subassembliesTjy solid carbon debris. In the case of PFR no reactivity effects were seen, possibly because the oil was retained in the pump cone for a prolonged period and... 

Lastly those events which are specific to the LMFR technology are again considered from the angle of the concepts retained for the European Fast Reactor (EFR). 

Differences in design practices have in the past handicapped developments that could have reduced costs. For example, some difficulties with past design practices are evident in examining application of the ASME Code and Code Case N-47 to pool type LMFRs, such as Super Phenix. The design of Super Phenix resulted in considerably thinner components. This required different buckling rules. Super Phenix design creep effects were negligible... 

The use of advanced computer modeling should allow reduced costs of fabrication, inspection and surveillance. In the longer term, it should be possible to develop uniform international code approaches for LMFR design that are less restrictive and allow improvements in plant availability and reliability. 

Because these developments are focused on intended applications in other fields and not problems of LMFR design there are significant differences in design lives, service conditions, materials, manufacturing practices, etc. The types of structures differ. The impact of these differences on such design information as constitutive models, material failure modes and models, and structural failure modes and consequences are sometimes difficult to assess. However, computer modeling, structural analyses methods, and analytic methods to understand materials behavior have advanced greatly in some of these non nuclear areas. 

It is obvious that application of these developments in non nuclear areas is not a trivial undertaking. In spite of the obvious difficulties adapting these developments to LMFRs offers considerable promise and should be aggressively pursued. The reductions in the overall level of LMFR development activities, with the attendant reduction in work specifically directed at LMFRs makes such effort doubly attractive. 

UNUSUAL OCCURENCES DURING LMFR OPERATION IAEA, VIENNA, 2000 IAEA-TECDOC-1180 ISSN 1011 289... 

Design of liquid metal cooled fast reactors (LMFRs) is still in evolution, and only a small number of LMFRs are in operation aroirnd the world. Specialists operating these LMFRs have gained valuable experience from incidents, failures, and other events that took plaee in the reactors. These unusual occurrences, lessons learned and measures to prevent recurrences are often either not reported in literature, or reported only briefly and without sufficient detail. Hence there is a need for specialists designing and operating LMFRs to share their knowledge on unusual occurrences. 

Considerable experimental and theoretical knowledge on various aspects of LMFR design construction, pre-operation testing and operation has been collected by several Member States with fast reactor programmes over the past decades. 

Large experience has been already gained with sodium cooled fast reactor operation. The use of sodium as a coolant poses fire danger in case of its leakage and interaction with air or water. Operating experience testifies the possibility of coping with the mentioned problem, but the quest for excellence calls for future improvement in LMFRs technology. 

Techniques to counter the heavy metal coolant disadvantages are being developed, but in spite of this work and the apparent disadvantages of sodium, the consensus in favour of sodium remains strong. This is demonstrated by fact that before lead-cooled fast reactor BREST-300 is built, MINATOM will first build a sodium-cooled LMFR BN-800 (E. Adamov, NW, 23 September 1999). Moreover, in the last few years sodium has been chosen in both China and the Republic of Korea for the respective fast reactor development project. This is a significant endorsement for sodium as a fast reactor coolant. 

However, given that there is now this tendency for countries to have their own viewpoint and their own preferred option evolutionary sodium-cooled LMFR models or an innovative one with new coolant (gas, steam, other than sodium liquid metal coolants), it is essential to clarify as far as possible the scientific issues related to the different innovative options and to exchange information on advances in development of traditional and innovative fast reactors. 

LMFRs(liquid metal cooled reactor) have been under development for more than 50 years. Twenty LMFRs have been constructed and operated. Five prototype and demonstration LMFRs (BN-350/Kazakstan, Phenix/France, Protot3q)e Fast Reactor/UK, BN-600/Russian Federation, Super Phenix/France) with electrical output ranging from 250 to 1200 MW(e) and large scale (400 MW(th)) experimental fast flux test reactor FFTF/USA have gained nearly 110 reactor-years. In total, LMFRs have gained nearly 310 reactor-years of operation. 

Significant technology development program for LMFRs is proceeding in several countries, namely in France, India, Japan and the Russian Federation. Activities are continued in a number of other countries at lower levels. 

Commercial introduction of fast breeder reactors in France has been postponed however, alternative LMFR application is being developed, namely transmutation of long-lived nuclear waste and utilization of plutonium. Continued operation of Phenix at 350 MW(th) is related to these requirements. One of the objectives of expanding the lifetime of the Phenix reactor is to perform the necessary irradiation experiments in support of the project identified as "Concept of Intensive Plutonium Reduction in Advanced LMFR. 

BN-600 with nominal power output of 600 MW(e) for 20 years (Fig. 2.1) with an average load factor of 70%, as well as construction of the largest fast reactor SPX in France, are milestones in the implementation of LMFR technology. 

Current efforts with regard to LMFRs in the Russian Federation are directed towards improving safety margins and economics. While these efforts will take some time, an immediate use is foreseen for fast reactors for energy production, as well as Pu and minor actinide utilization. In Russia, detailed design of commercial fast reactor BN-800 was completed, and license was issued for its construction on Yuzno-Uralskya and Beloyarskaya NPP sites. 

---