Monju is a 280MWe (714MWt) prototype fast breeder reactor (FBR) located at Tsuruga City in Fukui Prefecture, Japan. The reactor is a loop type sodium-cooled design, whose three secondary sodium loops outside containment provide heat to three helical-coil steam generators. It uses mixed uranium and plutonium oxide (MOX) fuel.

The former Power Reactor and Nuclear Fuel Development Corporation (PNC), now part of the Japan Atomic Energy Agency, began construction of Monju in 1985. It was to be Japan’s first industrial fast reactor demo, following on from Joyo, a 100MWt experimental scale fast reactor that went critical in 1977 (see also p23).

Monju achieved first criticality in April 1994 and was connected to the grid in August 1995. Six months later, on 8 December 1995, a sodium leak occurred from the secondary circuit at Monju. At the time the leak occurred, the reactor power was being raised for a plant trip test and had reached 43%. An alarm sounded at 19:47 due to an off-scale sodium temperature at the outlet of the intermediate heat exchanger (IHX) in the secondary circuit (loop C). A smoke detector went off at the same time and it was followed by a sodium leak alarm. Operators began normal shutdown operations at 20:00, after judging that a small sodium leak had occurred.

The reactor remained cooled and the safety of the reactor was secured throughout the incident. There were no adverse effects for operator personnel or the surrounding environment. A more detailed summary of the accident and post-accident response can be found in [1].

Investigations revealed that the secondary sodium had leaked through a temperature sensor, due to the break away of the tip of the thermocouple well tube that was installed close to the secondary circuit outlet of the IHX. The cause of the failure of the thermocouple well tube resulted from high cycle fatigue due to flow-induced vibrations. It was found that this flow induced vibrations were not caused by Von Karman vortex shedding, but a symmetric vortex shedding.

The original design of the thermocouple well was reviewed. PNC’s successor agency, the Japan Nuclear Cycle Development Institute (JNC), proposed a new design. In 2007, the Japan Atomic Energy Agency (JAEA) replaced 42 sodium temperature sensors in the secondary cooling loop with shorter well tubes to avoid the resonance vibration caused by the sodium flow. Another six sensors were removed from the plant.

Modification of the plant design began in 2005. The main improvements included installation of new sodium leakage detectors (TV monitors, smoke sensors etc.); remodeling of the drain system to shorten the sodium drain time; installation of a nitrogen gas injection system to extinguish sodium fires in the secondary circuit; and the division of the secondary circuit into four smaller zones to minimize the spread of aerosol. The design of the thermocouple well was also modified to prevent sodium leakage and applied in 2007.

After the modifications were complete, staff carried out a modified system function test (MSFT) from 18 December 2006 to 30 August 2007 to confirm the system-level function and performance of the modified facilities. After the long shutdown, station staff also carried out an entire system function test (ESFT) to confirm the plant-level function and performance of both the modified and unmodified facilities as well as the system-level function and performance of the unmodified facilities in shutdown/operation mode. The ESFT was conducted from 31 August 2007 to 12 August 2009. Neither test operated the reactor.

Following the ESFT, staff began preparations for a three-year system start-up test (SST). Staff were technically ready to restart Monju at the end of January 2010. Monju attained criticality on May 8. The SST consists of three steps. In the first core confirmation test, the performance of the core after long-term lay down is measured. Core characteristics are confirmed at the reactor criticality or very low output conditions. The minimum electric output for Monju is 40%. In the second 40% output confirmation test, the soundness of water/steam system and turbine/generator system will be checked. In the third rising power test, performance of whole plant will be checked in output ranges to 100%. Integrated controllability of reactor output, primary sodium flow rate, secondary sodium flow rate and water/steam flow rate are evaluated. Automatic safety shut-down characteristics of the plant are also checked by simulating abnormal situations. Required reactivity for each test (CCT, 40%CT, and PRT) will be supplied to the core by the refueling. Different core characteristics will be measured in the three different situations. After finishing these steps, full power operation will start. After reaching full power and a short stable operation, Monju will be stopped for periodic plant maintenance. It is free to do so, since Monju is a prototype reactor and is not under very strict economic requirements for outages.

Coverup and reorganisation

The limited technical problem of fixing the sodium leak was just one of several complex institutional and cultural issues that Monju operator PNC faced at the time of the leak in 1995. The difficulty of resolving these issues, and subsequent safety planning, delayed the restart of the reactor for 15 years.

The first issue emerged only weeks after the event, when PNC officials concealed parts of a video taken by personnel examining the leak just after the accident, and then released a misleading, heavily-edited version the day after the accident. Probes by the Fukui Prefectural Government and the then-Science and Technology Agency discovered the fraud [2]. PNC also delayed publication of a report about the accident.

The situation grew more complicated when an executive heading up an internal JNC investigation lied about when his team found out about the deception in a 12 January news conference. He said that his team found out 10 January; in fact his team knew about it almost three weeks earlier, on 25 December [2]. The executive, Shigeo Nishimura, committed suicide a day after the news conference. (A subsequent lawsuit from Nishimura’s family against PNC was dismissed in 2007 because the prosecution could not prove that PNC had forced the man to lie). Fifteen months after the press conference, in March 1997 PNC was involved in a nuclear accident at a bitumen solidification plant at the Tokai reprocessing plant (no one was hurt). After this event, the Science and Technology Agency again spotted a PNC false declaration, this time about Tokai. It quickly organised a restructuring team; its report in August 1997 spelled the end to the organisation as it then existed. (Two years later, a catastrophic nuclear accident small fuel fabrication plant in Tokai-Mura lead to the deaths of two people).

The review committee found that PNC had far too wide a scope. Its mandate included performing novel research and development, maintaining strict safety requirements and supplying highly-competitive technology. It found that PNC did not adjust to surrounding environmental changes and had poor board-level management. It found that management had bloated R&D to the detriment of safety and emergency response programmes.

To help fix PNC’s problems, the committee restricted the scope of the new organization to research and development in fast breeder reactors, related fuel cycle, and high level waste treatment and disposal. They said that public relations activities and information disclosure were not merely services for the public, but mandatory requirements for survival of the new organization. For example, at the time of the sodium leak, Monju staff had to obtain approval for information release by PNC headquarters in Tokyo. Now, information about unusual plant situations at Monju is generally transferred to outside within a half hour, day and night, on a routine basis.

So the organisation was renovated, renamed (to become the Japan Nuclear Cycle Development Institute, JNC), and reformed in management. It would eventually merge with Japan Atomic Energy Research Institute (JAERI) in 2002, to become part of the Japan Atomic Energy Agency (JAEA) in 2005.

Reorganisation work has continued until recently. A November 2009 JAEA report, approved in February 2010, suggested establishing an independent quality assurance system. At the time of the sodium leak accident, the employee in charge of quality assurance had another post under the director general of Monju. Now, a full-time quality assurance staff member has been employed; procedures for design review have been established and managed, and site management have been given more power to control the site by having the freedom to prioritise expert opinions, secure prompt management resources, and cultivate leadership independent of government agencies Nuclear and Industrial Safety Agency (NISA) and Ministry of Education, Culture, Sports, Science and Technology (MEXT).

In February 2009, a new station organization was launched, as well as new safety regulations (SR). The new fast breeder reactor research and development centre includes three departments (engineering, plant operations and plant maintenance engineering) and two offices (plant management coordination and safety and quality management). Previously, only one section managed maintenance at the site, in which one general manager had to lead nearly 100 workers. Now the section has been upgraded to a department with four sections to enable better quality assurance. Third, the quality management system of Monju, and the regional head office in Tsuruga, have been unified to promote the ‘plan-do-check-act’ sequence of quality management. A committee has been established to find nonconformity and take corrective action.

Public opinion

After the reactor shut down, the political climate also turned against fast breeder reactors. Foreign media claimed that FBR research and development should be stopped; moreover, the closure of France’s Superphénix plant in 1997 added fuel to the fire. In Japan, consensus about the national plutonium policy among the central and local governments needed to be formed.

Only one month after Monju received national approval for its planned modifications, on 27 January 2003, a regional court, the Kanazawa Branch of Nagoya High Court, ruled that the Monju pre-construction safety review was inadequate. It argued that it was possible that radioactive substances could be released from the reactor in the event of a secondary loop failure, that the review did not fully address the risk that the rupture of one steam generator tube might cause the rupture of another, and that the review did not provide adequate analysis of the prevention of core meltdown [3]. The ruling was appealed, but it took 30 months for the Supreme Court to reverse the Nagoya High Court decision (in summer 2005). It did so on the grounds that NSC’s safety assessment was ‘not unreasonable’ and that it did not ‘contain flaws that could not be overlooked’ [3].

The long duration of court deliberations made it difficult for the local government to actively support R&D at Monju.

The safety evaluation for seismic design of Monju was one of the biggest issues to be solved for the restart. After new seismic guidelines were published in 2006 [4] a number of geological investigations were conducted depending on the distances from the site: boring, trenching, surface layer digging survey, seismic reflection survey, airborne laser measurement, ocean bed survey, marine ultrasonic surveys and other methods. Results of the evaluation of the 2007 Niigata-ken Chuetsu-oki earthquake also played a role.

NISA and the Nuclear Safety Commission confirmed the seismic safety of Monju. The standard seismic motion Design Basis Earthquake Ground Motion (DBEGM) was originally 466 Gal, but now has been upgraded to the new standard of 760 Gal. The Monju facility as a whole has been evaluated and has been confirmed as resistant to the new 760 Gal figure. Evaluations were conducted for safety-related elements including equipment, piping, buildings, and civil structures, ground stability, and safety evaluations against tsunami and backland disruption. Equipment with little margin against the maximum DBEGM is being reinforced on a systematic basis. For example, the top of the ventilation stack is being equipped with dampers to mitigate the vibration caused by earthquakes.

To further reassure the general population about the safety of Monju, even in extreme circumstances, JAEA published a collection of possible accident scenarios, an analysis of their potential impact, how they might be dealt with, and the routine actions that prevent them from occurring. The booklet, ‘For the safety of Monju – how do we respond to the potential accidents and troubles?’ was also published in English in March 2010.

A political boost for the project came in October 2005, when the Japan Atomic Energy Commission published its Framework for Nuclear Energy Policy [5]. The document, which had taken five years to develop, strongly supported fast-breeders and Monju. It said: “The major task at this stage is the research and development of FBR and its fuel cycle technology. As FBR and its fuel cycle technology has the potential of contributing to long-term energy security and reduction of the effects of potential toxicity of radioactive waste, it is necessary to consistently promote the research and development toward its commercialization led by the Japan Atomic Energy Agency, while fully applying lessons learned from past experiences. Specifically, the operation of Monju, the core of the place for its research and development, should be resumed at the earliest possible time, and the priority should be placed on achieving the initial goals of demonstrating reliability as an operational power plant and establishing sodium handling technology, hopefully, within ten years or so. After that, we expect Monju to be utilized as a location for research and development activities toward commercialization of FBR and its fuel cycle technology, in cooperation with the research and development activities of fuel production and reprocessing technology, based on the prospect of offering fast neutrons to relevant research and development activities.”

In 2006, the third science and technology basic plan of Japan was published. Five sciences and technologies of national importance were chosen; the FBR cycle technology was selected, the only one in the energy field. Research and development of Monju is expected to play a key role.

Future plans

Near-term R&D plans at Monju include validation of the reliability of the plant’s operational safety and establishment of sodium handling technology. Plant inspection and repair techniques will be developed as a FBR-specific maintenance technology. These will be essential for the design and operation of future demonstration and commercialized fast reactors (FRs). FRs use high-temperature but low-pressure systems, while light water reactors use low-temperature but high-pressure systems. As a result, unlike LWRs, FRs never lose coolant rapidly in the case of leakage. But FRs should take account of the thermal fatigue caused by the larger difference of temperature in the plant. This is a basic requirement for FR maintenance technologies.

In-service inspection (ISI) devices are under development in Monju to confirm the soundness of reactor vessel, steam generator heat transport tubes, and primary sodium cooling loops. The devices will demonstrate their performances in Monju in a high-temperature and high-radiation dose environment before the first periodic maintenance outage expected in 2014. The results will be used in the development of ISI devices for the demonstration plant and succeeding plants.

Inspection, monitoring, aging diagnosis, repair techniques, and soundness evaluation technology will be developed and validated in Monju to ensure plant safety and realize rational maintenance. These technologies will be reflected back into the specifications and standards of the future FRs.

A research institute of nuclear engineering (RINE) attached to Fukui University, near Monju, was established in April 2009. RINE features FBR engineering. It is expected to cooperate with Monju.

In the longer term, the reactor will also be able to provide interesting data about the minor actinide Am-241. The fuel loaded into Monju was mixed plutonium-uranium oxide with fissile plutonium enrichment around 15-20%. By 2009, through natural radioactive decay, the fuel had only half of the Pu-241 it started with. Such a low concentration of Pu-241 makes achieving criticality impossible, so new fuel was loaded.

But the Am-241-dominated fuel is an important research asset; its neutronics characteristics are very hard to find elsewhere. To design and operate reactors, core physicists use calculation codes, which rely on real-world nuclear data, such as the cross sections for fission, neutron capture, and neutron scattering. The accuracy of the calculation code depends mostly on the nuclear data and calculation methods. To validate the calculation accuracy of a core containing Am-241, experiments are usually conducted in critical assemblies with very high Pu-239 concentration of over 90%. But in the actual core, the Pu-239 content is not very high, and the effect of Am-241 generated from the beta decay of Pu-241 must be taken into account. The Monju core has a sufficiently high Am-241 concentration to validate the accuracy of the calculation codes and existing nuclear data.

JAEA plans post irradiation examinations (PIEs) on Monju spent fuels, including confirmation of composition change in the fuel pellet. Detailed information will be provided through PIEs on the burning characteristics of Pu-241 in FBRs.

With cooperation with France and the USA, the global actinide cycle international demonstration (GACID) project intends to demonstrate the burning of minor actinides including neptunium, americium and curium in Monju to develop the future FBR fuels.

Author Info:

This article is based on a presentation given by Tsutomu Yanagisawa, senior advisor of Japan Atomic Energy Agency, to the general meeting of the Japan Society of Maintenology and published in online journal E-JAM: Additional writing and research by Will Dalrymple, with support from JAEA’s T. Kitabata and M. Tsuzuki

Related Articles


[1] Miyakawa A., Maeda H., Kani Y., Ito K., “Sodium leakage experience at the prototype FBR Monju,” IAEA Technical committee meeting on unusual occurrences during LMFR operation, Vienna (Austria), 9-13 Nov 1998, IAEA-TECDOC—1180, pp49-62,
[2] The Japan Times, ‘Court rejects lawsuit over Monju coverup suicide’ Tuesday, May 15, 2007,
[3] ‘Restarting Monju – Like Playing Russian Roulette’ Citizens Nuclear Information Center,
[4] Regulatory Guide for Reviewing Seismic Design of Nuclear Power Rector Facilities, Nuclear Safety Commission of Japan (NSCJ), September 2006,
[5] Framework for nuclear energy policy, Japan Atomic Energy Agency, October 2005,


Timeline: significant events in the Monju restart
Fast reactors around the world
Table 1: Monju main data