The safety of nuclear power plants is based on the principle of defense in depth: multiple layers of protection, or lines of defense, already present in the design of the facility. One of the major safety objectives of nuclear facilities is thus to contain radioactivity in all circumstances.
In a pressurized water reactor, the first containment barrier is the cladding around the fuel rods (fissile materials) which retains the radioactive products created in the fuel pellets. Poor heat removal or mechanical stress that is too great may result in cladding failure, and more or less significant damage to the fuel pellets. IRSN carries out R&D programs to understand the behavior of the first containment barrier in the various accident situations that could occur in a PWR.
Reactivity injection accident
In a pressurized water reactor, the envelope accident retained for the design basis to simulate a sudden and instantaneous increase in power, known as a power transient, typical of a reactivity-initiated accident (RIA), is ejection of a control rod assembly. An assembly consists of absorber rods which participate in controlling the nuclear reaction.
In the event of control rod drive mechanism failure, the control rod ejection accident is due to the difference in pressure between the reactor coolant system (155 bar) and the containment (at atmospheric pressure). This violent ejection causes a local runaway of the nuclear reaction for several tens of milliseconds (power pulse), leading to a rapid increase in fuel temperature. Neutron feedback limits the power transient before the reactor trip (intact control rod drop), which occurs in a second phase. The study of the short term phase of the accident, which is the goal of the Cabri tests, dimensions the limit insertions of the control rod clusters in the core, which depend on the reactor power level.
In addition to the experimental programs conducted in the Cabri reactor, IRSN performs a detailed analysis of physical phenomena that occur during the various phases of the reactivity accident by interpreting the contents of the international experimental database, which consists of the results of in-reactor tests (NSSR in Japan, BIGR in Russia, etc.) and analytical tests on cladding behavior (PROMETRA) and heat transfer (Patricia).
At the same time, IRSN carries out studies further upstream with the academic world by supporting doctoral research in France. The objectives of this research based on advanced modeling and smaller-scale experiments are to better understand elementary phenomena that occur during an RIA and to offer relevant models.
An experimental device developed with IMFT in Toulouse is used to study heat transfer between a wall simulating a fuel rod and a flow simulating water in the core, especially for very sudden heatups that initiate boiling and are characteristic of an RIA. Another experiment designed with IUSTI in Marseille characterizes the ejection speed of fragments simulating nuclear fuel outside a pressurized cylinder simulating a fuel rod failure.
Two other experiments with INSA in Lyon offer information for understanding the violent heat interaction between a hot body, simulating a fuel fragment, and a surrounding fluid, as well as the thermal mechanical behavior of cladding in post-DNB conditions, i.e., during very rapid temperature transients.
For modeling, laws concerning the mechanical behavior of fuel and cladding in temperature and stress conditions that characterize an RIA situation are being designed at the joint laboratory of IRSN, Montpellier University and CNRS-LMGC.
Lastly, laws of mechanical behavior specific to MOX fuel are currently being worked out with LMA in Marseille.
All of the acquired knowledge is then entered in IRSN’s SCANAIR software, which simulates the thermal mechanical behavior of a fuel rod during a reactivity insertion accident.
In a pressurized water reactor, a loss-of-coolant accident (LOCA) is caused by a rupture of the reactor coolant system, which causes a drop in reactor coolant system pressure and lowers the water inventory in this system. The resulting heatup of the fuel rods must be limited so that fuel damage does not jeopardize cooling of the reactor core, thus avoiding core melt. A LOCA is used for the design basis for the Safety Injection System and the reactor containment. A LOCA also results in loads on the reactor containment (pressure and temperature) which must stay within certain limits to ensure containment and the continued operation of equipment.
As part of programs funded by the French National Research Agency (ANR), IRSN oversees the PERFROI project to characterize experimentally the mechanical behavior of cladding during a LOCA and study the possibilities for cooling a deformed fuel assembly (rod ballooning) during LOCA.
In addition, IRSN participates in several international research programs whose goal is to understand fuel rod behavior during a LOCA situation (cladding embrittlement and deformation, fuel fragmentation, etc.). The main programs, carried under the auspices of the OECD, are the Halden Reactor Project in Norway and the Studsvik Cladding Integrity Project (SCIP) in Sweden.
At the same time, the knowledge gained during this research is added to DRACCAR, a thermo-mechanical multi-rods 3D software to realistically asses the behavior of irradiated fuel rods during a LOCA.
LOCA in spent fuel pools
Before and after use, nuclear fuel is stored in the spent fuel pool in the fuel building located next to the reactor building. Its main role is to store used fuel until its residual power is sufficiently low and it can be definitively removed from the site. During plant outages, new or already radiated fuel is placed here prior to loading in the reactor.
The spent fuel pools have a volume on the order of 1,000 m3 and a depth of 12 meters. The fuel assemblies are located at the bottom of the pool and covered with approximately eight meters of water. A system equipped with pumps and heat exchangers keeps the temperature of the pool below 50°C. It is commonly accepted that in the event of loss of pool cooling, the temperature of the pool will increase to boiling and the water level will decrease, potentially until the assemblies are uncovered after several days. To avoid uncovery of the assemblies, the procedure is to compensate water lost through evaporation with a water makeup, after opening the building to remove steam. The procedure is implemented until a cooling means is found.
The Fukushima accident nevertheless confirmed the importance of examining all possible strategies to ensure cooling of the spent fuel pools. The issues related to loss of cooling or any accident involving uncovery in spent fuel pools have shown the need for additional knowledge and understanding of the phenomena of these accidents.
For these reasons, since late 2012 IRSN has overseen the ambitious DENOPI research program as part of post-Fukushima projects cofunded by the French National Research Agency (ANR) to learn more about core uncovery in the spent fuel pool during an accident. It consists in performing experiments, modeling and validating calculation software. Its goal is to improve knowledge about physical phenomena during accidents in the spent fuel pool, as well as prevention or mitigation measures that may be implemented for these accidents.