Defected fuel rods identification in TRIGA Reactors: The experience at the ENEA Casaccia TRIGA RC-1 reactor

Abstract

Experience in running TRIGA reactors all around the world has shown that fuel cladding failures can occur. Fission products, especially in gaseous physical form, can exit the defected fuel rods being dispersed within the water primary coolant. Suspects of a cladding failure event can be confirmed by detection of short-lived fission products, i.e., Krypton, Xenon, and even Iodine isotopes in primary water, or within the ionic-exchange resins tank installed for the purification of the primary coolant loop. The magnitude of the release of those ‘key-indicators’ from the defected fuel element(s) is the driving method by which the defected rod(s) can be identified. ‘Significant’ releases can be detected with a direct online sampling of the water from the top of the suspected fuel rod with reactor at power, leading this water to an online high-resolution gamma spectrometry analysis system. By considering delays due to the water velocity in tubes and decay time of radionuclides identified within the gamma spectrum, it is possible to calculate the concentration of those radionuclides just emerging from the inspected fuel rod. Releases lower than the minimum detection capabilities of the previous online experimental configuration push to modify the detection method with an indirect identification of the release. This is the case when the ‘normal’ radioactive background of the activated water, when reactor is on power, is the dominant component in the gamma spectrum of the sampled water, and fission gases (even produced) are not identified promptly, i.e. a relationship to a specific fuel rod by the sampling circuit before could not be identified. The paper describes the experience carried out at the Italian ENEA TRIGA RC-1 reactor, deepening the technical aspects and solutions applied to solve the issue. In particular, a ‘significant’ release has been found in the instrumented fuel rod within ring B, i.e. the inner fuel ring, exposed to the maximum neutron flux of the reactor. The leaking element was found within a week, in two days of operation, being the sampling system designed on detection of minutes-shortlived fission gases (no need to ‘cool’ down the primary loop by waiting decay of hours-lived radionuclides). After the removal of the instrumented fuel rod in ring B, further days in searching other ‘significant’ release with the sampling circuit before have reported nothing detected. But in samples of water taken after reactor shutdown, some iodine emerged after decaying of ‘normal’ radioactive background of the activated water. This was sufficient evidence of another defected fuel rod in the pool. Identification of the latter defected fuel element is still ongoing, being based on: 1) selective removal of rod(s) from the reactor core, 2) run power of the reactor, 3) take samples of water after shutdown and measure iodine after decaying of ‘normal’ radioactive-background of the activated water, 4) identify the rod responsible of the remaining leakage when no further iodine is detected.