An Ultrasonic Fuel Identification System for Liquid Metal Cooled Reactors Resilient Against Multiple Transducer Failures

We describe a fuel assembly identification system developed for the MYRRHA reactor-a new multi-purpose flexible irradiation facility to replace the aging BR2. MYRRHA is a fast spectrum research reactor cooled with lead-bismuth eutectic (LBE) and conceived as an accelerator driven system capable of operating in sub-critical and critical modes. As liquid metal is opaque to visual light, the conventional optical fuel assembly identification system, as used by water cooled reactors, has to be replaced by a system not hindered by the opacity of the coolant. As already suggested in the late sixties, we use ultrasound for this purpose and present an encoding especially designed to enhance the robustness of the ultrasonic read-out. The encoding is based on notches of varying depth on the inflow nozzle of a fuel assembly. The depth of each notch is used to encode two bits and is measured by a dedicated transducer aligned over the notch. To increase the reliability of the fuel identification process, the identification number is protected by an error correcting code based on Hamming codes. We describe the ultrasonic system used to read out the vector of depths which is subsequently converted to a vector of bits. We explain the encoding of the twelve bit fuel identification numbers to a 22-bit error correcting code and discuss how Hamming decoding can be used to correct single bit errors, detect two bit errors or fill in the missing bits of a failing transducer. We shortly show how the confidence on the individual measurements can be taken into account using a Gaussian measurement error distribution assumption. We also present a method based on solving a linear system over Boolean variables to (partially) reconstruct the fuel identification number in case multiple transducers fail. We show that the probability on full reconstruction is 100% for up to two transducer failures, 98% for three, 79% for four and 20% for five failing transducers. Finally, we prese- t validation results in water and lead-bismuth eutectic for the differential measurement method used to measure the depth of the notches which form the basis for the requirements of the final system which will be installed on a robotic fuel manipulator.