Framatome Gmbh

Mathieu Segond

Christian Royere

A method for operating a nuclear reactor comprises acquiring a plurality of quantities characterizing the operation of the nuclear reactor; and calculating at least one Departure from Nucleate Boiling Ratio using a deep neural network. The entries of the deep neural network are determined by using the acquired quantities. The deep neural network includes at least two hidden layers of at least five neurons each. The method further includes calculating the deviations between the at least one calculated Departure from Nucleate Boiling Ratio and a plurality of predetermined reference threshold values and formulating a control signal for a reactor control system by using the calculated deviations. The control signal is an automatic reactor shutdown or alarm. The method also includes emergency shutdown of the nuclear reactor or emission of an alarm signal if relevant.

Method for operating a nuclear reactor using calculation of the departure from nucleate boiling ratio

A method for protecting a nuclear reactor includes reconstructing a maximum linear power density released among the fuel rods of the nuclear fuel assemblies of the core; calculating the thermomechanical state and the burnup fraction of the rods; calculating a mechanical stress or deformation energy density in the cladding of one of the rods by using the said reconstructed maximum linear power density, the calculated thermomechanical states and the calculated burnup fractions, by means of a meta-model of a thermomechanical code; comparing the calculated mechanical stress or the calculated deformation energy density with a respective threshold; and stopping the nuclear reactor if the calculated mechanical stress or the calculated deformation energy density exceeds the respective threshold.

Growing community of inventors

PARIS, France

Mathieu Segond