Abstract
A system to diagnose abnormalities on structural materials in the primary cooling systems of BWR by analyzing water chemistry data is proposed. In the system, patterns composed of more than two kinds of water chemistry data are compared with several stages of reference patterns obtained by the models which relate the water quality changes to the structural material abnormalities. Small abnormalities on the materials canbe expected to be detected at very early stages and future abnormalities can be predicted by the diagnosis system.
In this paper, the basic structures of the anomaly detection system using water chemistry data are explained.