Statistical modelling of external corrosion in buried piping

The assessment of external corrosion of buried piping is a major challenge due to access restrictions. This study utilizes field data collected by the U.S. National Institute of Standards and Technology (NIST) to model the external corrosion rate using a weighted least squares (WLS) multiple regression approach. The key parameters in the model include soil resistivity, pH, moisture equivalent, and the time of exposure. The model can be used to predict the corrosion rate and probability of leak over time, and hence directly support the risk-informed decision making and management of buried piping at nuclear power plants. (author)