Classification of atmospheric aggressiveness in Colombia using multivariate statistical methods

Abstract

Studies of atmospheric corrosion often require a classification of environments based on their aggressiveness, which involves the processing of a large amount of data related to weather and pollution variables. In this work, a classification of different Colombian environments, under which the main power transmission lines are placed, was made based on weather data and using the cluster analysis method. Based on this classification, 21 representative sites were chosen to perform field measurements of corrosionof steel and galvanized steel, deposition of SO2 and chloride, temperature and time of wetness. With this information, atmospheres were classified again and correlations were established between atmospheric variables, exposure time and corrosion, using regression models and multivariate statistical methods.