Several attempts have been made by different companies and research institutes to compare the field data with the predicted internal corrosion rate/corrosivity but the matching of the actual and predicted corrosion rates has not been successful in many cases. In fact, some of the important parameters collected from fields or obtained in laboratories do not have the same meaning as the data used in the prediction models. Many of the conditions existing in fields can not be produced in laboratories. The accuracy of the field data is also questionable in many cases and if this is the case, there is a risk of considering an accurate prediction as wrong. There is lack of knowledge for the impact of some important parameters, like small amount of H2S and organic acids, on the corrosion rate measured in the field. The mechanism of the localised corrosion does not allow the prediction of an accurate corrosion rate. The paper analyses in details the accuracy of some production and corrosion data collected from fields, explains why the fact that prediction models give different results is normal and the prediction of an accurate corrosion rate is actually neither possible nor required. It is suggested that the oil industry concentrates its efforts to predict properly the influence of important parameters on the corrosivity, improve the validity of the data and to develop an approach to predict low corrosivity, medium corrosivity and high corrosivity.
Key words: intemal corrosion, corrosion rate prediction, prediction models