Using Statistic Method to Calculating and Monitoring the Corrosion Rate of Petroleum Process Piping

Abstract

In chemical plants and most refineries, some process piping is suffered from metal thinning. Take carbon steel piping for example, each component such as elbow, tee, direct piping, branches, reducers are suffered thinning no matter the piping is still in using or not. So, with an eye to personal safety and smooth operation held in the chemical plants and refineries, every effort put in the inspections should be correctly getting knowledge about the deterioration conditions. Normally, to check the piping thinning condition is to perform the ultrasonic thickness spot gauging at each component of the piping. Through the thickness reading, inspectors could get the information about the piping remaining thickness, calculate the corrosion rate, and evaluate the remaining life at each part of the piping spools. However, when one puts an eye on the piping inspection management and deterioration conditions about the piping system in the operating unit, the variation condition of the inspection thickness data, in some view points, always play an important part as corrosion information hidden in the thickness data group. Besides, by considering the variation condition of the thickness data, the inspectors can get a more conservative and satisfactory corrosion rate data about the whole piping spool. By plotting the histogram of the corrosion data; inspectors can get a more clear idea about the deterioration condition of the whole operating unit. In this paper, the ultrasonic thickness survey is reported from the pre-selected susceptible areas of deterioration such as the injection points, elbow, tee etc. (as suggested by the API 570 Code and the API recommended practice 574). After collecting the thickness readings, the statistic skill is used to evaluate the corrosion rate of each piping spool in the operating unit. After the statistic calculations, the inspectors can analyze the evaluated corrosion rate data by plotting the histogram of the evaluated upper bound corrosion rate values. With the trend shown in the histogram, the inspectors can easily get knowledge about the deteriorating condition of the operating unit. Finally, in order to discuss the feasibility of the proposed method, the inspection data of two piping spools will be used as the examples. Also, a site survey result performed in the lubricant rerun unit shows the effectiveness of this statistic calculation method.