The Risk-Based Analysis And Management Of Thickness Measurement In Refining Enterprise

This topic studies on the static equipment(pressure vessels.piping,etc.) which is closely related to security of petrochemical production equipment by combining thickness measurement technology with risk assessment.A set of scientific corrosion management system is formed to establish a self-organizing system of the user, inspection program management,corrosion management and supervision departments by combining corrosion monitoring and risk based inspection (RBI).Two methods was used for corrosion analysis in this paper. The corrosion rate prediction model of BP neural network was established by matlab with the pipe material and the process parameters. Its input layer consisted of five nodes:sulfur content of the medium, pressure, acid content, velocity and temperature of the gas. The output layer was the corrosion rate. After repeated calculations and iterative,the result which can predict the average corrosion rate of the pipeline was best with eight hidden layer.The other method was to calculate average corrosion rate,the short rate,the long rate,and estimate residual life and the next inspection time by using the data obtained by the ultrasonic thickness measurement.Then the calculation results by PCMS and the theoretical calculation results were compared to verify the correctness of the calculation results.In this paper, failure probability of RBI was calculated by the corrosion rate weighted by the actual corrosion rate and the predicted corrosion rate. According to the risk level, inspection and maintenance program was constituted and the ultrasonic thickness distribution was corrected.The research methods was used in sulfur recovery unit of China Maoming Petrochemical to verify the effectiveness of the method.