Sampling and Risk Assesment Method for Metallurgical Inspection of Pipes and Tanks

This paper gives a general overview of a method developed to optimised NDT inspection planning and results analysis. This computerized tool offers sample size calculation, statistical analysis of thickness measurements and life prediction modules. In each step and module, calculations are driven by equations developed using data gathered from previous expertise and laboratory work. The equations are updated when new case studies are added to the system data base. The method was initially developed for radiographic and ultrasonic inspection of corroded pipes. It can be described in eight steps: 1) Assessment of degradation severity 2) Determination of inspection method 3) Initial sample size calculation 4) Field testing 5) Results analysis 6) Probability and reliability calculations 7) Predictions and degradation pattern 8) Results presentation. Steps 4 to 7 can be recomputed to get a better confidence index. Calculation and results presentation are customized based on a questionnaire filled by the project engineer. Introduction: Metallurgical, chemical, petrochemical, energetic and marine industries are using different types of pipes and tanks to carry and stock gaseous, liquid and solid products. These pipes and tanks are subjected to internal and external degradation caused by erosion and corrosion. For certain types of installation like pressurized tanks and pipelines, there exists regulation about inspection frequencies and minimum acceptable wall thickness [1-4]. In other cases, degradation control and inspection parameters are left to the goodwill of the installation’s owner. Non-destructive techniques (NDT) such as radiography (RT), ultrasound (UT) and Eddy current (ET) are widely spread inspection methods used to monitor pipes and tanks degradation [2,3]. They all give information about wall thickness and about the existence of metallurgical flaws. However, inspection is never full proof. Generally, the installations are inspected partially and the methods have limitations. Since corrosion and erosion of metals are rarely uniform, there exists a risk associated with recommendations and predictions based on partial inspections. This risk is related to sampling size, corrosion mode and inspection method. The objective of this paper is to give an overlook on the corrosion expert system developed by X-per-X inc. research group. This fully computerized diagnostic tool has been designed to fulfil two needs of the company’s NDT group. The system optimises sampling according to the corrosion severity and a needed confidence level. It also calculates the probability of failure a number of years subsequently to inspection. This paper is divided in three sections. Section 1 gives an overview of the system’s philosophy. Section 2 details the steps and calculations leading to a preliminary sample size. A description of the method used to analyse the inspection data is given in section 3. Finally, section 4 is a summary of the type of predictions and results presentation offered by the system. System overview: To design the system, the initial train of thoughts was decomposed in nine steps regrouped in three modules (Fig. 1). 1) Corrosion severity estimation 2) Choice of NDT 0) Information gathering 3) Sampling calculation MODULE 1 7) Prediction and risk evaluation 8) Report presentation 4) On-site inspection