Uso do Método Box-Behnken para diminuição de incerteza de medição da técnica de inspeção por PIG MFL

Abstract

PIG MGL (Pipeline Inspection Gauge Magnetic Flux Leakage) disclosure is one of the most used techniques in the disclosure of pipelines to detect internal and external losses, caused by the corrosion phenomenon. Although the technique is widespread in the industry, it brings with it a high definition uncertainty, causing some defects to be over-dimensioned or under-dimensioned. This work analyzes some parameters intrinsic to the MFL technique, such as: PIG movement speed, magnetization level and wave amplitude of the MFL signal, using the Box-Behnken statistical method to expose the factors that most influence the MFL tool agreement uncertainty. The study was carried out in loco with 17, using an ultrasound conduction technique to assess a real condition of metal losses. It was observed that the parameter with the greatest influence on the measurement uncertainty is a wave amplitude, however, a combination of the three previous elements, special a better response in determining the definition uncertainty. A significant factor is that the wave amplitude expresses different geometries for losses of processing thicknesses in the tube body, if compared with losses in the TAZ (Thermally Affected Zone). This can be explained due to the modification of the crystalline structure of the material in the TAZ. In this way, pipeline monitoring and maintenance inspection routines, based on the combination of the three agreement parameters, will be able to prevent damage from spreading and improve the reliability of the entire pipeline and pipe network.