High resolution Magnetic Flux Leakage (MFL) inspection tools are very adept at showing the current static state of corrosion in a pipeline. A single inspection shows only a partial reflection of the actual corrosion activity. Multiple inspections and an accurate corrosion growth analysis are required to have a more complete understanding of the corrosion activity. It should then be possible to find features growing at a potentially critical rate and to differentiate between areas of new growth and areas of continued growth. The principal advantage of this is to more accurately predict future corrosion growth in a line. This paper is a highlight of two corrosion growth analyses and a method that has been developed to provide a more complete understanding of corrosion activity and growth. The method that has been developed enables a minimization of the primary errors involved in a corrosion growth analysis which has been through the usage of the raw signal.
The knowledge of the rate at which corrosion grows in a given pipeline is key to determining the optimal time between MFL inspections, finding hot spots of high corrosion growth, and possibly preventing catastrophic failure of the pipeline. BJ Pipeline Inspection Services inspected a 36 inch pipeline in 2003 for TransCanada PipeLines (TCPL). This inspection was performed five years after a previous inspection of the same pipeline. At the request of the pipeline operator, a corrosion growth analysis comparing the two inspections was carried out. The first calculation method that was utilized to estimate corrosion growth rates involved analysing the original reported box data (length, width, depth) for each inspection. Even though both sets of data were accurate [1 ], it was found that the results from this method were dependent on sizing methods and models that had been changed due to improvements over time. The second calculation method involved comparing certain aspects of the raw signal data collected by the MFL tool. Utilizing the raw MFL signal data between inspections has led to improvements in accuracy and has dealt with many of the short falls of the simple 'box' matching method. This method significantly reduces errors in the analysis and brings this type of calculation to a higher level of accuracy and reliability. To date, 300 km of pipeline have undergone a growth analysis involving 36 and 42 inch pipelines. Corrosion growth rates have been calculated for the complete dataset totalling over 90,000 features.