The characterization requirements used for an MFL inspection depend on the pipeline operator's goals in performing the inspection. As discussed earlier, some operators are interested in identifying locations where imperfections or defects are forming. In regions where the coating is damaged, for example, water and soil conditions can lead to corrosion defects. Other operators are interested in identifying critical defects only, and there is less emphasis on locating small defects and imperfections. Characterization accuracy is important in differentiating between defects, imperfections, pipeline components, and non-relevant indications.

Detection Threshold Requirements shows a typical boundary between defects and imperfections. Similar curves can be used to assess how sensitive an anomaly's severity is to errors in length and depth. Understanding the severity of a metal-loss region as a function of length and depth is important in understanding when accuracy in length and depth is needed.

Sensitivity to Inspection Errors
(36-inch-diameter x 0.375-inch wall thickness X60)

Shown above several curves of defects that fail at different pressures. The bottom curve is for defects that fail at 100 percent of the yield pressure; the middle curve is for defects that fail at 95 percent of the yield pressure, and the top curve is for defects that fail at 90 percent of the yield pressure. As before, the ordinate (x-axis) is the length of the defect and the abscissa (y-axis) is the depth of the defect.

Length accuracy is important when an error in length causes a significant error in estimating the severity of an anomaly. For a 30-percent deep 14-inch long defect above, a two-inch error in length does not appreciably change the failure pressure. So, an error in length will not significantly affect the calculated severity.

For deeper defects, errors in length become increasingly more important. For a 60-percent deep 4-inch long defect, an error of two inches leads to a much larger error in the severity. So, length accuracy is more important for short deep defects than for long shallow defects.

Most inspection tools have good accuracy in estimating the length of individual defects. Difficulties arise, however, when several defects are in close proximity. In inspection reports, many vendors will "group" individual defects together as a composite defect; that is, two or more defects are reported as a single defect. This practice can be very conservative, especially when several deep defects are grouped.

Most integrity analysis methods allow corrosion pits (short deep defects) to be treated as individual pits when their separation is quite small. The pits can be treated as individual defects when the separation is greater than, for example, three times the wall thickness or one inch. Clearly, reporting four 2-inch long defects as one 8-inch long defect will cause serious errors estimated severity.

Depth accuracy is important when an error in depth causes a significant error in estimating defect severity. For the two defects shown above, the effect of an error in depth is similar: a 10-percent error in depth causes an error of about 5 percent in the estimated severity of each defect. So, while length accuracy is most important for short deep defects, depth accuracy is important for all defects.

Characterization accuracy is also a function of the number of inspection indications found on a line. On a line with few indications, a pipeline operator can investigate all indications. Here, detection may be of paramount importance, while characterization is less important. On a line with many indications, characterization accuracy is more important. Similarly, characterization can be critical if an indication cannot be easily or independently investigated.