Since the market launch of Ultrasonic crack detection tools, the conventional crack depth sizing is based on four depth classes or buckets. A more differentiated, continuous depth sizing is becoming increasingly relevant for the pipeline operators and especially for pipelines with large populations of planar anomalies (SCC colonies, lack-of-fusion in ERW seam-welds, etc.). The ILI industry is introducing a continuous crack depth sizing. Next to the better differentiation and the linearity of the depth reporting, the main advantage of the continuous depth sizing is the direct comparability to the results of the field verifications. The continuous depth sizing improves the ability to assess the performance validation of the depth sizing and thus, contributes to a general improvement of the crack depth sizing. This paper describes the development and implementation of a continuous crack depth sizing approach and shows its advantages in comparison to the conventional depth classes. A sizing model is introduced, making use of an empirically derived function, that relates the amplitude measurement to the defect depth. The continuous depth sizing applies to crack-like defects with depths ranging from 1mm to 4mm. The parameters of the model are derived from performance tests based on artificial flaws. In addition, the model is validated by means of field verification results. The depth sizing accuracy and confidence levels are obtained from the performance test data in accordance to API 1163  and POF 2009 .
In addition, the paper discusses the extraction of the crack depth profiles from inspection data, making use of the newly developed continuous depth sizing model. In comparison to standard reporting of maximum depth and length, crack depth profiles deliver more accurate and more valuable input to the integrity assessment for pipeline operators. Examples of a direct comparison of these crack depth profiles to field verification data are included.