Casing Wear Estimation Without a Baseline Log - A Distorted Ellipse Methodology

Abstract

The estimation of casing wear from USIT or Caliper logs requires a proper evaluation of the shape of the tubular inner surface prior to casing wear. In the absence of a viable baseline log, this shape must be deduced through analytical means. Manufacturing deviations from perfect circularity result in eccentricity and ovality, so that the pre-wear shape is in general, elliptical in nature. Owing to the fact that a standard ellipse may not be adequate to capture this shape in all situations, an approach using a distorted and shifted ellipse is proposed. The shift from the geometrical centerline of the ellipse co-ordinates in order to capture the effect of improper centralisation of the logging tool. The approach uses a least-squares minimisation technique to determine the distorted ellipse parameters such that the overall radial discrepancies between the pre- and post-wear shapes are minimised. A multi-pass approach wherein detected wear grooves are censored results in better predictions of the pre-wear shape as was determined from several validation test cases with known wear groove depths. The methodology was also applied to an actual USIT log with over 20,000 stations with 72 radial measurements at each station. The methodology also enables the detection of true locations of high wear from spurious ones. As proposed, the method is also capable of using either internal radius or wall thickness as the evaluation metric, and is stable in the presence of significant noise in the measurements.