Testing Electromagnetic Multi-Pipe Thickness Tools in Azimuthally Constrained Defects for Deepwater Applications

Abstract

Abstract Tubular corrosion in high water-cut wells has long been an industry concern, typically accounted for during the initial completion design phase. Direct evaluation of metal loss in production tubing is an established practice, with a range of tools on the market to deliver accurate azimuthal measurements. Some of these same techniques can be employed to evaluate casing through tubing but have reduced resolution and are currently omnidirectional in their measurement. A yard test was performed with several electromagnetic wireline tools to understand the lower detection limits of the tools and their overall accuracy when logged through tubulars containing multiple azimuthally constrained engineered defects. For the purposes of testing, it was supposed that spatial corrosion in tubulars can occur either homogenously or localized. Detecting pitting through pipe has historically been more difficult due to tool resolution. A tubular could theoretically contain severe localized corrosion, but the electromagnetic tools could measure little to no corrosion since bulk metal volume loss is low. Defects were machined into several joints of casing; some in the form of general defects (pipe OD reduction) and others in the form of localized defects (drilled features on pipe OD). Chrome tubing (25%) was inserted into the machined string of casing, with control (non-defect) casing on either end. Four logging tools were selected to participate in the test. All four tools identified all machined defects in the casing. Generally, the quantitative analysis of the metal loss was conservative – the measured wall thickness was less than the actual machined defect wall thickness. Metal loss measurements were generally more accurate in measuring the general defects metal loss than the localized defects. Defects with less metal loss were generally better resolved than those with larger metal losses. The test ultimately illustrated that azimuthally constrained defects (not 360-deg circumferential defects) can still be identified with all four tools, even with only 5% bulk metal loss (over a 5-foot interval), but the tools were not always accurate in their measurements of absolute metal loss. Overall, this controlled test reaffirmed that electromagnetic measurements are more appropriate for use as metal loss identification tools than wall thickness measurements to be relied upon quantitatively when dealing with real world scenarios. The size and shape of the potential corrosion or pipe defect was shown to be significant; different defect types can lead to the same interpreted wall loss.