Maintaining the integrity of wells in the oil and gas industry is of critical importance to ensure operational safety, environmental protection, and cost-effective production. One very important aspect in terms of the pipe integrity status in a well is casing deformation, which can be caused by formation swelling, formation subsidence, tectonic activity, salt creep, corrosion-induced issues, and thermal stresses. Some other conditions include completion defects or operation envelope. Evaluation of casing deformation is therefore an important requirement for well integrity management in these conditions; however, existing logging technologies require the removal of tubing in the well in order to evaluate the casing behind it. Tubing retrieval is a time-consuming and costly operation, and through-tubing casing inspection is a highly desired service for reducing these interventions and making casing deformation monitoring more accessible and routine. Nevertheless, through-tubing casing inspection with conventional electromagnetic technology encounters significant challenges in two principal areas. The first challenge is the tubing shielding effect, where only approximately 20% of the magnetic flux density reaches the casing area behind the tubing, resulting in little casing response to provide a reliable analysis. The second challenge is the tubing position inside the casing, which could be eccentric and variable along the zone of interest, introducing substantial uncertainties in the processing of the logging data.
A novel approach is proposed in this paper in which data-driven modeling and the ensemble Kalman filter are utilized to address these challenges. Recognizing the dynamic nature of the system and the influence of tubing shielding on the casing, an enhanced dynamic mode decomposition (DMD) based on Koopman operators was developed to manage these conditions. This approach treats the measurement process as a dynamic system, accounting for the casing deformation change within the well as a state change. In the ensemble Kalman filter, an observation model was built with the training data, which is acquired based on finite element simulation results. The simulation model considers different wellbore scenarios and logging conditions. The variable parameters include casing and tubing diameters, weights and location in the well, tool position, along electronic parameters. Moreover, the ensemble Kalman filter considers the physical constraints inherent in through-tubing casing inspection. This approach merges Koopman-based dynamical system learning with physical constraints as an innovative approach to overcome the tubing shielding effect in order to effectively improve the accuracy of through-tubing casing inspection. Notably, the solution quantitatively estimates tubing eccentricity, allowing for the removal of its nonlinear effects, even in extreme cases of eccentricity where the tubing touches the casing.
The performance of this tool and the processing advancements have been validated both in laboratory setup and in the field, across various well conditions and casing/tubing combinations. In the simulated through-tubing assessments, remarkable results of an estimated 5% deformation ratio accuracy for casing diameters up to 13-3/8" were achieved. Applications in the energy sector for such techniques encompass through-tubing well integrity monitoring for production, injection, gas storage, and geothermal wells, stuck pipe diagnostics without pipe manipulation, pipe eccentricity assessment for plug and abandon operations, and tubing clamp orientation detection for control line positioning analysis. Measurements from this through-tubing casing deformation technology can be integrated with other conventional single or multi-barrier pipe inspection logging tools, such as multi-finger calipers and electromagnetic multi-pipe thickness tools, to provide a comprehensive pipe integrity evaluation solution.
