A hybrid model based on Physics of failure and Data-driven algorithms is developed that can estimate remaining useful life of production casing (well barrier). The state of integrity of the well barrier is assessed by updating the reliability under operational loads. The interactions between the casing and surrounding formation, and effects of tribocorrosion on the casing are considered. Tribocorrosion is the process of degradation of a material resulting from a sequential process of (i) mechanical wear (due to sliding, friction, or impact) followed by (ii) a corrosive action of the surrounding environment. The model includes simulating casing wear due to drilling, and enhanced degradation due to conditions in the well. The main capability of the model is to help well integrity analyst with insight of future health states of a monitored well. This is achieved in two main steps; the first being the offline module comprised of degradation models. The second is the pattern recognition based on well log and features mapping, and estimation of remaining useful life of well barrier. The production casing grade P-110 undergo reduction in strength due to wear during drilling, induced stress and hydrogen induced cracking. The remaining useful life is calculated for the depths of interest and time. A comparative analysis is carried out using the industry standard soft-string model versus a more comprehensive stiff-string model to estimate wear. The paper presents a unique approach to predict the remaining useful life of a well barrier and the dynamic state of the well's operational integrity. The prediction is not solely based on statistical modeling but also incorporates barrier engineering and physics of failure in the model.
展开▼
