Visual Data Analytics: Morphological Image Processing Approach on the Evaluation of Foamed Cement Quality

Abstract Foamed cement is widely used during cementing operations in which lightweight slurries are necessary. Foamed cement helps prevent gas migration and is often used in high-stress environments. A stable foamed cement has a uniform distribution of distinct bubbles, which helps ensure the gas will not break out of slurry. An unstable cement can result in gas coalescence with uncemented sections caused by channeling in the well and density inhomogeneity. Evaluation of foamed cement is necessary before use. This paper proposes a methodology for evaluation of foamed cement quality using two-dimensional (2D) images and an efficient k-means clustering algorithm under image processing. A combination of image-processing techniques is applied to segment bubbles in foamed cement computed tomography (CT) scans. These scans include image compression, morphological image processing (MIP), and image segmentation. Based on the image-processing results, a computer program is developed for foamed cement qualification using a k-means clustering algorithm. The algorithm uses evaluation values calculated from the foamed cement qualification equation, which provides a critical point and separates the data into two clusters. When an image is inserted, the program calculates the evaluation value automatically and assesses its qualification.