Abstract
Cement annulus is essential to prevent the migration of geological fluids along the well. However, temperature and pressure variations in the casing may damage the annulus and create leakage paths, either cracks in the cement or debonding (microannulus) at the interface with the casing or the rock. Understanding when damage occurs, and how severe it is, is essential to ensure safe operations. In this paper, we present experimental results obtained with the small-scale setup described in (Feneuil et al., 2024), and we study microannulus formation and geometry in two different samples. Both samples are prepared in the laboratory, they comprise of a Portland G cement cylinder (diameter 5 cm, height about 4 cm) placed around a 17-mm diameter steel casing. The setup allows for the evaluation of debonding between the casing and the cement by three different techniques: Pulse-Echo acoustic logging, gas-leakage tests and 3D visualization with X-ray Computed Tomography. With the first sample, we evaluate how the casing-cement bond changes during two temperature cycles from 23°C to -20°C. We show with the acoustic logging tests that cooling down the sample by 30°C below the curing temperature did not affect the casing-cement bond. Clear debonding was triggered after cooling down by 40°C, the former microannulus is heterogeneous, and does not provide a leakage path. The second sample contained a large pre-formed microannulus. The CT pictures show that the microannulus thickness reaches locally several 100 µm, however, the leakage tests reveal that the flow is constrained by the regions of smaller debonding, as the effective microannulus size is about 1 µm. In addition, we observe that the microannulus size is affected by the value of the confinement pressure. The setup and method described here will be further used to compare the efficiency of different sealing materials and as input for simulation work.