Understanding and mitigating leakage pathways in oil and gas well cements

点击翻译按钮获取中文摘要

Operating companies in oil and gas industry are becoming increasingly concerned with ensuring the long term integrity of their wells are maintained. Bulk cement as well as cement/casing and casing /rock interfaces are likely to be the weak spots in wells with regard to leakage and long term well integrity. De-bonding of cement from casing and/or formation, and subsequent loss of zonal isolation, is a significant contributor to fugitive methane emissions from surface casing vent flow and gas migration in oil and gas wells. Understanding the mechanisms of failure and de-bonding and how this impacts fluid movement through the cement is crucial for assessment and improvement of cement performance in wells, whether it is producing or plugged and abandoned. Direct imaging and subsequent modelling of the pore structure and casing/formation interface pre and post-failure, as proposed in this project, will allow for the design and use of more effective mitigation techniques. Experimental observations of cement bonding to wall made of steel or sandstone suggested that de-bonding may be associated with the microstructural changes that takes place in the near-wall zone, which develops along the cement/ casing and cement/rock interfaces. The mechanisms behind stresses inducing fractures along the interface and normal to it are yet to be fully understood. The questions remain for example: Why is the structure in the near-wall region different from that in the bulk material? Why do large radial cracks appear in cement mostly at the cement/rock interface? Does the rock type (e.g. cement bonding quality against impermeable shale vs permeable sandstone formation) has any impact on the magnitude of the radial cracks? How would the change in in-situ stresses due to thermal cycling and/or reservoir depletion effect formation of the cracks developing radially into the cement body or de-bonding along to the interface? Whether the observed leakages are caused by de-bonding, cracks, or both. Furthermore, no information is available on what the leak paths actually look like or how the fluid flows through these leak path. A comprehensive experimental and numerical study is, therefore, proposed.