All microannuli are not created equal: Role of uncertainty and stochastic properties in well leakage prediction

Abstract

Carbon Capture and Storage (CCS) as a method for climate mitigation relies on CO2 being locked up in subsurface reservoirs with a long-term perspective. Active and abandoned wellbores are among the major potential leakage paths for CO2, and their success as "gate keepers" relies on the quality of well cement. The length requirements and recommended practices for cement sheaths in active wells vary from country to country. In this study, we investigate how stochastic properties of the microannulus may affect the recommended "safe length " of cement sheath, should assigning such length be attempted for CO2 wells. We demonstrate, by means of a simple numerical model, that variation in the width of the microannulus along the well makes it a challenging task to assign a figure to the "safe cement-sheath length", even though longer cement sheaths do indeed reduce the risk of leakage. Variation in the mean value or standard deviation of the microannulus width by only 10% changes the recommended length of the cement sheath by up to an order of magnitude. This finding sheds new light on the use and added value of more accurate leakage prediction models: due to the uncertainties in and the lack of information about the properties of microannuli, such models should focus on investigating the effects of different operational and in-situ factors on the leakage potential rather than on attempting to accurately predict the leakage flow rate, however tempting such prediction exercise might seem at first glance.