Time dependent leakage of CO₂ saturated water along a cement fracture

Leakage of CO₂ saturated fluid along wellbores has critical implications for the feasibility of geologic CO₂ storage. Wells, which are ubiquitous in locations ideal for CO₂ storage, develop leaks (e.g. fractures) for many reasons and at different points in their age. Small leaks pose the most significant risk to geological CO₂ sequestration because they are difficult to detect and provide a direct pathway through which fluid can escape the storage formation. This dissertation shows that due to complex coupling between reaction and flow, leaking wells will tend to self-seal via secondary precipitation of calcium carbonate in the open pathway. Residence time, fluid reactivity, and initial fracture aperture all play a key role in determining the time required to seal the leakage pathway. To test the self-sealing hypothesis, laboratory experiments were conducted to inject reactive fluids into naturally fractured cement. Restriction of the leakage pathway, i.e., the fracture, was inferred from the relationship between flow rate and pressure differential. Precipitation was observed in both constant flow rate and constant pressure differential experiments. In the former precipitation resulted in an increasing pressure differential, while precipitation caused a decrease in flow rate in the latter. Analysis by electron microprobe and x-ray diffraction, and corroborated with effluent chemical analysis, showed that the reacted channel was depleted in calcium and enriched in silicon relative to the original material. The remaining silicon rich material prevents widening of the reacted channel and development a self-enhancing (e.g. wormhole) behavior. Self-limiting behavior is caused by calcium mixing with carbonate ions in high pH slow flow regions where local residence time is large and calcium carbonate is insoluble. Secondary precipitation initially develops next to the reacted channel and then across the fracture surface and is the source of pathway restriction and the self-sealing behavior. Results from the experiments are used to develop a simple analytical model to forecast well scale leakage. Future work is needed to test a broader range of experimental conditions (e.g. brine salinity, cement formulations, cement-earth interface, effect of CO₂ saturation, pressure, and temperature), to improve our understanding of both the fundamental behavior and the leakage model.