Man-made versus natural CO (sub 2) leakage; a 400 k.y. history of an analogue for engineered geological storage of CO (sub 2)

To evaluate sites for long-term geological storage of CO (sub 2) and optimize techniques for monitoring the fate of injected CO (sub 2) , it is crucial to investigate potential CO (sub 2) migration pathways out of a reservoir and surface leakage magnitudes. For the first time, we calculate CO (sub 2) leakage rates and volumes from ancient fault-related travertines and from an abandoned borehole. U-Th-dated travertine along two faults near Green River, Utah (western United States), shows that leakage has occurred in this area for over 400 k.y. and has switched location repeatedly over kilometer-scale distances. One individual travertine was active for at least 11 k.y. Modern leakage is predominantly through the active Crystal Geyser, which erupts from an abandoned exploration well. Using age data and travertine volume, we calculate magnitudes and rates of CO (sub 2) emission. Fault-focused leakage volume is twice as great as diffuse leakage through unconfined aquifers. The leakage rate from a poorly completed borehole is 13 times greater than the long-term time-averaged fault-focused leakage. Although magnitudes and rates of any leakage from future storage sites will be highly dependent on local geology and pressure regime, our results highlight that leakage from abandoned wells is likely to be more significant than through faults.