document preview

Abstract
Leakage through new or existing wellbores is considered a major risk for carbon dioxide (CO2) geological storage. Long-term effective containment of CO2 is required and the presence of millions of suspended or abandoned wells exacerbates the potential risk in mature hydrocarbon provinces. Accurate estimates of risk profiles can support the acceptance of geological storage and the adoption of economically effective risk prevention and mitigation measures.

Reliable data about long-term containment of CO2 is almost non-existent, so wells that exhibit a similar risk profile (such as gas storage, gas production and steam injection) should be used as proxy to assess failure rates and consequences for cemented wellbores.

Statistical data about occurrence of leaks and their consequences are analyzed to determine the risk profile of CO2 leaks. A smaller sample of data about leak rates is also analyzed to provide their statistical distribution. Rates and consequences are then compared to try and assess the order of magnitude of major and catastrophic leaks.

Hydrothermal CO2 leaks in natural analogs are also reviewed to compare the distribution of leak rates and the consequences on health, safety and environment of CO2 releases to soil and atmosphere.

Analysis of existing data will show that major leaks are likely to occur in less than 2 wells per 1,000, with the overwhelming majority of CO2 leaks being small and with limited or negligible consequences.

Given their risk profile, CO2 wellbore leaks should be addressed through a routine risk management approach. Their frequent occurrence requires effective prevention measures, such as understanding leaks and adapting and deploying practices to minimize their occurrence. On the other hand, their low impact ensures maximum effectiveness of mitigation measures, such as monitoring: since leaks can be detected long before damage ensues, they can be observed to predict their long-term consequences and to plan the most effective intervention without unnecessary immediate operation shut-downs.

In conclusion, the recommended course of action is to focus on risk prevention and early detection. This implies the evolution from a “no leaks” attitude (even for negligible leak consequences) to one that seeks no damage and relies on tight surveillance.

Introduction
Leakage risk
Leakage through wellbores is a major source of concern for the large-scale deployment of geological carbon storage. Reliable estimates of both the quantity of CO2 that can migrate through wells and the consequences of this migration are needed in order to properly manage the leakage risk and show that CO2 can be safely stored underground for very long periods of time.

These two tasks are subtly different: “risk” implies that the leaked CO2 affects a target (or stake) and causes a loss, to people (health and safety), environment, assets or infrastructure. Large quantities of CO2 could leak to the atmosphere, for instance in a desert, without causing any damage; conversely small leaks, if allowed to accumulate, could result in substantial losses. Estimating occurrences of leakage rate is thus solely dependent on the leakage mechanism, whereas risk depends on the additional uncertainty of whether targets are affected and what is the possible level of loss. On the other hand data on losses, thus on risk, are easier to come by since the general public and regulators are involved. Furthermore, even if CO2 leaks to the atmosphere without causing immediate damage, it will still defeat the purpose of storing it in the ground. We can then speak of “global risk”, as opposed to the local risk of losses at or near the storage site.