document preview

Abstract
The RSE-1 well operated by Total is located within the Rousse field in the southwest of France. The 43 year-old well, a gas producer from a now depleted reservoir has undergone extensive research in order to determine its suitability for CO2 storage. The main objective, in terms of containment, is to avoid that the injected CO2 migrates back to the atmosphere or to potable aquifers. Existing wells are often identified as a potential weak point in the containment system, because cement, casing and elastomers can be degraded by CO2 over time. Since any defects within the cement sheath may accelerate this aging mechanism, the material behind the casing should be fully characterized especially across the caprock.

Cement bond log data was available from the 43 year old RSE-1 well, which only gave a qualitative indication of the casingcement bond, but did not allow a detailed characterization of the material behind the casing. To gain more detailed information, new cement evaluation logs were acquired over the lower 905 m of a 2000 m thick caprock using a new generation 3D ultrasonic cement imager coupled with an array sonic imager. The use of different, independent measurements revealed good casing-cement bond that could be compared to the original logs from 1967 to confirm no degradation over the production life of the well. However, annulus imaging revealed breakouts and a very eccentered casing (almost touching the borehole wall), both of which led to the presence of mud pockets. Detailed evaluation of these pockets indicated that they are not connected and that zonal isolation over the 836 m cemented section of the caprock is indeed achieved. Modeling of cement imaging data also suggested good cement-formation bond, without any external microannulus presence.

Detailed characterization of the material behind casing, casing eccentralization and intermittent mud pockets provided an improved understanding in terms of well integrity. The results showed that the RSE-1 well nevertheless provides adequate hydraulic isolation for the injection of CO2 and that the cement sheath can therefore be eliminated as a potential weak spot in terms of containment.

Introduction
Geological storage of CO2 is recognized worldwide as part of a suite of options towards mitigating climate change. To be an effective mitigation tool, it is essential that the injected CO2 remains within the storage reservoir and the migration to potable aquifers or the atmosphere be avoided. The presence of wells intersecting this reservoir creates a potential weak link in the containment system, since casing, cement and elastomers can be degraded by long-term exposure to CO2.

Cement, although reactive if exposed to CO2, has a very low permeability in the order of 0.5 to 5 μD, which means that most of the CO2 will travel through the cement matrix by diffusion, a very slow process even over a length scale of one meter [Loizzo et al, 2008]. However, any defects within the cement sheath may accelerate the movement of CO2. Possible defects might include liquid channels within the cement sheath and microannuli at the casing-cement or cement-formation interface, which could provide direct pathways for CO2 migration; high water/cement ratio increases cement permeability and decreases the resistance to CO2 aggression, thereby providing a further possible slow pathway for CO2 migration. Some defects may be created after cement has set, even by the CO2 injection process itself. In order to effectively mitigate possible well integrity risks it is necessary to fully characterize the geometry and properties of the material behind the casing as well as its defects, especially across the caprock.

We will describe the study of a 43 year-old gas production well, which has undergone extensive research to determine its suitability for CO2 injection. As part of the research, new cement and corrosion data was acquired to evaluate the cement hydraulic isolation across the caprock. Using the acquired data, the material behind casing and its possible defects were studied and compared to cement bond logs run right after well construction in 1967 to determine possible degradation over time.