Analysis of field samples
The actual impact of CO2 injection on well integrity depends on whether cements in existing wells are more or less susceptible to attack than those used in the laboratory. This question is beginning to be answered by cements obtained in 2004 from a 19 year-old well at the Rocky Mountain Oil Testing Center in Wyoming. George Scherer of the Storage group worked with Schlumberger and RMOTC staff to design and execute an innovative drilling project that retrieved one intact sidecore (composed of cement, casing, and formation rock) and several fragmented cement samples from depths of 3000-5000 feet.
Initial analysis of the samples revealed a surprisingly complex microstructure that differed greatly from what was expected for the type of cement described in the cement log for the well. However, through collaboration with NETL it was confirmed that the sample was from the lead cement, which contained fly ash, rather than the cement that was expected to be in the location from which the sample was taken. A report on the analysis is in preparation.
Field estimates of well-bore permeability
In large-scale models of injection and leakage, the most important parameter values, and the ones that have the largest uncertainty, are the permeability values assigned to all of the well segments in the domain being analyzed. To our knowledge, these values have never been measured, nor are there any estimates in the open literature. Therefore we decided to design a field experiment that we believed would allow these critical parameters to be determined. The test is described in Gasda et al. (2007), and is shown schematically in Figure 8, where a constant pressure is imposed at a location along the well annulus and the pressure response is measured at another location along the annulus. In the published paper, we considered limits based on fracture pressures and on instrument accuracy to determine ranges of detectability for the effective well permeability. This test was applied for the first time a few months ago as part of a bp/Schlumberger CCP2 project, and the results are encouraging. Initial analysis of the resulting data shows that we are able to estimate the effective permeability of the cement section outside of the well casing, and that the estimated values are consistent with other measurements. We believe this kind of experiment can provide the greatest reduction of uncertainty in leakage estimates, and we will work to a continuation and expansion of these experiments at different field sites.