Summary
Foamed-cement systems are widely used in deepwater-cementing operations because of their various favorable attributes compared with conventional cement systems. For instance, in the Gulf of Mexico, foamed cement is one of the most commonly used systems for shallow-hazard mitigation. However, because current standard laboratory equipment cannot accurately simulate the foam-cementing process in the field, knowledge of the actual properties of foamed cement produced in field operations is limited.
In this study, the microstructure of foamed cement produced by use of field equipment in yard tests is examined in detail. Set foamed-cement samples were analyzed by use of X-ray microcomputed tomography (micro-CT) at different length scales with voxel resolution ranging from 2 to 20 µm. This study establishes the fundamental criteria and procedures necessary to obtain accurate gas-bubble-size distribution of foamed-cement samples by use of micro-CT technology. The test results suggest that foamed cement should be analyzed at multiple length scales to obtain a better characterization of the gas bubbles in the sample. Although a larger region of analysis is useful to obtain a statistically meaningful size distribution of the larger bubbles, small core samples (diameter smaller than 0.5 in.) and fine scan resolutions (5 µm or smaller) are typically required to obtain an accurate measure of the small gas bubbles in foamed cement.
By comparing foamed cement produced by use of field equipment with that produced by use of the traditional multiblade laboratory blender—i.e., the standard American Petroleum Institute (API) method—this study identifies the key characteristic differences of foamed cement derived from different methods of generation. Analysis of the CT-scan images reveals that gas bubbles in foamed cement generated by field equipment approximately follows a log-normal distribution with a wide size-distribution range, from less than 20 µm to more than 1000 µm, and the bubble-size distribution appears to show little dependence on foam quality. Conversely, the gas-bubble-size distribution of foamed cement generated by the API method shows a completely different behavior. It approximately follows a Gaussian distribution, with both distribution range and median varying significantly with foam quality. This research serves as a first step toward predicting the influence of gas-bubble-size distribution on the stability and various other properties of foamed cement to better understand the foam-cementing process in the field.
Influence of the Gas Bubble Size Distribution on the Ladle Stirring Process