Single-well Sonic Imaging: High-Definition Reservoir Cross-sections from Horizontal Wells

Abstract

Abstract The limited resolution of conventional seismic surveys often results in challenges to successful reservoir development. Reservoir structure and thickness may not be determined from the seismic to the level of detail necessary, affecting the placement of the horizontal well. We describe a new sonic imaging technique that resolves these features. In addition the position of the well in relation to the reservoir perimeter can be determined with great accuracy. Single-well sonic imaging places the acoustic transmitter and receivers in the same well, analogous to recording a micro-scale seismic reflection survey downhole. Using modified sonic logging technology, sound is broadcast into the formation and reflections are converted into a two- or three- dimensional spatial image of the zone around the borehole. Acoustic reflectors such as the top of the reservoir, calcite stringers, or fractures are mapped at distances of 10–15 meters from the well with resolution on the order of one-half meter. We show several examples of surveys conducted in horizontal wells to illustrate applications of the technique. Introduction The traditional paradigm for development of a hydrocarbon reservoir consists of drilling a large number of near vertical wells, each entering the reservoir through the cap-rock and extending through the reservoir into the formation below. This provides an excellent measure of the reservoir thickness, fluid contact horizons and petrophysical properties of the formation at each of the well locations. Logs are acquired in each well during the discovery, appraisal and early development phases. The resulting composite data set consists of measurements collected along vertical lines at many locations within the reservoir. Although each log provides extremely high-resolution information along the borehole, depth of investigation is typically only a few borehole radii. Characterization of the rest of the reservoir requires extrapolation or interpolation between wells using simple geometrical assumptions about the geology, combined with seismic or geostatistical methods as a guide. This approach results in a steady increase in information about the reservoir, and a decrease in risk, during the early stages of the reservoir's life. A further consequence of this paradigm is that logging tools, inversion algorithms, and interpretation techniques were developed to fit the model of measurements recorded in vertical wells. However, the increasing use of horizontal and highly deviated wells and the trend towards drilling them earlier in the development process has introduced a new set of challenges for traditional measurements and interpretation. For example, there are fewer horizontal wells per reservoir. Each horizontal well samples only a limited vertical region of the reservoir, perhaps just a few feet below the cap-rock, as compared to the full vertical cross-section typical in the conventional well. The inversion and interpretation of logs also becomes more complex - for example a variation in the gamma ray log may be due to either a local change in shaliness or to the well trajectory straying closer to the shale cap-rock. Added to these reservoir delineation challenges are issues associated with the construction of horizontal wells themselves. Often the reservoir is below the resolution of the seismic cross-section, yet the optimum positioning of the well requires its placement at a specific location within the reservoir. Steering the well and verifying its position relative to geological markers in order to optimize completion and production requires new analysis techniques for horizontal well construction.