Research on Fluid Identification Methods for Cambrian Dolomite Reservoirs Based on New Characteristic Response Parameters

Most carbonate reservoirs have poor physical properties, low porosity and permeability, strong heterogeneity and significant anisotropy. It is difficult to accurately identify the fluid properties of complex oil and gas reservoirs by single logging method. Therefore, to address the challenge of fluid identification in carbonate rocks, this study takes the Cambrian dolomite reservoirs in the Tarim basin as an example. Based on the existing data from the Tarim oilfield, a new characteristic parameter, resistivity per unit porosity RA, is proposed. Furthermore, this paper explores multiple fluid identification methods based on the new characteristic response parameters. These methods include two cross-plot methods, namely RA—C1 and RA—Δt, and the artificial intelligence-based random forest identification method. The results of case studies show that, in terms of the accuracy of the three methods, the random forest intelligent method has an edge over the cross-plot methods. Specifically, the fluid identification accuracy of the RA—C1 cross-plot method reaches 86.79%, that of the RA—Δt cross-plot method is 84.55%, and that of the random forest intelligent method is 88.56%. However, the cross-plot methods exhibit higher accuracy in identifying gas-bearing water layers and water layers. In conclusion, the comprehensive application of these methods can significantly improve the accuracy of fluid identification in carbonate rocks, providing an effective approach to solve the practical problem of accurately discriminating the fluid types in carbonate reservoirs during oil and gas exploration and development.