Integrated characterization of ultradeep reef-shoal reservoir architecture: A case study of the Upper Permian Changxing Formation in the giant Yuanba gas field, Sichuan Basin, China

Abstract

The Yuanba giant gas field in the Sichuan Basin produces from Upper Permian Changxing heterogeneous reef-shoal reservoirs. It is an ultradeep tight gas reservoir with burial depth of 7000 m. This paper proposes an integrated methodology incorporating the reservoir geology description, petrophysics characterization, seismic response analysis, forward seismic modeling, pseudo-acoustic inversion, and reservoir architecture imaging to identify the geologically complex reef-shoal reservoirs and predict the distribution of the producible reservoirs. The facies of reef cap, reef core, reef base and shoal are first identified based on observations and interpretations of cores, thin sections, mud logging, wireline logs, and formation microscanner images (FMIs). High-energy shoals, reef cores, and reef caps are the most favorable reservoir microfacies. Seismic responses of reef-shoal reservoirs are then investigated through forward modeling and well-seismic calibration. Four reef belts are identified based on seismic reflections and edge detection by seismic paleogeomorphologic features. Reservoir distribution and quantitative properties are predicted using the facies-controlled postattack seismic inversion technique through the correlation between reservoir properties and seismic attributes, which is aided by pseudo-gamma acoustic inversion to exclude the shale-bearing non-reservoir and pseudo-neutron acoustic inversion based on its good relationship with porosity. The reef reservoir thicknesses range from 10 m to 130 m, and the thickest reservoir is located in the northwestern Yuanba gas field. Finally, 3-D visualization is used to characterize the spatial variation of the interior architecture, connectivity, and distribution of a good reservoir in the reef system. The accuracy rate of prediction for reservoirs and non-reservoirs reaches 88% using this method. The technique has also been successfully used for the deployment of development wells and well trajectory optimization, and the rate of encountering the predicted reservoirs is 100%. This study provides a practical approach to identify and predict the tight carbonate gas reservoir with similar petrophysical responses of density and acoustic to non-reservoirs in the complex carbonate system.