Granite "buried hill" oil pool is an unconventional oil pool which can be formed a large and highly effective oilfield in some basins such as Bach Ho oilfield in Vietnam and Kharir oilfield in Yemen.However, the diversity of reservoir types and combination,highly heterogeneity, and monotony method of reservoir prediction restrict the reservoir prediction accuracy of the granite buried hill. With the development of acquisition and processing of wide azimuth seismic data, much more new methods can be used to improve the prediction accuracy in this paper.
Because of the complexity of the granite reservoir formation, it is too difficult to predict and evaluate reservoir only by one method. Through the past years study, workflows are developed mainly including three steps. The first step is classify the reservoir according to the reservoir space type based on the comprehensive analysis of core,outcrop,wireline, FMI data. The second step is summarize seismic response according to the different types of granite reservoir. The third step is use different seismic techniques to establish the corresponding predication workflow based on the seismic response and data situation.
Granite buried hill reservoir is divided into four types in vertical : weathering-leaching reservoir (A), fracture-vug reservoir (B), fractured reservoir (C) and tight granite (D). Different types of reservoir have obvious differences in core slice, wireline and FMI data. On seismic sections, reservoir A shows good continuity and high amplitude; reservoir B is different--some have good continuity and mid-high amplitude, and others have poor continuity and low-mid amplitude; reservoir C has poor continuity and low-mid amplitude.Reservoir A has the characteristics of low velocity and density, and is different in P-wave impedance from overlying clastic rock and underlying bedrock,thus it can be identified and predicted used by P-wave impedance inversion and amplitude spectrum gradient attribute.Reservoir B demonstrates the abnormal features of moniliform/punctate amplitude in amplitude spectrum gradient attribute which shows the development degree of solution fractures and vugs. Reservoir C can be predicted by post-stack attributes of volume curvature and coherence cube in a large scaleand then refined prediction used wide azimuth seismic data in a small scale, the fracture intensity and strike prediction results are matched the FMI data. The pre-stack fracture prediction can not only predict the fracture intensity, but also the strike. Totally, 11 of 13 well prediction results of fracture density and strike are matched well with drilled wells, and the coincidence rate is nearly 84%.
Using this technique series have obtained a good result in B Basin, which provide a technical reference for similar complex reservoir prediction and play more and more important role in granite reservoir prediction.