Pre-Stack Detailed Frequency Variation Study and Application in Complex Sandstone Reservoir Hydrocarbon Detection

Bohai oilfield is an important offshore oil and gas producing area in China. The fluvial sandstone reservoir is an important production series, which accounts for about 45% in the total proved reserves. However, in recent years, the fluvial sandstone exploration encountering new challenges. The target fluvial sandstone series are buried shallow, about 1000-1600 meters underground, due to the large sedimentary accommodation space, huge river channels are developed. Now, almost all huge river channels have been drilled, the remaining complex river channels that drastic changes in both vertical and horizontal directions have become important targets for future exploration. Because of the frequent oscillation and migration in river channels, the vertical superimposition of these fluvial sandstone reservoirs is poor, making it difficult to evaluate the oil and gas reserves by old drilling schemes, otherwise, high-priced exploration costs will make it difficult to commercially develop these reservoirs. The key to reducing exploration costs lies in accurately assessing whether complex river sand bodies contain oil and gas before drilling. Hydrocarbon detection can be used as an efficient tool for screening potential targets and drilling risk assessment, but it requires high quality seismic data. Bohai Oilfield has basically achieved full coverage of 3D seismic data, including both towed cable data and submarine cable data. Considering the shallow burial depth of this complex river sandstone reservoir, the seismic data quality is generally sufficient, but the hydrocarbon detection results cannot meet the precise research requirements, the oil layer detection accuracy is very low. Through numerous examples comparison and analysis, it is believed that there are two main reasons for the low accuracy of hydrocarbon detection, first is caused by the hydrocarbon detection method, and the other is caused by the seismic data itself. Generally, hydrocarbon detection methods can be divided into two categories: pre-stack hydrocarbon detection methods and post-stack hydrocarbon detection methods. There are many methods for post-stack hydrocarbon detection, but they mainly rely on the reservoir reflections amplitude and frequency. In the past decade, the widespread application of post-stack hydrocarbon detection methods in Bohai oilfield is mainly due to the discovery that oil layers proven by drilling have typical characteristics of strong amplitude reflection. But in subsequent exploration, it was found that the amplitude of water bearing reservoirs can also be very strong, even stronger than oil layers. Once the hydrocarbon filling capacity decreases, amplitude detection will lead to the total failure of drilling. Although forward analysis has found that there are more significant differences between oil and water layers in the frequency spectrum, it is difficult to effectively improve the accuracy of post-stack hydrocarbon detection in practical applications, as the strong reflection water layers cannot be suppressed. Pre-stack hydrocarbon detection methods mainly include PG attribute construction, pre- stack inversion, etc. In the pre-stack data, as the incident angle changes, the intercept and gradient of oil and water layers vary, which is the basic principle of PG attribute detection method. By using angle limited stacks, pre-stack inversion can simultaneously invert P-impedance, S-impedance, Vp/Vs and density data. Simultaneously with these volumes, Vp, Vs and Vp/Vs volumes were computed whether there is oil or water in sandstone reservoirs. The commonality of these pre-stack hydrocarbon detection methods is that they require high-quality pre-stack gathers to obtain accurate hydrocarbon prediction results. In one target area, multiple sets of seismic data with different processing methods may have significant reflection differences in pre-stack profile, although there is little difference in post-stack profile. Currently, there is no clear conclusion on how to determine whether a set of seismic data is suitable for pre-stack hydrocarbon detection research, which restricts the application of pre-stack hydrocarbon detection methods and directly leads to a low drilling success rate in complex sandstone reservoirs exploration.