Derisk Exploration Targets Offshore NW Scotland by Combination of Angle Domain Imaging and Machine Learning Techniques

West of Shetland (WoS) area is one of location of the UKCS largest remaining hydrocarbon reserves. For de-risking the WoS exploration Generalized Radon Transform (GRT) migration has been applied to preserve true amplitude and output exact angle gathers which are two important factors for Reservoir Characterization. GRT is ray-based migration scheme but it carried out in angle domain. Machine learning methods have been integrated into Quantitative Interpretation workflow. First, the Fuzzy C-means clustering algorithm is used to quickly scan for AVO anomalies from a large area. The machine learning picked AVO anomaly area matches with Extended Elastic Impedance (EEI) fluid (+200) slice at target horizon very well, and both can confirm with well information (dry well and hydrocarbon well), which means the higher confidence can be achieved on the potential reservoir since two sides of inputs show very similar outputs. After the target area has been chosen unsupervised machine learning algorithms like Principal Component Analysis (PCA) and Self- Organizing Map (SOM) are applied on seismic attribute volumes to pick geo-bodies/sweet spots. The final conclusion is GRT migration integrated with machine learning methods in QI workflow can derisk WoS exploration and to precisely understand the amplitude anomaly for any further field development.