Publications

The Volve Field is one of the developed fields with proven oil reserves located in the North Sea, Norway. The productive reservoir sandstone was the syn-rift Hugin Formation. Reservoir characterization can be identified through Quantitative Interpretation (QI) using both deterministic and probabilistic methods. The probabilistic method is an inversion technique that uses a geostatistical approach, specifically, Stochastic Inversion and Probabilistic Neural Network (PNN). This study aims to compare deterministic and probabilistic methods and improve the seismic resolution for reservoir characterization based on elastic parameters sensitive to lithology and fluid content. In this study, which involves rock physics sensitivity, the lithology-sensitive parameter was Vp/Vs as sandstone, Whereas the elastic parameter sensitive to fluid was Poisson's ratio as sandstone lithology where hydrocarbons were present. The accurate delineation of hydrocarbon zones within clastic reservoir intervals was achieved through the use of Vp/Vs < 1.84 for reservoir lithology and Poisson's Ratio < 0.26 for fluid discrimination as elastic parameters, demonstrating a clear advantage over deterministic methods using acoustic parameters. Stochastic Inversion enhanced layer continuity and boosted well log data matching by 0.75 over the deterministic approach of 0.72. The productive zone of the Hugin Formation is characterized by good reservoir quality, as evidenced by porosity values between 18% and 23% in the Probabilistic Neural Network (PNN) cube prediction. The application of the PNN method resulted in a better correlation of 0.125 and an error reduction of 5% compared to the deterministic method for the distribution of porosity using external attributes from the stochastic inversion volume. This proves the significant enhancement of the stochastic and probabilistic applications of this research method.