Log Facies Evaluation and Property Modelling of a Turbidite Reservoir, the Gulf of Mexico

Abstract

This master thesis presents a hierarchical modelling workflow in a slope canyon of the turbidite sandstone reservoir in the Formation X at the PA field in the Gulf of Mexico. The Formation X with approximately 400 m thick was deposited during Late Paleocene to Late Middle Eocene and composed of alternating shale and sandstone intervals. Depositional lithofacies, petrophysical well characteristics, 3D seismic data and geological facies models have been used for characterizing and evaluating of this turbidite reservoir. The reservoir properties and heterogeneities are typically controlled by facies architecture. Depositional lithofacies have been interpreted using core description in few wells, and have been tried to link with the predicted electrofacies from well logs. Two main logs representing depositional heterogeneity -GR and Vsh- have been interpreted using neutron and density logs and have been used to predict electrofacies in the Formation X. Three electrofacies have been predicated and considered as facies associations. The facies associations include sand, shaly-sand and shale. The core descriptions, regional studies and seismic data are valuable data to understand the conceptual geological model that can be used a guide to distribute the facies associations between the wells. Well tops from well logs have been interpreted by variations in depositional characteristics and have been tied to seismic surfaces in order to construct the stratigraphic framework. A total of eight surface resulted in defining seven geological zones to populate reservoir properties. The geological grids have been defined according to the possible depositional heterogeneities in lateral and vertical directions. Well data including facies and porosity in a scale of 15 cm have been scaled-up to geological grids with 1 m thick. Due to uncertainty in the conceptual depositional model two different alternative facies modelling methods have been tried in order to learn the modelling processes and to understand the challenges which may influence the final results. Object-based models have been used to model the sand facies as channel, shaly-sand facies as overbank and shale facies as background. The pixel-based facies models have been built based on random distribution of the facies according to the results from data analysis in available wells. Petrophysical models have been constrained to the pixel-based facies models using random function simulation algorithm. According to data analysis and the results from modelling, the depositional facie control reservoir properties. Sand facies have higher porosity values than shaly-sand and shale facies. The result from geological description, petrophysical evaluation, seismic interpretation and property modelling has shown that the Formation X is a complex turbidite system. Such complexities results in higher uncertainty in the results and thus uncertainty assessments are necessary in this kind of reservoirs.