Generation of Black-Oil PVT Properties Tables Using Tuned EOS Model by Integrating PVT Laboratory Test Datasets

Abstract

This paper describes the workflow used to develop a systematic approach for integrating a large number of pressure-volume-temperature (PVT) datasets and generating representative PVT tables for use in a black-oil reservoir simulation. The field in this study is a faulted carbonate reservoir with multiple producing zones. A variation in PVT properties was evident from laboratory tests conducted on the collected fluid samples. The challenge was to design a systematic approach for capturing PVT property variations, performing oil compositional variation modeling, and understanding PVT compartmentalization. A simplified approach is proposed that combines PVT properties variation and PVT compartmentalization into a single workflow. PVT properties variations were used to identify various trend lines, and PVT samples belonging to a trend line were assigned to one PVT compartment. An equation of state (EOS) model was tuned for each PVT compartment, and the tuned EOS was used to model compositional variation and properties variation with depth. Validation of the workflow was performed by assessing the quality of the EOS tuning and EOS calculated compositional gradient match with laboratory measured data. It was discovered that EOS models were able to capture PVT properties variation and compositional variation with depth. The workflow was found very useful for generating a new set of PVT tables required for history matching reservoir pressure.