This research investigates the way laboratory oil formation volume factors (FVF or Bo) from differential liberation (DL) and flash separation tests are used by engineers to represent real reservoir production processes. A survey about Bo was performed among industry professionals. The work proposes standardisation in these approaches, including a way of adjusting laboratory DL data to represent a true differential process and a new technique for obtaining composite Bo data from combined DL and separator tests. Finally, 3-D reservoir black-oil and compositional models were developed to calculate Bo and explore which phase behaviour processes ECLIPSE 100 and ECLIPSE 300 are mimicking.
This paper shows laboratory Bo data are commonly misunderstood by engineers, and recommends greater awareness and better education. A novel analysis of laboratory DL studies demonstrates that these can seriously overestimate true differential processes. The work tests and discusses the feasibility of using 3D compositional simulation models to derive more realistic Bo values for the combined reservoir and production process. A procedure has been written based on the findings of this work to help petroleum and PVT engineers to standardise the reporting and proper use of laboratory Bo measurements.