Generation of Representative PVT Properties through an EOS Model by Integrating Multiple Partial Datasets

Abstract

The field in study has multiple stacked reservoirs with 10-20m oil column overlain by medium to large gas caps. PVT analysis from DST in the gas zones were available, which showed gas and condensate production at surface. Composition, PSAT and CCE data were available from surface samples. DST tests in oil zone provided only GORs, surface oil and gas gravities. Reservoir oil and gas gradients were also available from RFT/MDT.The challenge was to generate representative PVT properties for the oil zone for reservoir simulation by integrating these partial datasets. Due to condensate production, it was necessary also to incorporate and characterize the vaporized oil fraction in the gas, which is not possible by using black oil correlations. This paper describes the workflow used to integrate these partial datasets through an EOS model to generate representative PVT properties for the reservoir oil and gas. The composition from the gas cap sample was used to build an EOS Model which was tuned using CCE data and gas cap DST results. This partially tuned EOS was used to simulate a compositional gradient experiment which was tuned to match the PSAT at GOC obtained from RFT/MDT and equilibrium oil composition at reservoir temperature and pressure at GOC was calculated. This oil composition was then plugged back into the EOS Model and model was further tuned with reservoir oil density from RFT/MDT, oil API and production GOR from the oil zone DST. This tuned EOS was then used to export Live Oil and Live Gas PVT tables for simulation.This approach helped to generate representative PVT tables which produced a good history match in the dynamic model within the range of uncertainty. This approach is now being tried with other reservoirs where PVT data is inadequate and is becoming helpful to tune PVT models.

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

“…EOS was tuned using experimental data, such as constant composition expansion (CCE) (bubble point, relative volume), differential liberation (DL) (Bo, viscosity, Rs, oil density), and separator (oil API or density) tests, and the measured fluid composition (C 1 -C 7ϩ plus or C 30 plus). This was performed for the first depth bin, and after EOS tuning and quality checks (QC) were conducted, a compositional gradient calculation (Das et al, 2014) was used to generate a new oil sample at a deeper depth bin. The predicted composition was compared with the measured composition at that depth, and EOS tuning was repeated with the Rs, Bo, oil viscosity, CCE, and API data to generate a representative black-oil table at that depth.…”