Publications

Fluid characterization is a very critical process for early stages of field development since it generates Equation of State (EOS) model which will be used in reservoir studies such as: early volumetric estimation, material balance, well test analysis, and dynamic simulation (history matching). Inaccurate EOS model could provide misleading information resulted from any of those reservoir engineering studies which can lead to improper reservoir management strategy and inefficient investment. In this paper, fluid characterization from a deepwater gas condensate field, where reliable fluid samples from exploration/appraisal wells is not available, will be systematically explained. The first step is from a decision to collect new fluid samples from development well and how to select sampling method by considering some challenges in our field that can affect sample quality, for instance collecting fluid sample in a well drilled with Oil Based Mud (OBM). The next step was to perform QA/QC process to sampling procedure/execution, handling, sample quality, and laboratory testing. Sample quality is evaluated based on OBM contamination level using skimming and subtraction method. Only samples with the lowest OBM contamination level were sent for further advance tests (CCE/CVD and separator flash). Then some evaluation takes place to assess PVT data quality from those tests using: Mass Balance Plot, Hoffman Plot, Y-function Plot, and Gas Density Comparison between CCE-RDT gradient. Finally, modelling was performed using validated PVT data and selected EOS (Peng Robinson 78 with constant Peneloux volume shift). Sample quality assessment indicates that maximum OBM contamination level is still below 8% of Stock Tank Oil (STO) mol %. Moreover, laboratory PVT data indicate good quality based on all but one assessment and consistency evaluation techniques. Only gas density comparison does not show good agreement between CCE and RDT pressure gradient data. This evidence leads to additional process to yield for another EOS model based on fluid recombination to match RDT pressure gradient. From validation of EOS model using first year of production data, it can be observed that EOS model tuned to RDT pressure gradient can give better match in condensate yields (±10%). This experience demonstrates that even in a deepwater environment with OBM used during drilling, representative fluid sample can still be obtained from BHS if the proper procedure is followed.