Reservoir fluid properties play a crucial role in the upstream field development cycle. Petroleum engineers extensively utilize Pressure-Volume-Temperature (PVT) studies in applications such as calculations of pipelines’ pressure drop, and assessment of Enhanced Oil Recovery (EOR) strategies. These studies are generated from a series of lab experiments conducted on reservoir fluid samples in high pressure-high temperature (HPHT) lab environments, and commonly matched using Equation of State (EOS) software.

Feeding and characterizing the composition of a reservoir fluid in a PVT software play a central role towards understanding its behavior. These steps are heavily affected by the last carbon number measured and the lumping scheme used in the simulator. This paper investigates the application of splitting the plus fraction, and utilizing Saturates, Asphaltenes, Resins and Aromatics (SARA) analysis in enhancing viscosity prediction at atmospheric conditions.

In this study, three oil samples from fields with suspected flow assurance issues were selected. A fingerprint study was first conducted on all samples to ensure that they are representative of the original reservoir fluid, and free of any drilling fluid contaminants. The methodology used in this study is based on conducting compositional analysis and viscosity test on the selected samples. Furthermore, SARA analysis was conducted to enhance the characterization of reservoir fluid, and confirm asphaltene presence. Lastly, splitting technique and SARA-based lumping scheme were used to predict viscosity values at atmospheric pressure and were compared to experimental data.

The results of this work demonstrated the effectiveness of SARA-based lumping scheme on atmospheric viscosity prediction, which captured the plus fraction concentrated in the dead oil without compromising the computational time. Furthermore, the EOS software used studied the sensitivity of the simulation results to different compositions.

You can access this article if you purchase or spend a download.