Measurement and modelling of interfacial tension and viscosity of reservoir fluids

Abstract

The knowledge of reservoir fluids physical properties is crucial in upstream and downstream processes of petroleum industry. Viscosity and interfacial tension are among the most influential parameters on fluid behaviour. These properties have considerable effects on fluid flow characteristics and consequently in many oil and gas production and processing aspects from porous media to surface facilities. Hence, accurate estimation of the mentioned fluid properties plays a significant role in reservoir development. However, experimental data are scarce at high pressure and high temperature (HPHT) conditions. The work presented in this thesis is an integrated experimental and modelling investigation of viscosity and interfacial tension of petroleum reservoir fluids over a wide range of pressure and temperature conditions. Several series of experimental data on the viscosity of reservoir fluids were generated at high pressure and high temperature conditions (up to 20,000 psia and 200 °C). Experiments were conducted on three binary hydrocarbon systems and three synthetic and real multi-component mixtures, in addition to investigating the effect of dissolved water on the viscosity of the above fluids. Besides, the influence of oil-based mud filtrate on the viscosity of various dead oil samples also was studied as part of this thesis. The effect of different salt concentrations on the interfacial tension of gas-brine systems over a wide range of pressure and temperature conditions also was studied experimentally. The experimental data generated were employed to evaluate, improve and propose predictive models to estimate the mentioned physical properties. A new approach to retrieve the viscosity of original fluid (clean dead oil) from contaminated sample was introduced. Also a novel technique for predicting the gas-water (brine) interfacial tension was outlined. The proposed techniques and models were evaluated against independent experimental data generated in this work and the data gathered from open sources. Predictions of the developed methods were in good agreement with the experimental data.