Abstract
Abstract Phase alteration is a physical phenomenon that constantly takes place in reservoirs, during production or injection, due to the regional change in fluids’ pressures or temperatures. The transformation of phases also occurs frequently in enhanced oil recovery (cold solvent injection, CO2 injection, etc.) or thermal enhanced oil recovery (steam injection, solvent-thermal application) at which injected fluids condense gradually throughout the matrix owing to heat loss or an increase in pressure. Achieving an accurate prediction of phase change in the porous media is important for attaining trustful forecasting of hydrocarbon recoveries, retrieval of the injected solvents, vapor-liquid equilibrium calculations, and modelling of two-phase envelopes. Generally, the capillary effect has an impact on the properties of fluids when medium sizes are tight enough to affect the molecules and empower pore-molecule interactions. One of the fluid properties influenced by high capillary pressures is the phase-change. When pore sizes become tighter than 100 nanometers, vaporization and condensation temperatures/pressures of confined fluids tend to shift from bulk measurements. Understanding the nature of physical state alteration (liquid-to-gas or vice versa) has attracted the attention of researchers because of its enormous importance in pressure-volume-temperature (PVT) calculations and reservoir simulation. In petroleum industries, cubic equation-of-state (EoS) models are used to predict the phase behavior of hydrocarbons in the reservoirs. One of the major limitations of the commonly used cubic EoS (Peng-Robinson EoS and Soave Redlich-Kwong EoS) is that they do not consider the confinement effect on the phase-alteration behavior of fluids. Such a drawback causes these cubic models to be inaccurate in modelling two-phase envelopes of rock fluids in extended tight reservoirs (shales, tight sands) or even in permeable rocks (sands under thermal injection). This paper experimentally investigates the phase-change behavior of hydrocarbons in various categories of the porous system. The experimental journey was initiated by studying the boiling behavior of single-component hydrocarbon liquids in silica-glass Hele-Shaw cells which is represented as simple capillary spaces with different sizes. Shifted boiling temperatures were observed in the glass cells, due to the confinement effect. As a next step, the vaporization of hydrocarbons was analyzed in homogeneous and heterogeneous silicate-glass microfluidic chips. Early vaporizations of solvents were observed in the micromodels, as a result of capillary effects in the porous systems. The analysis was shifted forward to focus on the phase-change behavior of hydrocarbons in real reservoir rocks. Owing to the existence of nanopores, the vaporization of tested liquids took place in the rocks at temperatures lower than normal boiling points and calculated boiling temperatures by the Thomson equation. Such reductions were also observed in the permeable rocks although the volume percentages of extended confined pores (< 100 nanometers) were less than 5%. The second set of experiments paid attention to measuring the vaporization temperature of single-component and multicomponent hydrocarbon liquids in different rocks at various pressures. The pure-component solvents were representing the injected solvents in cold solvent injection or as additives to steam, and the multicomponent solvents were representing non-complex light oil. The experimental results were, then, compared with the normal boiling points and calculated phase-change temperatures by the original version of the Peng-Robinson EoS. Noticeable deviations of measured vaporization temperatures of liquid solvents from the bulk and computed values were observed. As the medium gets smaller, interior pore surfaces begin to have influences on boiling temperatures of hydrocarbons, due to the pore-molecule interactions. Studying the condensation of propane in various reservoir rocks under isothermal and non-isothermal conditions was also a part of the investigation. The vaporization of propane was inspected in different rocks at various temperatures. The experimental observations were then compared with the bulk vaporization pressures and computed saturation pressures by the Peng-Robinson EoS and Kelvin equation. The recorded vapor pressures, in the rocks, were 7% lower than the bulk values and calculated vapor pressures by the Kelvin equation. Meanwhile, the propane vapor pressures, in the rocks, were 15% (on average) lower than the pressures modelled by the Peng-Robinson EoS.
Published Version
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have