Experimental and theoretical studies of solvent bubble nucleation and liberation processes in different heavy crude oil−solvent systems

Highlights

• CH₄ can induce a higher supersaturation in a heavy oil in comparison with CO₂ or C₃H₈.

• The heavy oil can entrap more CH₄ bubbles than CO₂ or C₃H₈ bubbles.

• The liberation of CH₄ bubbles is much slower than that of CO₂ or C₃H₈ bubbles.

• CH₄ induce the strongest and most stable foamy oil compared with CO₂ and C₃H₈.

Abstract

In this paper, solvent bubble nucleation and liberation processes in the heavy crude oil−CO₂ systems, heavy crude oil−CH₄ systems and heavy crude oil−C₃H₈ systems were experimentally studied and theoretically analyzed. First, two respective series of tests were conducted for different heavy crude oil−solvent systems. The first series included eleven conventional isothermal constant-composition-expansion (CCE) tests and the other series consisted of three new isothermal constant-composition-expansion & compression (CCEC) tests. Second, the amount of the evolved gas (i.e., the dispersed gas and free gas) in each pressure reduction step was determined from the measured CCE test data to study the solvent bubble nucleation process in each heavy crude oil−solvent system. A new quantity named the bubble nucleation index (BNI) was introduced and used to represent the solvent bubble nucleation strength. Third, the respective amounts of the dispersed gas and free gas in each pressure reduction step were obtained from the measured CCEC test data to examine the solvent bubble liberation processes in three heavy crude oil−solvent systems. A second new quantity named the bubble liberation index (BLI) was defined and applied to represent the solvent bubble liberation strength. It was found from the CCE and CCEC tests that the measured P_cell−v_t data for each heavy crude oil−solvent system had three distinct regions. Region I was the one-phase region. Region II was the foamy-oil region, in which the solvent bubble nucleation started and the solvent bubbles were dispersed in the heavy oil. Region III was the two-phase region, in which the free-gas phase was formed and started to dominate the total compressibility of the heavy crude oil−solvent system. In addition, the solvent supersaturation vs. reduced pressure data indicated that the heavy crude oil−CH₄ system had the lowest solvent supersaturation at the same reduced pressure in comparison with the heavy crude oil−CO₂ or C₃H₈ system. Thus CH₄ was easier to be nucleated from the heavy oil in comparison with CO₂ or C₃H₈. Moreover, the BNI vs. solvent supersaturation data showed that the BNI of the heavy crude oil−CH₄ or C₃H₈ system was slowly increased at lower solvent supersaturations but quickly increased at higher solvent supersaturations, whereas the BNI of the heavy crude oil−CO₂ system was almost linearly increased with solvent supersaturation. Furthermore, the BLI vs. reduced pressure data revealed that in comparison with C₃H₈/CO₂, CH₄ was the most difficult solvent to be liberated from the heavy oil once its bubbles were nucleated. A large amount of CH₄ bubbles could be trapped in the heavy oil to induce the strongest and most stable foamy oil in comparison with C₃H₈ and CO₂. The above experimental findings help to better understand the foamy-oil strengths and stabilities in different heavy crude oil−solvent systems and determine the most suitable solvent to optimize a solvent-based enhanced oil recovery (EOR) process in a heavy oil reservoir.

Introduction

In the 1980s, foamy-oil studies started from some field observations of anomalously good primary production performances in the heavy oil reservoirs in Venezuela, Canada and China, which included high oil production rates, high primary heavy oil recovery factors and low gas−oil ratios (GORs) under the solution-gas drive and foamy-oil flow (Chen and Maini, 2005). The terminology of “foamy heavy oil” is defined as a heavy oil that contains a large number of dispersed small solvent bubbles (Xu, 2007). The foamy heavy oil is formed because a highly viscous heavy oil has an ability to trap the nucleated solvent bubbles during a pressure depletion process in a heavy oil reservoir (Chen et al., 2015; Wang et al., 2008). Theoretically, however, such dispersed solvent bubbles in the heavy oil are not in a thermodynamic equilibrium state and will be liberated from the heavy oil to become the free gas eventually (Sun et al., 2020). The foamy-oil phenomenon begins from the solvent bubble nucleation in the solvent-saturated/diluted heavy oil due to the reservoir pressure depletion and attenuates due to the solvent bubble liberation. Hence, the solvent bubble nucleation and liberation processes have long been studied by many petroleum engineers and scientists worldwide in order to generate a strong and stable foamy-oil flow and enhance heavy oil recovery in a solvent-based enhanced oil recovery (EOR) process in a heavy oil reservoir.

The solvent bubble nucleation occurs because of a solvent supersaturation in the heavy crude oil−solvent system (Sheikha and Pooladi-Darvish, 2012). The solvent supersaturation is defined as a non-equilibrium transient state of a heavy crude oil–solvent system that contains more dissolved solvent than that under the equilibrium conditions at the same pressure and temperature (Bondino et al., 2009a). The dissolved solvent concentration difference denotes the degree of the solvent supersaturation (Kennedy and Olson, 1952). When the pressure of the heavy crude oil−solvent system is lower than its bubble-point pressure during a pressure-depletion process, it is likely to become supersaturated. Then the solvent bubbles begin to be nucleated once the solvent supersaturation reaches certain degree (Kamath and Boyer, 1995). In general, the solvent supersaturation is affected by the characteristics of the heavy crude oil−solvent system (Li et al., 2015), the operating conditions (Sheng et al., 1999), as well as the pressure depletion level and rate (Kortekaas and van Poelgeest, 1991). The pressure drawdown rate is the most extensively studied factor. It has been pointed out that a low pressure drawdown rate will lead to a low solvent supersaturation (Bauget and Lenormand, 2002; Firoozabadi et al., 1994).

The solvent bubble nucleation process has been experimentally and theoretically studied in the past. Typically, the solvent bubble nucleation starts when a certain degree of solvent supersaturation exists in the heavy crude oil−solvent system, which is called the critical solvent supersaturation (Arora and Kovscek, 2003). It is worthwhile to note that the solvent supersaturation state occurs only when the pressure is depleted faster than that at slow equilibrium conditions. The critical solvent supersaturation depends on many factors, including the oil viscosity (Wang et al., 2009), interfacial properties (Moulu, 1989) and pressure depletion rate (Firoozabadi et al., 1994). Some experimental techniques were used to study the solvent bubble nucleation in porous media, such as transparent micromodels (Firoozabadi and Aronson, 1999; Bora et al., 2000), computerized tomography (CT) (Meyer et al., 2007), and ultrasonic measurements (Hoyos et al., 1990). Also, several theoretical and numerical models were developed to study the solvent bubble nucleation process. For example, a mechanistic bubble-population-balance model was formulated to characterize the homogeneous and heterogeneous bubble nucleation processes (Du et al., 2019). A five-component kinetic model was used to examine the bubble ex-solution in porous media, which divided the free-gas formation into four different stages (Shen, 2015). A pore-scale numerical model was proposed to investigate the heterogeneous bubble nucleation and the foamy-oil flow, which included the inter-pore diffusion, capillary-controlled gas expansion, buoyancy-driven migration and oil shrinkage (Bondino et al., 2009b).

The solvent bubbles that are nucleated during a pressure depletion process tend to remain dispersed in the oil phase for a reasonably long time. However, the dispersed solvent bubbles are thermodynamically unstable and will eventually break out or burst from the oil phase (Chen et al., 2020). The bubble separation time depends on the solvent bubble stability. The nucleated solvent bubbles in the light crude oil can be quickly liberated from the oil phase to form the so-called free-gas phase (Yao and Gu, 2021), whereas those in the heavy crude oil can remain dispersed for an extremely long time because of its high viscosity (Zhou et al., 2017). It is these dispersed solvent bubbles that cause the unusual foamy-oil flow and anomalously high oil recovery during the primary production in a heavy oil reservoir. It is found that the solvent bubble liberation can be affected by several factors, including the pressure depletion rate (Zhang, 1999; Wu et al., 2011), oil viscosity (Kumar and Pooladi-Darvish, 2001; Callaghan and Neustadter, 1981), and porous media (Nyre et al., 2008). The pressure drawdown rate was found to be the most important parameter to especially affect the solvent bubble liberation (Zhang, 1999). It was observed in a visual micromodel experiment that smaller bubbles were coalesced to form larger bubbles or continuous gas clusters when the pressure decline rate was low (Wu et al., 2011). On the contrary, larger bubbles or continuous gas clusters could be broken into smaller bubbles when the pressure drawdown rate was high (Kumar and Pooladi-Darvish, 2001). The solvent bubble liberation determines the foamy-oil stability, which strongly depends on the oil viscosity. It was found that the foam stability in the refined mineral oils was increased linearly with the kinematic viscosities of the oils tested (Kumar and Pooladi-Darvish, 2001). Callaghan and Neustadter's experimental data showed that an increase of the bulk crude oil viscosity led to an almost linear increase of the foam lifetime (Callaghan and Neustadter, 1981). The porous media can also strongly affect the solvent bubble liberation. It was experimentally proven that the solvent bubble liberation in a porous medium was much slower than that in a bulk phase (Nyre et al., 2008).

Although the solvent bubble nucleation and liberation processes in the heavy oil have been extensively studied in the past, so far it is still not well understood yet how solvent type will affect these two processes as well as the foamy-oil strength and stability. They can largely determine the ultimate oil recovery in a heavy oil reservoir, where a solvent-based EOR process is applied. In this paper, the solvent bubble nucleation process is studied by conducting the constant-composition-expansion (CCE) tests and the solvent bubble liberation process is studied based on the experimental data from the constant-composition-expansion & compression (CCEC) tests. Then two theoretical analyses are undertaken and two new indexes, namely the bubble nucleation index (BNI) and the bubble liberation index (BLI), are proposed to quantify the solvent bubble nucleation and liberation strengths during the CCE and CCEC tests, respectively. Three different solvents, CH₄, CO₂ and C₃H₈, are studied because CH₄ is the most common solvent in the primary production, whereas CO₂ and C₃H₈ are the commonly used solvents in the solvent-based EOR processes, such as cyclic solvent injection (CSI). In this way, the detailed effects of solvent type on the solvent bubble nucleation and liberation processes in different heavy crude oil−solvent systems are examined and compared. In practice, the oilfield production performance of a solvent-based EOR process strongly depends on the solvent-induced foamy-oil strength and stability during a pressure depletion process in a heavy oil reservoir. Therefore, it is anticipated that this study helps to better understand the solvent-induced foamy-oil evolution and more importantly to optimize the solvent-based EOR process.