AmericanChemicalSociety.com

Introduction

Crude oil mobility in porous reservoirs is controlled by several physical parameters such as capillary force and surface tension existing between the oil and solid media, and the geometry of the pore spaces. As a result, petroleum mobilization is controlled by the heterogeneity of the porous medium and the related distribution of oil. The outcomes of this research are beneficial to understanding the pore-scale interfacial processes controlling tertiary oil recovery from existing reservoirs. The purpose of this research is to quantify the oil distribution as a function of crude oil fraction and heterogeneity of the porous medium.

Multiple columns were packed with three different types of sediments with increasing heterogeneity. The columns were saturated with water, and injected with three different fractions of crude oil. The columns were flushed with anionic surfactants in two episodes. Utilizing the Advanced Photon Source Facility at Argonne National Laboratory, synchrotron X-ray microtomography was used to capture high resolution (9-10 µm) images of the oil and water distributions within the columns before and after each flooding event. Three fractions of crude oils (Chevron and BP) were chosen as model oil-phase liquids with high (14.0°), medium (29.6°) and light (41.4°) API gravity. A 0.1 vol% alkylpropoxy sulfate solution was used as the anionic surfactant of choice for the flooding experiments. This surfactant was chosen for its ability to develop low interfacial tension at very low concentration and because it is not sensitive to insensitive to salinity.

Results and Discussions

Selected results are presented for a series of homogeneously packed columns (Cu =1) distributed with 3 different oil fractions after 2 episodes of sequential surfactant flushing. Results show that all three oil fractions show significant differences in oil morphology and distribution after surfactant flushing. Overall, oil recovery behavior after sequential surfactant flushing was similar for the light and medium oil fractions whereby approximately 50% of the oil was removed after the first flushing episode. After the second surfactant flush (5 total pore volumes), there was no remaining light or medium oil within the porous-medium filled columns. This indicates that nearly complete oil recovery is likely if the system is flushed long enough, at least for light-fraction and medium-fraction oils distributed within a homogeneous porous medium. Conversely, the heavy-fraction oil showed no significant recovery after the first surfactant flushing episode. However, after the second surfactant flush (5-PV) approximately 13% of the initial oil volume was recovered.

The distribution of blobs for the light and medium fraction oils varied considerably from the initial residual saturation conditions to after the first surfactant flush. In general, the distribution of both oil fractions (light and medium) increased in heterogeneity over this first flushing episode. The coefficient of variation, which describes the heterogeneity of the oil distribution, increased from 2.45 to 27.5 for the light-fraction oil. The light oil fraction distribution changes most drastically in terms of blob heterogeneity from initial saturation to after the first flushing event. In addition, the overall size fraction of the light oil blobs decreased significantly after the first flushing event. The median diameter changes from 0.275 mm (initial) to 0.047 mm (after first flush). The coefficient of variation for the medium oil distribution increased from 4.64 to 9.39 from initial saturation conditions to after the first flushing event. The overall size fraction of the medium oil blobs did not decrease substantially after the first flushing event. In contrast to the light and medium oil fractions, the heavy oil exists as a continuous fluid phase essentially through all stages of flushing and therefore a representative variation in individual blob distribution cannot be effectively quantified.

Morphology of the oil blobs was quantified using a relationship that characterizes the blob shapes in terms of blob deviations from a perfect sphere.  As the oil blobs deviate from this relationship the blobs become more ganglia-like, possessing larger surface area to volume characteristics. For the light oil fraction, the larger blobs deviate most significantly from a sphere, and an overall increase in very small spherical blob sizes result after the first flushing event (2-PV). The larger blobs from the medium oil fraction change in morphology as well, however not as significantly as the lighter oil fraction. For the medium oil blobs, an overall increase in small spherical blob sizes also result after the first surfactant flushing event (2-PV).  The heavy oil-fraction morphology is primarily present as a continuous ganglia-type distribution within the column, even after each surfactant flushing event (i.e. 2 PV and 5 PV). As a result, there is no strong trend observed in the blob morphology distribution. This preliminary result appears to have significant control on the poor recovery observed for the heavy oil fraction.

Acknowledgements

Acknowledgment is made to the Donors of the American Chemical Society Petroleum Research Fund for support (or partial support) of this research.

Table 1: Data for light crude oil before and after 2PV surfactant flush

Table 2: Data for medium crude oil before and after 2PV surfactant flush

Table 3: Data for heavy crude oil before and after 2PV and 5PV surfactant flush