Enhanced Oil Recovery with Surfactant Flooding

Enhanced oil recovery (EOR) is being increasingly applied in the oil industry and several different technologies have emerged during, the last decades in order to optimize oil recovery after conventional recovery methods have been applied. Surfactant flooding is an EOR technique in which the phase behavior inside the reservoir can be manipulated by the injection of surfactants and co-surfactants, creating advantageous conditions in order to mobilize trapped oil.

Correctly designed surfactant systems together with the crude oil can create microemulsions at the interface between crude oil and water, thus reducing the interfacial tension (IFT) to ultra low (0.001 mN/m), which consequently will mobilize the residual oil and result in improved oil recovery. This EOR technology is, however, made challenging by a number of factors, such as the adsorption of surfactant and co-surfactant to the rock during the injection and chromatographic separation of the surfactant and co-surfactant in the reservoir.

Therefore it would be a significant step forward to develop single surfactant systems, as this would minimize the consequences of adsorption and separation. Furthermore the surfactants must be resistant to and remain active at reservoir conditions such as high temperatures, pressures and salinities.

Understanding the underlying mechanisms of systems that exhibit liquid-liquid equilibrium (e.g. oil-brine systems) at reservoir conditions is an area of increasing interest within EOR. This is true both for complex surfactant systems as well as for oil and brine systems. It is widely accepted that an increase in oil recovery can be obtained through flooding, whether it is simple waterflooding, waterflooding where the salinity has been modified by the addition or removal of specific ions (socalled “smart” waterflooding) or surfactant flooding.

High pressure experiments have been carried out in this work on a surfactant system (surfactant/ oil/ brine) and on oil/ seawater systems (oil/ brine). The high pressure experiments were carried out on a DBR JEFRI PVT cell, where a glass window allows observation of the phase behavior of the different systems at various temperatures and pressures inside the high pressure cell.

Phase volumes can also be measured visually through the glass window using precision equipment. The surfactant system for which an experimental study was carried out consisted of the mixture heptane, sodium dodecyl sulfate (SDS)/ 1-butanol/ NaCl/ water. This system has previously been examined at ambient pressures and temperatures but this has been extended here to pressures up to 400 bar and to slightly higher temperatures (40 °C, 45 °C and 50 °C).

Experiments were performed at constant salinity (6.56 %), constant surfactant-alcohol ratio (SAR) but with varying water-oil ratios (WOR). At all temperatures it was very clear that the effect of pressure was significant. The system changed from the two phase region, Winsor II, to the three phase region, Winsor III, as pressure increased.

Increasing pressures also caused a shift from the three phase region (Winsor III), to a different two phase region, (Winsor I). These changes in equilibrium phase behavior were also dependent on the composition of the system. A number of different compositions of the surfactant system were studied. The effect of increased pressure became more significant when combined with increasing temperature.

The experiments performed on the oil/ seawater systems were similar to the high pressure experiments for the surfactant system discussed above. Oil was contacted with different brine solutions with varying sulfate concentrations at a WOR of 70/30. A series of experiments were performed on two crude oils; a Latin American crude oil and a Middle East crude oil.

The two crude oils showed significantly different phase behavior when exposed to elevated temperatures and pressures. The Latin American crude showed a decrease in oil viscosity with an increase in sulfate concentration in the brine solution after contacting in the PVT cell. The Middle East crude oil formed emulsions in the PVT cell with increasing temperature and pressure which was more pronounced at higher sulfate concentrations.

Further characterization of the two crude oils using gas chromatography and SARA analysis confirmed that the heavier components in the crude oils, (in the case of the Latin American crude oil), are correlated to the observed decrease of viscosity, where the viscosity decrease may be explained from change of the shape of the heavy components with the increase in sulfate concentration after contacting at high pressures and temperatures.

A third model system consisting of heptane and seawater solutions was also studied. This system formed emulsions in the PVT cell similar to the Middle East crude oil, which indicates that the lighter components in the Middle East crude oil (compared to the Latin American crude oil) are responsible for the observed formation of emulsions.

The final part of the thesis is a phase behavior modeling study of alkane/ alkanol/ water systems relevant for surfactant flooding. Existing thermodynamic models, such as equations of state, while able to predict and correlate phase equilibrium in two liquid phases (with varying degrees of success) cannot account for the formation of a microemulsion phase.

The presence of electrolytes in the surfactant systems further complicates the problem, and the incorporation of electrolytes into equations of state is a problem that, while old, has not been satisfactorily solved. Furthermore the effect of pressure is presently not well accounted for. The simplified PC-SAFT equation of state is used to model the phase behavior of several binary systems.

Typically, introducing a small binary interaction parameter, kij, results in good correlations. However, the interaction parameter must be fitted to each individual binary system. A glycol ether/ water binary system was also included in the phase equilibrium modeling study. This system is so difficult to model adequately that an additional binary interaction parameter, lij, was introduced to see if the correlations of this system could be improved – especially with regard to the significant effect of pressure on the phase behavior.

It was concluded that this additional binary parameter was not sufficient to substantially improve the performance of the model.