Established technologies like primary recovery methods using gas pressure and other natural forces in the reservoir, and secondary recovery by water flooding can only approximately recover one-third of the crude oil present in known reservoirs. The overall recovery of a reservoir is a product of microscopic displacement efficiency, ED, and macroscopic displacement efficiency, EV,. In equation form, E=E_D E_V (1) ED measures the effective ness of the displacing fluid in mobilizing the oil at the pore scale. In water flooding reservoirs, ED is usually around 0.6-0.7, which means the residual oil saturation, Sor, in the regions contacted by the displacing water is 0.3-0.4. The primary reason for this high residual oil saturation is the capillary trapping at the pore throat, since in most sandstone reservoir water is the wetting phase. Researchers have recognized that a dimensionless capillary number N_c=μν/σ controlled the residual oil saturation, where ν is the interstitial velocity, μ is the viscosity of aqueous solution and σ is the oil-water interfacial tension (IFT) (Taber 1969; Stegemeier 1977; Melrose 1974; Foster 1973). Correlations between residual oil saturation and capillary number find as the capillary number increases to 10-2 magnitude, residual oil saturation can be reduced to lower than 0.05 (Abrams 1975). Surfactant is such a chemical that is added into aqueous solution to reduce oil-water IFT, hence increase the capillary number. Surfactant flood processes have been well developed in the past decades. Various new surfactants and formulations were invented and tailored to fit the reservoir of interest. The general procedures of a surfactant flooding project is shown in Figure 1-1. A candidate reservoir suitable for surfactant flooding is firstly screened. And formulation is designed according to the reservoir conditions at lab. Considered reservoir conditions include brine salinity, reservoir temperature and oil properties. Then coreflood is conducted to evaluate the displacement efficiency of the designed formulation. In next step, coreflood simulation is performed to explain the coreflood results and obtain parameters that can capture the multiphase displacement process. These parameters are then used as input in pilot test simulation to predict the oil recovery in field scale, hence the economics can be evaluated.

Figure 1-1 Procedures of a surfactant flooding project

In these procedures, microemulsion phase behavior plays a critical role. During formulation design, microemulsion phase behavior tests are conducted to screen candidate surfactant and optimize the surfactant formulation, which is time consuming and highly dependent on the experiences of formulation researchers. And in coreflood process, different microemulsion type corresponds to different displacement mechanism. Moreover, it is Type III microemulsion that most efficient in reducing oil-water IFT and mobilized trapped oil. In compositional surfactant flooding simulators, microemulsion phase behavior model is an important package to calculate phase composition, phase saturation and interfacial tension, etc. Therefore, an incorrect phase behavior model or improper input will lead to an unreliable simulation results. This work aims to provide solutions to these problems in surfactant flooding technology, by using a novel hydrophilic-lipophilic deviation (HLD) and net-average curvature (NAC) called thereafter HLD-NAC equation of state. In this dissertation, Chapters 2 and 3 focus on microemulsion modeling. Chapter 2 validated the HLD-NAC model for surfactant/brine/crude oil systems. Microemulsion phase behavior of various formulations were reproduced by using only one fitting parameter, the surfactant tail length L. The contribution has appeared in the Journal of Petroleum Science and Engineering. Chapter 3 further evaluated the predictability of the HLD-NAC equation of state. Without using any matching parameter, four optimum formulations and corresponding microemulsion phase behavior are predicted with the HLD-NAC model using laboratory characterized parameters as input. This contribution has been accepted in 2016 SPE Improved Oil Recovery Symposium and been submitted to SPE Journal for peer review. Chapters 4 and 5 combine the HLD-NAC model with numerical and analytical methods to study its advantages in modeling surfactant flooding. In Chapter 4, a new chemical flooding simulator is developed by implementing the HLD-NAC EOS into UTCHEM. An algorithm is invented to describe composition distribution on a ternary phase diagram using surfactant, water and oil as the pseudo-component. As a replacement of Hand’s rule phase behavior model in UTCHEM, the HLD-NAC EOS shows various advantages in both physical significance and computational efficiency. Chapter 5 attempts to analytically study three-component, two-phase surfactant flooding by coupling the HLD-NAC EOS and coherent theory. The analytical solution is compared with the results from numerical simulator developed in Chapter 4, to prove that the algorithm is correctly implemented into UTCHEM. Using the analytical solution, the effects of phase behavior dependent parameters on surfactant flooding can by systemically studied. Finally, Chapter 6 presents some concluding remarks of this work and recommendations for the future studies.