A Laboratory Study of Microemulsion Flooding (includes associated papers 6395 and 6396 )

Abstract

Abstract According to the literature, microemulsions haveopen been evaluated using core floods run at highflow rates sometimes much higher than could beachieved in a reservoir, except close to an injectionwell. The rationale for this is, presumably, thatmicroemulsion flooding is a "miscible-type"process.However, it is necessary to determine what is meantby these words and what relation they bear to a"miscible" process. In this paper an effort hasbeen made to experimentally distinguish misciblefrom immiscible aspects of microemulsion floodingand to evaluate the contributions of each to oilrecovery. Some of the principal conclusions relateto the rate dependence of microemulsion flooding, bypassing of resident fluids, nature of the externalphase, surfactant retention, and mechanisms of oilrecovery. Introduction Results of a surfactant field test conducted inthe Loudon field in Illinois demonstrated that afresh-water preflush could not be employed to"condition" the reservoir so as to provide optimalsalinity before surfactant injection. Even thoughthe Loudon pilot sand was not highly heterogeneous (Lorenz coeffecient = 0.35), a 1.4-PV preflush offresh water did not adequately displace the highsalinity resident brine from the reservoir. As aresult, the surfactant was exposed to higher salinitiesthan could be tolerated. It was concluded that"pockets" of highly saline water were flushed outbecause of mobility improvement when surfactantand polymer slugs effectively swept the reservoir.it follows that it will be necessary to design theentire surfactant flooding process on the basis ofresident brine and oil; i.e., both the surfactant slugand polymer slug must be effective in resident fluids.Preflushes can be successfully used in laboratory floods, but these floods do not model reservoirmixing.It would be desirable, then, to devise surfactantsystems that are highly effective over a large rangeof salinities. Not having these, it is necessary todesign existing for resident salinities.These observations have affected experimental procedures used in this work. Specifically, the samebrine that is used as resident water and final drivewater is used to prepare the microemulcon andpolymer slug. This mode initial residual oil inthe core is always the same as the oil component ofthe microemulsion slug. This mode of operation hasthe added advantage that laboratory core-floodingresults are more readily interpreted than would beotherwise possible.Microemulsion displacement of oil and water is acomplex process not well understood in all itsdetails. A number of questions have been raised inthe literature and some of these require further, more detailed study. For example, are there rateeffects? Does micellar structure type make anydifference? That is the duration of miscibledisplacement? What role is played by immiscibledisplacement after slug breakdown? What are theeffects of bypassed water and oil and do thesepersist in long core floods? Some of these questionswill be answered here, others explored, but therewill still remain much to be done beforemicroemulsion flooding is thoroughly understood. CONCEPTS AND DEFINITIONS Microemulsion displacements are referred to asmiscible, or of a miscible-type, with some misgivings, possibly in view of the micellar structures alwayspresent. For some, it is difficult to think of adisplacement as miscible when just below thethreshold of visibility there are discontinuities having dimensions on the order of hundreds ofangstroms. Thus it appears that the concept ofmiscibility refers to some scale and has to do with the presence or absence of interfaces. However, the scale is arbitrary and it is convenient to usevisible white light as the measure.The following brief discussion is included forpurposes of consistent use of terminology. Aspreviously, one is concerned with the approximateternary representation of oil-water-surfactant systems(Fig. 1). SPEJ P. 87^