Surfactant Flood Process Design for Loudon

Abstract

Summary Laboratory steps required to design a successful surfactant flooding processfor the high-salinity Loudon field are described. These steps involved (1)phase-behavior tests to identify two synthetic surfactant components; (2) Bereacoreflood tests to establish the desirability of incorporating polymer in themicroemulsion and to optimize blend ratio, microemulsion salinity, drive-watersalinity, and bank size; and (3) Loudon rock corefloods to provide a finalcheck in reservoir core material. The process recovered significant quantitiesof residual oil in two 1-acre-pilot tests without the use of a cosurfactant, acosolvent alcohol, or a low-salinity preflush. Introduction The Loudon field, operated by Exxon Co. U.S.A. in Fayette County, IL, is aprime tertiary recovery target from a technical standpoint. The field is in anadvanced stage of depletion after 13 years of primary production and 38 yearsof waterflooding. After waterflooding is completed, nearly one-half theoriginal oil in place will likely remain unrecovered. Most of the reservoircharacteris-tics are favorable for state-of-the-art microemulsion (surfactant)flooding-i.e., low temperature (78 deg. F), moderate permeability (140 mdaverage), low oil viscosity (5 cp), and fairly low clay content (- 3 %). Thehigh salinity of the resident brine-about 104,600 ppm (10.5%) total dissolvedsolids (TDS), including more than 4,000 ppm of divalent ions (see Tablel)-represents a significant challenge for surfactant flooding. An earlier surfactant flood pilot I conducted at Loudon in 1969 used alarge-volume low-salinity preflush to displace resident brine and to reducesalinity to a level required for efficient oil displacement by the petroleumsulfonate surfactant system. The main conclusion reached from petroleumsulfonate surfactant system. The main conclusion reached from that test, whichrecovered only about 15 % of the residual oil in the test area, was thatpreflushes are likely to be ineffective unless the surfactant system iseffective over a broad range of salinities. Much of the subsequent surfactantflooding research was directed at surfactants that are effective inhigh-salinity reservoirs without requiring a preflush. preflush. This paperdescribes the final design stages of a high-salinity surfactant floodingprocess for the Loudon field. We include a brine discussion of phase behavior, but a complete treatment of phase behavior to identify the generic surfactantstructure is beyond the scope of this paper. The bulk of work reported here isdivided into two parts dealing paper. The bulk of work reported here is dividedinto two parts dealing with Berea coreflood tests and with Loudon rockcoreflood tests. The section concerned with Berea corefloods begins with adiscussion of mobility control, and this is combined with oil recoverycomparisons between small- and large-cross-section (wide-model) corefloods toestablish the desirability of incorporating polymer in the microemulsion. Subsequent stages of the process design are discussed essentially inchronological order, including optimization of surfactant blend, microemulsionsalinity, drive-water salinity, and bank size. Phase Behavior Phase Behavior The process design work detailed in this paperused surfactants selected from a group of compounds described in Ref. 3 andrepresented by the general formula R1O(C3H6O)m(C2H4O)nYX. In the specificmolecules considered here, R is an isotridecyl alcohol radical, m and n havevalues from 1 to 6, Y is the hydrophilic sulfate group, and X is the monovalentsodium cation. These compounds were manufactured by reacting propylene oxideand then ethylene oxide with isotridecyl alcohol in two separate steps, followed by sulfation and neutralization with sodium hydroxide. Two majoradvantages of these surfactants, detailed in the patent, are that (1) optimalsalinity is relatively high and is not a strong function of surfactantconcentration or divalent-ion concentration and (2) chromatographic separationof components is expected to be minimal with negligible effect on surfactantperformance. Fig. 1 shows optimal salinity contours on a grid representing differentnumbers of propoxyl and ethoxyl groups in the surfactant molecules. (The valuesm and n are actually averages for distributions of m and n in each singleproduct or blend of surfactants.) Generally, when other parameters are heldconstant, optimal salinity increases with more ethoxyl groups and decreaseswith more propoxyl groups. The contour lines are labeled as a percentage of TarSprings brine (TSB) composition (shown in Table 1). The salinity of TSB is verysimilar to that of the resident brine -M the Loudon reservoir. From the grid in Fig. 1, it is evident that an infinite number of combinations of m and n couldgive a surfactant system with an optimal salinity near 100% TSB. The systemselected for use in the Loudon microemulsion formulation consisted of a blendof the two surfactants identified on Fig. 1: i-C13H27O(PO)4(EO)2SO3Na andi-C13H27 O(PO)3(EO)4SO3Na. This selection was based on many preliminary phase-behavior tests andlaboratory corefloods conducted before the final design steps described in thispaper. A discussion of these tests is beyond the scope of this paper. A blendof surfactants provides a means for quality control through adjustment of theblend ratio to account for compositional differences that may result fromunintentional variations in manufacturing operations. Later sections in thispaper discuss situations where readjustment of the optimum blend ratio wasnecessary. A 60/40 blend of the surfactants PL612 and PL613 [(PO)4(EO)2 and (PO)3(EO)4tridecyl alcohol sulfates, respectively] was required to give an optimalsalinity near Loudon resident-brine salinity. Fig. 2 shows solubilizationparameters vs. salinity for microemulsions made with two different oils at a1:1 WOR and 2 % (wt/vol) of the 60/40 blend. To keep microemulsion costs low, diesel oil initially was selected as the oil component for microemulsionformulation. The diesel oil used to develop Fig. 1 gave an optimal salinitynear that obtained with Loudon crude oil at 78 deg. F. However, somesignificant differences in phase behavior were observed. Diesel oil producedclassical phase-behavior transitions of lower- to middle- to upper-phasemicroemulsions with increasing salinity, as shown in Fig. 2a. On the otherhand, Loudon crude oil produced nonclassical phase behavior; as salinityincreased, Loudon oil uptake by the lower-phase microemulsion increased andthen dropped abruptly, as shown in Fig. 2b. Further salinity increases produceda middle phase that contained little or no solubilized oil. Because no trueoptimal salinity exists for such nonclassical systems, we arbitrarily took" optimal salinity" to be the salinity where the discontinuity in oilsolubilization occurred. In Fig. 2b, this salinity is 105% TSB. At reservoirsalinity (100% TSB), the oil-solubilization parameter is about eight. Theoil-solubilzation parameter with diesel at reservoir salinity is about 17.