Flooding alkali

The purpose of alkali flooding (Jennings et al., 1974) is to introduce alkali into a reservoir where it can react with organic acids in the oil to produce organic salts, which act as surfactants. The surfactants (or petroleum soaps ) generated reduce the surface tension between the oil and water and this in turn reduces the level of capillary trapping of the oil. Thus, more oil is recovered because less of it remains trapped in the formation s pore spaces. 

This type of flood can be successful only if, as the fluid moves through the reservoir, a sufficient amount of the alkali remains in solution to react with the oil. Reaction of the flood with minerals and fluid in the reservoir, however, can consume the flood s alkali content. Worse, the reactions may precipitate minerals in the formation s pore space, decreasing permeability near the wellbore where free flow is most critical. A special problem for this type of flood is the reaction of clay minerals to form zeolites (Sydansk, 1982). 

The effectiveness of alkali flooding, and, in fact, most reservoir treatments, varies widely from formation to formation in a manner that is often difficult to predict. Quantitative techniques have been applied to model the migration and consumption of alkali as it moves through a reservoir (e.g., Bunge and Radke, 1982 Zabala et al., 1982 Dria et al., 1988). There have been fewer attempts, however, to predict the specific chemical reactions that might occur in a reservoir or the effects of the initial mineralogy of the reservoir and the composition of the flood on those reactions (Bethke et al., 1992). 

To consider how such predictions might be made, we model how three types of alkali floods might affect a hypothetical sandstone reservoir. The floods, which are marketed commercially for this purpose, are NaOH, Na2C03, and Na2Si03. We take each flood at 0.5 N strength and assume that reaction occurs at a temperature of 70 °C. 

The reservoir rock in our model is composed of quartz grains, carbonate cement, and clay minerals in the following proportions, by volume  

The effectiveness of alkali flooding, and, in fact, most reservoir treatments, varies widely from formation to formation in a manner that is often difficult to predict. Quantitative techniques have been applied to model the migration and 

---

In this chapter, in an attempt to devise methods for helping to foresee such unfavorable consequences, we construct models of the chemical interactions between injected fluids and the sediments and formation waters in petroleum reservoirs. We consider two cases the effects of using seawater as a waterflood, taking oil fields of the North Sea as an example, and the potential consequences of using alkali flooding (i.e., the injection of a strong caustic solution) in order to increase oil production from a clastic reservoir. 

Fig. 30.3. Variation in pH during simulated alkali floods of a clastic petroleum reservoir at 70 °C, using 0.5 N NaOH, Na2C03, and Na2Si03 solutions. Pore fluid is displaced by unreacted flooding solution at a rate of one-half of the system s pore volume per day.

Fig. 30.4. Changes in the volumes of minerals in the reservoir rock during the simulated alkali floods (Fig. 30.3) of a clastic petroleum reservoir using NaOH, Na2CC>3, and Na2SiC>3 solutions. Minerals that react in small volumes are omitted from the plots. Abbreviations Anal = analcime, Cc = calcite, Daw = dawsonite, Dol = dolomite, Kaol = kaolinite, Muse = muscovite, Parag = paragonite, Phlog = phlogopite, Qtz= quartz, Trid = tridymite.

In the simulations, a significant fraction (about 50% to 80%) of the alkali present in solution is consumed by reactions near the wellbore with the reservoir minerals (as shown in Reaction 30.6 for the NaOH flood), mostly by the production of analcime, paragonite, and dawsonite [NaAlC03(0H)2]. In the clastic reservoir considered, therefore, alkali floods might be expected to cause formation damage (mostly due to the precipitation of zeolites) and to be less effective at increasing oil mobility than in a reservoir where they do not react extensively with the formation. 

Amy Berger helped me write Chapter 10 (Surface Complexation), and Chapter 31 (Acid Drainage) is derived in part from her work. Edward Warren and Richard Worden of British Petroleum s Sunbury lab contributed data for calculating scaling in North Sea oil fields, Richard Wendlandt first modeled the effects of alkali floods on clastic reservoirs, and Kenneth Sorbie helped write Chapter 30 (Petroleum Reservoirs). I borrowed from Elisabeth Rowan s study of the genesis of fluorite ores at the Albigeois district, Wendy Harrison s study of the Gippsland basin, and a number of other published studies, as referenced in the text. 

Table 22.2 summarizes the simulation results. In each case, the flood dissolves clay minerals and quartz. The simulations, however, predict the production of significant volumes of analcime, in accord with the observation that zeolites are prone to form during alkali floods. The volume of analcime produced in each case is sufficient to offset the volumes of the dissolved minerals and lead to a net decrease in porosity. Of the simulations, the Na2SiO3 flood leads to the production of the most analcime and to the greatest loss in porosity. 

---