1 Vc

Figure 10-8b. Salinities less than 20 g/l, impairment caused by dispersion, and by deflocculation at salinities less than 1 g/l.

where k is the observed permeability, k0 the permeability when no clay is present in the pack, p is the percent clay, c the concentration of the sodium chloride in grams per liter, and A and B are constants.

At salinities less than 20 g/l the clays became "unstable" (i.e., dispersed). Bardon and Jacquin point out that 20 g/l is almost the identical concentration of sodium chloride at which Norrish observed the sudden expansion in clay lattice spacing, and close to the salinity which marks the change from aggregation to dispersion, as measured by the clay volume and optical density tests (23 g/l, see the section on aggregation and dispersion, Chapter 4).

Finally, they observed that when distilled water was flowed through the pack, it plugged completely if the pack contained 10% montmorillonite, but if the pack contained only 5% montmorillonite, the permeability increased, and clay was discharged from the end (see Figure 10-8b). These sharp changes in behavior are understandable when it is remembered that at salinities below about 1 g/1 clay particles are deflocculated as well as dispersed, and therefore much more mobile than when flocculated and dispersed. (See Chapter 4)

Similar tests with natural sandstones containing illite or kaolinite agreed qualitatively with the results obtained with the sand packs and montmorillonite (see Figure 10-9).

Most natural sands and sandstones contain considerably less clay minerals than did the packs in Bardot and Jacquin's experiments; for example, sands in southern California contain from 1 to 2% clay minerals.I4b Consequently, the decrease in permeability caused by crystalline swelling is much less than that shown in Figure 10-8a. Crystalline swelling causes the volume of montmoril-lonite to no more than double (see "Clay Swelling Mechanics," Chapter 4); if the clay is present only as a thin coating on the pore walls, doubling the volume has little effect on the permeability. On the other hand, the flow paths through natural sand and sandstone formations are more tortuous than those in artificial sandpacks; consequently, dispersed and deflocculated clay particles are usually trapped, with consequent sharp drop in permeability. Thus the shape of the curves in Figure 10-9 is typical of those obtained with natural sands and sandstones, even when montmorillonite is present.

Work by other investigators15 •17 ■18'19 leaves no doubt that the decrease in permeability at low salinities is caused by the displacement and dispersion of the clay or other fines from the pore walls by the invading fluid, and by subsequent trapping at the pore exits (as shown in Figure 10-10). This mechanism is now commonly referred to as clay blocking. The mechanism is analogous to the formation of external mud filter cakes, but in this case multitudinous internal microfilter cakes are formed on the pore exits. The very low permeabilities observed under the deflocculating conditions of distilled water floods are analagous to the low permeabilities of the filter cakes of deflocculated muds compared to those of flocculated muds. The effect of deflocculation can also be shown by flooding sand packs containing montmorillonite with sodium chloride brines containing a deflocculant such as sodium hexametaphosphate. As mentioned in Chapter 4, sodium hexametaphosphate raises the flocculation point of montmorillonite in sodium chloride brines to 20 g/1, so that the particles are both deflocculated and dispersed at lower salinities. Consequently, permeabilities are much lower than those obtained when no hexametaphosphate is present (as shown in Figure 10-11). The point has considerable practical importance because it means that deflocculants such as the complex phosphates and tannates should never be added to dispersed muds when drilling through water-sensitive formations.

Another factor that affects clay blocking is the rate of reduction of salinity. Experiments by Jones20 showed that permeability reduction was much less if the salinity of the brine floods was reduced gradually in steps, rather than an abrupt change from a concentrated brine to fresh water. Mungan16 showed

0 0

Post a comment