Intermediate wettability by chemical treatment

Journal of Petroleum Science and Engineering 24 Ž1999. 123–130 www.elsevier.nlrlocaterjpetscieng Intermediate wettability by chemical treatment M. Fleury a a,) , P. Branlard b, R. Lenormand a , C. Zarcone c Institut Français du Petrole, 1 et 4, aÕenue de Bois-Preau, BP 311, 92506 Rueil-Malmaison Cedex, France ´ b RHODIA Silicones, France c Institut de Mecanique des Fluides de Toulouse, France ´ Abstract To study the effect of wettability on recovery, we used a chemical additive that changes the surface properties of natural rock samples while keeping interfacial tension and viscosity constant. The wettability change is weaker than for other chemical treatments Že.g., silane. and is therefore closer to reservoir cases. The chemical additive is a potassium methyl siliconate that is soluble in water. A polymerization of the adsorbed layer on the pore surface is obtained when the pH is lowered near neutral. Outcrop sandstone samples were treated either with or without the presence of oil Žat Swirr. using carbon dioxide as a neutralizer. Efficiency of the treatment is determined mainly by comparing centrifuge Žnegative. imbibition capillary pressure curve on treated and untreated samples. Micropore membrane technique was also applied to measure the positive imbibition capillary pressure curve. The main results are Ž1. the final oil saturation ŽSor. is reduced from 0.4 to 0.3 and 0.1 depending on how the treatment was performed; Ž2. there are only minor modifications on the primary drainage capillary pressure curve; and Ž3. water relative permeability at Sor measured after centrifuge forced imbibition is not modified by the treatment. q 1999 Elsevier Science B.V. All rights reserved. Keywords: capillary pressure; wettability; siliconate; relative permeability 1. Introduction Wettability has a strong impact on water-flooding processes and it is well-established that better recovery rates are obtained for intermediate or mixed wetting conditions Žfor a review, see Anderson, 1987a,b; Jadhunandan and Morrow, 1995.. However, reservoir wettability estimation and restoration in the laboratory are still controversial matters because full reservoir conditions Žtemperature, pressure, fluids, ) Corresponding author. Tel.: q33-1-4752-6285; Fax: q33-14752-7072. E-mail address: marc.fleury@ifp.fr ŽM. Fleury.. etc.. cannot be reproduced in most laboratory experiments, not to mention the cost. Wettability changes due to the presence of crude oils are very complex chemical processes dependent also on the rock surface properties Že.g., type of clays or minerals.. In this context, this work has two objectives: – – to obtain intermediate wettability using a simple and rapid laboratory procedure with the aim of testing laboratory experiments Žflooding, capillary pressure and resistivity measurements, etc.. and, to explore the possibility of field wettability treatment to improve recovery at the field scale. 0920-4105r99r$ - see front matter q 1999 Elsevier Science B.V. All rights reserved. PII: S 0 9 2 0 - 4 1 0 5 Ž 9 9 . 0 0 0 3 6 - 4 124 M. Fleury et al.r Journal of Petroleum Science and Engineering 24 (1999) 123–130 In addition to the well-known and efficient method of aging with crude oil Žfor instance Zhou et al., 1993., different chemicals have been tested to change the wettability of a rock, either in the laboratory or during reservoir flooding. Several studies use organosilanes to change the wettability from water wet to intermediate or oil wet. The final wettability and persistence of the treatment depend on the temperature and the conditions of treatment. Coley et al. Ž1956. used a solution of ‘‘Drifilm’’ Ždimethyldichlorosilane. in toluene at different concentrations. They studied the effect of wettability on capillary pressure curves, relative permeabilities and oil recovery. Fatt and Klikoff Ž1959. also used Drifilm but only in the vapor phase to treat sand and to make media of well-defined fractional wettability by mixing treated and non treated sand. Skauge and Aarra Ž1993. also used Drifilm, in hexane solutions, to study the effect of wettability on oil recovery during a WAG process. The treatments were performed by injecting several pore volume of the solution. Takach et al. Ž1989. showed that organosilanes were more efficient and more stable in vapor phase than in solution. They also showed that the optimum temperature was in the range 2208C–3508C. A more efficient method is by ‘‘silanation’’, i.e., the chemical grafting of silane molecules on the silicate rock surface ŽLombard and Lenormand, 1993; Lombard et al., 1991.. This method involves boiling of the rock in sulfonic acid to obtain the maximum of reactive sites, and then immersing the core in a boiling solution of silane in toluene. This method leads to very strong oil wettness, much stronger than that in reservoir situations. Other chemicals have also been tested. Morrow et al. Ž1973. screened several polar compounds present in crude oil that may be responsible for the change of wettability of reservoir rocks. The more efficient alteration of dolomites was obtained with octanoic acid. For laboratory wettability alteration, Maini et al. Ž1986. used ‘‘Quilon-S’’ to obtain a stable intermediate wettability from water-wet cores. Quilon is a water soluble fatty acid ŽC14–C18. polymerized and bonded to the rock surface by chromium. The Quilon saturated core was cured at 958C for at least 5 days. Some wettability treatments focused on oil recovery during water-flooding as an alternative to reduction of surface tension by surfactants. The principle of the treatment is to alter the rock wettability from water-wet to neutral that leads to maximum recovery ŽJadhunandan and Morrow, 1995.. Michaels and Stancel Ž1964. and Michaels and Porter Ž1965. showed that recovery in water-wet sand pack was improved by the injection of an aqueous hexylamine slug. The mechanism was a temporary change towards oil wettability, followed by a desorption leading to water wetness. Leach et al. Ž1962. also obtained recovery improvement in laboratory and field tests using a slug of sodium hydroxide. Some chemical treatments are designed to change wettability from oil-wet to water-wet. Shen et al. proposed trisodium trimetaphosphate in aqueous solution. The need for wettability reversal is also crucial for oil-wet fractured reservoirs. Wettability reversal agents were then used to change matrix wettability to preferentially water wet. Oil can then be recovered by spontaneous imbibition. A list of agents based on alkali salts or hydroxides is provided by Sarem et al. Ž1993.. In this work, we studied a water-soluble siliconate that is easy to use for laboratory and field applications. The change of wettability is viewed essentially through the change of capillary pressure curves, especially the imbibition curves where surface effects are in competition with porous structure. This approach is similar to classical wettability tests ŽAmott-IFP, USBM. when aging procedures are used to change wettability. Schematic capillary pressure curves for water-wet and intermediate wet situations are shown in Fig. 1. Fig. 1. Typical capillary pressure curves in drainage and imbibition for two wettability states. In the water-wet situation, there is no change of saturation for negative capillary pressure. M. Fleury et al.r Journal of Petroleum Science and Engineering 24 (1999) 123–130 2. Description of the chemical additive It is well-known that silane or silicic acids derivatives can change wettability and they are utilized in various industrial applications as water repellents for construction materials Žairrwater. as proposed by Elliot and Krieble Ž1950. and Bosh et al. Ž1976.. The chemical additive used was potassium methylsiliconate that is available commercially ŽRhodorsil w .. It is used basiSiliconate 51T from Rhone-Poulenc ˆ cally to prevent capillary rise in building materials such as tiles or concrete. In practice, the product is diluted before use and sprayed on or injected to the porous material to be treated. An important property of our application is that the additive is water-soluble. As a commercial product, the Siliconate 51T is delivered at a high concentration Ždry material content of 47% - active material content of 28%.. Therefore, the pH is high Žabout 13.. It is usually diluted down to 1–5%. We checked that interfacial tension is not modified in this range of concentration. For example, the measured interfacial tension between brine and dodecane is 36 " 2 mNrm, while the interfacial tension between brine q siliconate at 1% and dodecane is 35 " 2 mNrm. The basic reaction mechanism is indicated in Fig. 2. A polymer chain is formed when the pH is lowered below 7, using carbon dioxide for example Žan acid could also be used.. The minimum concentration should not be smaller than about 1% for the polymerization to occur inside or outside the porous 125 medium. At that concentration, the pH value is lowered to about 11. It is also known that the polymer obtained can sustain temperatures higher than the usual oil field temperature. In a porous medium, it is expected that free oxygen bonds will interact with the solid surface and therefore change the wettability due to the hydrophobicity of the polymer obtained. However, we did not study the complex adsorption mechanism occurring between the polymer and different rock surfaces. 3. Experimental method for treatment We used two methods to change the wettability of outcrop sandstones. The first one was performed in the absence of oil in the plugs and is therefore a laboratory application, while the second one is performed during a water-flooding and more representative of field applications. These procedures were tested on outcrop sandstone samples. 3.1. Porous medium and fluids The plugs used for centrifuge experiments were drilled from the same block of Vosges sandstone. They were 40 mm in diameter and 50 mm in length. They were cleaned using isopropylic alcohol. For the four samples used in this study, porosity values were between 22.4% and 22.9% and permeability between Fig. 2. Reaction mechanism for the polymerization of the chemical additive used in this study. It is expected that free oxygen bonds will interact with the porous medium surface. 126 M. Fleury et al.r Journal of Petroleum Science and Engineering 24 (1999) 123–130 71 and 104 mD. The plugs were also characterized in terms of capillary pressure before treatment. Drainage and forced imbibition centrifuge capillary pressure curves for the two samples are shown in Fig. 3. The fluids used in all experiments were brine ŽNaCl 20grl. and dodecane Ždensity 0.75 and viscosity 1.3 cp.. Both the irreducible water saturation and residual oil saturation for the different samples are very close Ž0.27–0.29 and 0.39, respectively., indicating that the sandstone was homogeneous. The small differences in the shapes of the capillary pressure curves ŽFig. 3. are not significant. Similar samples were used for the micropore membrane experiments but they were not companion plugs of the centrifuge plugs. For these experiments, the length was reduced to 28 mm with a diameter of 40 mm. Experimental details concerning the micopore membrane technique can be found in Longeron et al. Ž1995.. with a solution of brine and siliconate at a concentration of 5%. At this point, a reaction is not obtained because the pH is still high. Then, carbon dioxide is simply injected at a low flow rate. When the saturation is stable and no production of solution is observed, the sample is dried at 608C for further polymerization in contact with air. After complete drying, the sample is flushed again with brine to remove salt deposit formed during the reaction. Note that the concentration has little importance here if it is larger than 1%. 3.2. Procedure without oil in place – In the absence of oil, the sample was first saturated with brine using standard vacuum technique Žnote that brine is used only to stabilize clay and should have no impact on the polymerization mechanism.. Then, the sample is flushed in a Hassler cell – 3.3. Procedure with oil in place For potential field applications, full drying of the porous medium is not appropriate and we tested a procedure relevant to water flooding. The sequence of manipulation is as follows: – – – – – Fig. 3. Centrifuge capillary pressure curves before treatment for two companion sandstone samples used in the study. sample centrifuged under oil to irreducible water saturation sample mounted in Hassler cell and water flooded with a brine-siliconate solution for 23 h at a flow rate of 3 ccrh the sample was left for 72 h to allow diffusion of the siliconate into brine that have not been flushed during previous operation carbon dioxide is injected for 18 h at a flow rate of about 10 lrh and pressure of about 200 mBar to neutralize the solution. Note that oil and water are produced at this step. the sample was left again for 72 h to allow CO 2 diffusion brine was injected to stop polymerization sample was flushed with isopropyl alcohol and dried at 608C and fully resaturated with brine. The stability vs. time of the treatment has not been tested extensively. However, both procedures with and without oil in place included cleaning with isopropyl alcohol to remove the polymer not adsorbed on the surface. Because the product was designed to have a long term stability on various materials, there is no evidence that desorption could occur within a short time Že.g., a month.. M. Fleury et al.r Journal of Petroleum Science and Engineering 24 (1999) 123–130 127 4. Results We tested the change of wettability in three ways: drainage and forced imbibition capillary pressure curves by centrifuge, drainage and imbibition capillary pressure curves by the micropore membrane method, and end-point water-relative permeability after centrifugation. 4.1. Centrifuge capillary pressure curÕes–Sor Õalues When the treatment is performed without oil in place as described above, the forced-imbibition centrifuge capillary pressure curves are severely modified ŽFig. 4.. The oil residual saturation is decreased from 0.33 down to 0.1, which is a typical effect of intermediate wettability. Moreover, spontaneous imbibition of water is not observed at zero imposed pressure when the sample is immersed in brine. For negative capillary pressure, there is an entry pressure of y70 " 10 Ža value between y60 and y80 mBar corresponding to the first two speed steps of the centrifuge. When the treatment is performed with oil in place Žsee Section 3.3., the wettability effect is reduced. The residual oil saturation is decreased from 0.39 to Fig. 5. Effect of Siliconate treatment on centrifuge capillary pressure curves. The treatment is performed in the presence of oil. The saturation after forced imbibition varies from 0.61 to 0.72. 0.28 ŽFig. 5.. There is no entry pressure at negative capillary pressure and the existence of a spontaneous imbibition process during immersion in brine is uncertain. The drainage curves are not modified. We conclude that the polymerization process and adsorption process on the surface is less efficient in the presence of oil but still significant. With or without oil, the irreducible water saturation Swirr at the end of the primary drainage is unchanged ŽFig. 4 and Fig. 5. at the same level of capillary pressure, and there is no spontaneous drainage. There is a difference in the curvature around S s 0.4 ŽFig. 4., but this is due to an inadequate equilibrium time which was modified later in the other experiments. The drainage curve is modified only slightly at the plateau in the case of the treatment with oil; the entry pressure at Sw s 1 changed and decreased from 112 " 7 mBar down to 66 " 6 mBar. 4.2. Micropore membrane capillary pressure Fig. 4. Effect of Siliconate treatment on centrifuge capillary pressure curves. The treatment is performed in the absence of oil. There is a small negative entry pressure and a large variation of Sor is observed for the treated sample. We used the micropore membrane technique to measure positive capillary pressure curves Ždrainage and spontaneous imbibition.. Samples were treated without oil in place. For primary drainage, the results obtained with the centrifuge technique ŽFig. 4. are confirmed: there is a variation of the entry pressure 128 M. Fleury et al.r Journal of Petroleum Science and Engineering 24 (1999) 123–130 pressure drop measured ŽFig. 7.. The measurements indicate that there is no rate effect in the range considered Ža linear relation ship between flow rate and pressure drop is found.. For the sample shown in Fig. 5 Žtreatment with oil in place., we found out that: KrwŽSwor s 0.60. s 0.05 Žwithout treatment. KrwŽSwor s 0.74. s 0.04 Žwith treatment, procedure with oil in place. Fig. 6. Effect of Siliconate treatment on drainage and imbibition capillary pressure curves. The treatment is performed in the absence of oil. The hysteresis between drainage and imbibition is amplified. and the capillary pressure of the plateau ŽFig. 6. which are determined more precisely with this technique. The Swirr value is also slightly smaller Žfrom 0.29 down to 0.26. when the sample is treated. The spontaneous imbibition curves are very different when the sample is treated ŽFig. 6.. The saturation change is almost zero for the treated sample, in contrast with the untreated sample where the drainage-imbibition hysteresis is smaller. The nearly vertical spontaneous imbibition curve is also consistent with the existence of an entry pressure at negative capillary pressure observed with the centrifuge technique ŽFig. 4.. This behavior is expected for intermediate wettability. For negative capillary pressure, the large variation of Sor value obtained with the centrifuge ŽFig. 4. is also confirmed Žhowever, negative capillary pressure measurements below y200 mBar are not available due to leaks.. There is a plateau region in the forced imbibition curve for the treated sample that is not observed for the water-wet untreated sample. where the relative permeability is normalized using the absolute permeability. Therefore, there is no significant change in end point water relative permeability. Because the oil saturation is lower when the sample is treated, the water relative permeability should be higher and we conclude that there is a partial plugging caused by the polymer. Based on literature, the lack of change of end-point Krw was also not expected: general trends Že.g., Anderson, 1987b, part 6, Gauchet, 1993. indicate that the endpoint KrwŽSwor. is often increased for intermediate wetting conditions. Note that the centrifuge yields a non uniform saturation profile in the sample and the measurement is performed actually at an average water saturation slightly smaller than Swor. However, this difference is only relevant for the oil relative permeability. 4.3. End point water relatiÕe permeability We used a standard procedure to measure KrŽSwor. at the end of the centrifuge forced imbibition displacement. The sample is removed from the centrifuge core holder and placed in a Hassler cell; then, brine is injected at different flow rates and the Fig. 7. Relative permeability measurement at the end of the centrifuge forced imbibition for a treated sample at different flow rates Žsquares.. The value KrŽSwor. obtained from the slope Žfull line. is not changed by the treatment. M. Fleury et al.r Journal of Petroleum Science and Engineering 24 (1999) 123–130 5. Conclusions A chemical treatment to obtain an intermediate wettability was presented. The procedure is simple, and the chemical used is not hazardous. The chemical reaction is basically a polymerization process occurring when the pH of the solution is lowered to near neutral. A useful property of the product used is that it is soluble in water. The change of wettability is demonstrated through the measurements of capillary pressure curves, especially imbibition curves, using the centrifuge technique and the micropore membrane technique for positive Pc values. The results from both techniques are coherent and indicate that the change in the capillary pressure curves is similar to what was obtained when samples were aged in crude oil: – – – drainage curves were not strongly modified; the drainage-imbibition hysteresis was severely amplified; and the Sor values were much smaller for the treated samples. However, the end point relative permeability KrwŽSwor. is only slightly modified by the treatment, even though saturation differ significantly. 6. Nomenclature Krw Pc Sw Swirr Swor Sor water relative permeability, function of water saturation capillary pressure defined as the oilwater pressure difference water saturation irreducible water saturation residual oil saturation expressed in term of water saturations 1-Sor residual oil saturation References Anderson, W.G., 1987a. Wettability literature survey: Part 4. Effects of wettability on capillary pressure. J. Pet. Technol., 1605–1623, October. Anderson, W.G., 1987b. Wettability literature survey: Part 6. The 129 effects of wettability on waterflooding. J. Pet. Technol., 1605– 1623, December. Bosh, E., Brunsperger, K., Gluck, H., Pirson, E., Roth, M., 1976. Method for Imparting Water-Repellency to Construction Materials, US Patent 3,956,570. Coley, F.H., Marsden, S.S., Calhoun, J.C., 1956. A study of the effect of wettabiity on the behavior of fluids in synthetic porous media. Prod. Mon., 29–45, June. Elliot, J.R., Krieble, R.H., 1950. Process for Rendering Materials Water-Repellent and Compositions Therefor, US Patent, 2,507,200. Fatt, I., Klikoff, W.A., 1959. Effect of fractional wettability on multiphase flow through porous media. Petr. Trans. AIME 216, 426–432. Gauchet, R., 1993. Waterfloodings on intermediate wet porous media: new considerations on two phase flow properties determination. In: Worthington, P.F., Chardaire Riviere, ` C. ŽEds.., Advances in Core Evaluation III. London Gordon and Breach, pp. 251–273. Jadhunandan, P.P., Morrow, N.R., 1995. Effect of wettability on waterflood recovery for crude-oilrbrinerrock systems. SPERE, 40–46, Feb. Leach, R.O., Wagner, O.R., Wood, H.W., Harpke, C.F., 1962. A laboratory and field study of wettability adjustment in water flooding. J. Pet. Technol., 206–212, Feb. Lombard, J.M., Lenormand, R., 1993. Fractional wettability and petrophysical parameters of porous media. In: Worthington, P.F., Chardaire-Riviere, ` C. ŽEds.., Advances in core evaluation: III. Reservoir management. Gordon and Breach Science Publishers, pp. 411–438. Lombard, J.M., Robin, M., Lenormand, R., 1991. Fluid flow in intermediate wettability porous media. Capillary pressure: experiments and simulations. In: 6th Research Conference on Exploration-Production. Physical Chemistry of Colloıds ¨ and Interfaces in Oil Production, Saint-Raphael, ¨ 4–6 Sept 1991. Longeron, D., Hammervold, W.L., Skjaeveland, S.M., Water-Oil Capillary Pressure and Wettability Measurements Using Micropore Membrane Technique, paper SPE 30006 presented at the International Meeting on Petroleum Engineering in Beijing, PR China, 14–17 November 1995. Maini, B.B., Ionescu, E., Batycky, J.P., 1986. Miscible displacement of residual oil — effects of wettability on dispersion in porous media. JCPT, 36–41, May–June. Michaels, A.S., Porter, M.C., 1965. Water-oil displacements from porous media utilizing transient adhesion-tension alterations. AIChE Journal 11 Ž4., 617–624. Michaels, A.S., Stancell, A., 1964. Effect of chromatographic transport in hexylamine on displement of oil by water in porous media. Soc. Pet. Eng. J., 231–239. Morrow, N.R., Cram, P.J., McCaffery, F.G., 1973. Displacement studies in dolomite with wettability control by octanoic acid. Soc. Pet. Eng. J., 221–232, Aug. Sarem, A.M.S., Hutchins, R.D., Friedman, S.R., 1993. Enhanced Imbibition Oil Recovery Process, US Patent 5,247,993. Skauge, A., Aarra, M., 1993. Effect of Wettability on the Oil Recovery by WAG, presented at the 7th European IOR-Symposium, Moscow. 130 M. Fleury et al.r Journal of Petroleum Science and Engineering 24 (1999) 123–130 Takach N.E., Bennett, L.B., Douglas, C.B., 1989. Generation of Oil-wet Model Sandstone Surfaces, SPE paper 18465, presented at the 1989 International Symposium on Oilfield Chemistry, Houston. Zhou, X., Torsaeter, O., Xie, X, Morrow, N.R., 1993. The Effect of Crude-Oil Aging Time and Temperature on the Rate of water Imbibition and Long-Term Recovery by Imbibition, paper 26674, presented at the 68th Annual Technical Conference and Exhibition of the Society of petroleum Engineers, Houston, USA.