Petroleum Science and Technology, 29:24322440, 2011 Copyright Taylor & Francis Group, LLC ISSN: 1091-6466 print/1532-2459

9 online DOI: 10.1080/10916461003735061

The Effect of Solvent Nature and Dispersant Performance on Asphaltene Precipitation from Diluted Solutions of Instable Crude Oil
J. C. PEREIRA,1;2 J. DELGADO-LINARES,2 A. BRIONES,2 M. GUEVARA,2 C. SCORZZA,2 AND J.-L. SALAGER2
1

Downloaded by [J. C. Pereira] at 12:56 18 October 2011

Lab. Petrleo, Hidrocarburos y Derivados (PHD), Universidad de Carabobo, Facultad Experimental de Ciencias y Tecnologa, Departamento de Qumica, Valencia, Venezuela 2 Laboratorio de Formulacin, Interfases, Reologa y Procesos, Facultad de Ingeniera, Escuela de Ingeniera Qumica, Mrida, Venezuela
Abstract The asphaltene precipitation experiments were studied on El Furrial crude oil from western Venezuela, which is known to exhibit serious instability problems. A Turbiscan backscattering apparatus was used to evaluate the precipitation of asphaltenes with different solvents. The transmittance variation with time was studied as the crude was diluted with heptane, pentane, and cyclohexane. Linear alkanescontaining systems exhibit a two-stage behavior, whereas only one is found when diluting with cyclohexane. Dispersing agents were tested by using the precipitate height as a criterion of effectiveness. Results are reported for ethoxylated nonylphenols and a commercial dispersant. Keywords asphaltene, crude oil, dispersant, precipitation

Introduction
During crude oil production and rening, unwanted asphaltene precipitation can take place in pipelines, separation equipment, and conversion units. Asphaltenes are generally dened as the petroleum fraction soluble in aromatic solvents (i.e., benzene and toluene) but insoluble in n-alkanes (i.e., n-pentane and n-heptane; Acevedo et al., 1985; Speight 2004; Labrador et al., 2007). The precipitation of asphaltenes has been shown to be intimately related to their behavior in solution, which has been modeled according to liquidliquid or solidliquid equilibria (Qin et al., 2000; Angle et al., 2005). Most publications on asphaltenes considered some kind of aggregation and aggregate peptization by resins (Pfeiffer and Saal, 1940; Priyanto et al., 2001; Pereira et al., 2007). This aggregation often boosts the asphaltene lms adsorbed at the interface that stabilize water-in-crude emulsions (McLean and Kilpatrick, 1997). Ekulu and coworkers (2005) have studied the microstructure of asphaltenes in crude oil with differential scanning calorimetry. Their results suggest that the thermal effects produced when the crude is mixed with n-heptane or toluene are characteristics of the
Address correspondence to Juan Carlos Pereira, Lab. PHD, FACYT, Universidad de Carabobo, Avenue Salvador Allende, 2001 Carabobo, Venezuela. E-mail: jcpereir@uc.edu.ve

2432

Effect of Solvent on Asphaltene Precipitation

2433

Downloaded by [J. C. Pereira] at 12:56 18 October 2011

solvent. This conforms with the well-known inuence of the medium on the aggregation of asphaltenes (Yudin et al., 1998; Castillo et al., 2001; Ortega-Rodrguez et al., 2003). Synthetic dispersing agents have been used as stabilizers of asphaltenes in crude oil, which is essentially the role played by some species of resins to stop the aggregation at some aggregate size and avoid precipitation. Carnahan et al. (1999) found that Boscan crude resins are better stabilizers than those from Hamaca crude. An increase in stabilization results from the addition of Boscan resins to Hamaca crude. A correlation was suggested between the resin polarity and its stabilizing capacity. It has been shown that the effectiveness of asphaltenes dispersants depend on their molecular structures (Wiehe and Jermansen, 2003). Deasphalted oil (DO) and resins have been used by Al-Sahhaf et al. (2002) as inhibitors of the asphaltene precipitation. Various surfactant molecules were tested for the same application, and a correlation was found between their effectiveness as asphaltene dispersants and the acidity of their polar group (Rogel and Len, 2001; Rocha et al., 2006). Recent studies on the aggregation and precipitation of asphaltenes (Marugn et al., 2009) were based on the variation of light reectance and transmittance versus time. Kraiwattanawong et al. (2009) related the effectiveness of several asphaltene dispersants to the average transmittance of the blend (crude oil, dispersants, n-heptane), which is essentially a direct measurement of the sample turbidity. In the present article, the variation of transmittance versus time will be used to study the effect of solvent type. Additionally, the inuence of synthetic dispersants on asphaltene precipitation was monitored for the height of the precipitate.

Materials and Methods

Crude Oil Asphaltenes and maltenes were extracted from El Furrial petroleum, a crude oil from western Venezuela, which is known for its strong tendency to precipitate asphaltenes. Some properties of this crude oil are presented in Table 1 (Pereira, 2000). Asphaltenes Extraction A 10 g sample of crude oil was diluted with 400 mL of n-heptane. The sample was stirred for 1 hr and then left to rest for 24 hr at ambient temperature. The precipitate was ltered and washed with hot n-heptane in a Soxhlet extractor until the solvent became colorless. The remaining n-heptane was eliminated by evaporation

Table 1 Properties of El Furrial crude oil and its asphaltene (Pereira, 2000) Asphaltenes in the crude, wt% 7.6 Elemental analysis, wt%a C 85.5 H 6.9 Ob 2.5 N 1.73 S 3.4 H/C 0.97

API 21
a In b Is

asphaltene. calculated by difference with the other elements (C, H, N, and S).

2434

J. C. Pereira et al.

under vacuum (18 mmHg) at 60 C until a constant mass was reached. Asphaltenes were stored in dark glass asks to prevent oxidation. Synthetic Dispersants Lauric alcohol was donated by Lipesa SA (El Tigre, Anzoategui, Venezuela), and polyethoxylated nonylphenol nonionic surfactants with different degrees of ethoxylation (labeled nonyl-phenol [NPEON] where EON is the average number of ethylene oxide group per molecule) were obtained from Witco Chemicals (Houston, TX). A commercial asphaltene dispersant of the ethoxylated nonylphenol-formol resin was supplied by Resinas Multiples SA (Cagua Aragua, Venezuela). The dispersants were rst dissolved in toluene before being added to the crude according to the procedures discussed next. Sample Preparation

Downloaded by [J. C. Pereira] at 12:56 18 October 2011

Solutions containing toluene and crude oil (50:50 by volume) were prepared to achieve a low-viscosity uid. Thirty milliliters of n-heptane was mixed with 0.16 mL of the toluene/crude oil solution and 0.16 mL of dispersant solution to be tested. The dispersant was added so that its concentration in the nal mixture (containing crude C toluene C heptane C dispersant) was 1 and 25%. The latter value was only for dodecanol, which is effective only in very high proportions. A part of the mixture (10 mL) was then poured into a glass tube (Turbiscan cell) and stirred for 5 min in an ultrasonic bath. Precipitation experiments were carried out as indicated next. Experiments of Asphaltene Precipitation Asphaltene precipitation was determined by measuring the variation of transmittance versus the sample height; see Figure 1. These measurements were carried out as a function of time, with a Turbiscan MA2000 apparatus from Formulaction (LUnion, France). The curves of percentage transmittance (%T) versus height of test tube transmittance of asphaltene solution. Each curve corresponds to a particular time (see column on righthand side of gure). The values of transmittance increased as a function of time due to a reduction in the amount of asphaltene in solution. This value of %T was constant throughout the tube test. Furthermore, it is shown in the circle the started curve %T for each system, which is indicative of an increase in the height in the asphaltene precipitate for each system. Initially, the transmittance was essentially zero, as a consequence of the high opacity of the sample. When asphaltenes begin to precipitate, the liquid becomes clearer and the transmittance increases in the tube, whereas the precipitate settles at the bottom. This is the principle of the technique proposed in recent publications to evaluate the performance of asphaltene dispersants (Kraiwattanawong et al., 2009). The transmittance of the sample is measured at some selected time intervals by scanning a nephelometric device from the top to the bottom of the tube. The transmittance of the solution is essentially the same everywhere in the upper part of the tube. The asphaltene precipitate that settles at the bottom of the tube blocks the light beam and keeps the transmittance at zero; this allows the determination of the precipitate height, which is used here to estimate the effectiveness of the tested dispersing agent. The experiments were carried out at ambient temperature (22 2 C).

Effect of Solvent on Asphaltene Precipitation

2435

Downloaded by [J. C. Pereira] at 12:56 18 October 2011

Figure 1. Curves of percentage transmittance (%T) versus height (mm) of the test tube. Each curve corresponds to a particular time (see column on right-hand side). The %T value at solution bulk is constant throughout the tube test. (color gure available online)

Results and Discussion

Effect of the Medium Figure 2 shows the variation of transmittance (%T) of the solution as a function of time (hr) for systems containing El Furrial crude oil and different diluents: heptane, pentane, and cyclohexane. Although toluene was added to the crude at rst (50:50 volume) in order to reduce its original viscosity without precipitating asphaltenes, the amount of toluene in the nal system was negligible; that is, less than 0.3%. There was a considerable excess of diluent (375:1 vol) with respect to the crude (and toluene) to warrant the precipitation of asphaltenes and to result in a low optical density of the liquid. Such an amount of diluent was found to be appropriate to monitor the asphaltene precipitation process. The zero time in Figure 2 actually corresponds to the rst scan, which was carried out a few

2436

J. C. Pereira et al.

Downloaded by [J. C. Pereira] at 12:56 18 October 2011

Figure 2. Asphaltene precipitation from El Furrial crude oil by different solvents through the measurement of the solution transmittance versus time. Each point of %T corresponds to the value of transmittance in solution.

minutes after the diluent was added, stirred in an ultrasonic bath for 5 min, and poured into the tube. The non-zero (but close to zero) transmittance at zero time indicates that the dilution provided a suitable initial turbidity for the purpose of the technique. Each one of the curves plotted in Figure 2 indicates an increase in transmittance of the upper solution as time elapses. The heptane and pentane curves were found to rise slowly over the rst 300 hr and to exhibit some irregularities that are not very signicant but are denitely larger than the experimental error. This could indicate the presence of different stages in the precipitation process. Cyclohexane exhibited a much quicker and smoother variation with time. A plateau value was attained after 2025 hr and corresponded to about 60% transmittance; that is, much less than the value attained with heptane after 300 hr. The initial increase in transmittance is shown in Figure 3, which is a blowup of Figure 2 data during the rst hours. The initial slope may be considered as an indirect measurement of the initial precipitations, which is seen to be much faster for cyclohexane than for both n-alkanes; that is, 10 times or more in the 1- to 4-hr intervals. Because the agglomeration-precipitation process is likely to start with an association into small size agglomerates sometimes referred to as aggregation, it may be conjectured that cyclohexane favors such a rst step, which results in the reduction of the scattering centers from molecules to core aggregates and thus a reduction of the turbidity. Cyclohexane is a better asphaltenes solvent than n-alkanes; hence, it might keep a part of the asphaltenes in a molecular state while favoring the association of only some other part, probably the higher molecular weight species. The lower molecular weight asphaltenes would hence remain as single molecules or in a very low aggregation (dimer or trimer) state and be responsible for the relatively high turbidity that remains after 2025 hr, when some equilibrium appears to be reached. With n-alkanes, the initial precipitation is much slower, thus indicating that the formation of the rst aggregates is delayed. This may be explained by the fact that the resins

Effect of Solvent on Asphaltene Precipitation

2437

Downloaded by [J. C. Pereira] at 12:56 18 October 2011

Figure 3. Initial kinetics for three solvents used. Data correspond to initial values shown in Figure 2.

are less soluble in n-alkanes than in cyclohexane and hence are more likely to interact with asphaltenes to keep them as smaller aggregates. The fact that the precipitation is even slower for pentane than for heptane is consistent with this interpretation, because resins are less soluble in pentane. On the other hand, Figure 2 indicates that the transmittance continued to increase over 200 hr; that is, that the formation of aggregates, probably of larger and larger size, proceeded with the systems containing n-alkanes. This is consistent with the fact that less asphaltene will be soluble at the end in these solvents, particularly n-heptane, which is the most aliphatic (McLean and Kilpatrick, 1997; Spiecker et al., 2003). Effect of Synthetic Dispersants The dispersant activity of some surfactants (NPEON, 1-dodecanol, and commercial dispersant) was evaluated with El Furrial crude oil. The dilution and asphaltene precipitation were carried out with n-heptane, which is the most effective solvent for precipitation of asphaltenes. The excess heptane added to the crudetoluene mixture led to asphaltene precipitation and allowed evaluation of the effectiveness of the dispersants. The Turbiscan was used to measure the precipitate height in the test tube as time elapsed, and the precipitation efciency criterion was taken as this height after 200 hr, which, according to Figure 2, is probably close to the plateau equilibrium value. The % dispersion index is dened by the following equation: .HnoD HD/ %Dispersion index D 100 (1) HnoD where HnoD is the precipitate height without addition of any dispersant and HD is the corresponding height with dispersant D at the indicated concentration. This criterion provides an estimate of the asphaltenes that would have been precipitated by the n-heptane alone but were kept suspended in the solution by the dispersant.

2438

J. C. Pereira et al. Table 2 Height of asphaltene precipitate for several dispersants after 200 hr Precipitate height, mm 2.57 2.87 2.85 2.52 2.15 2.86 1.96 0.68 % Dispersion index 2.0 16.3 23.7 73.5

Dispersant No dispersant NP4 NP5 NP7 NP11 Dodecanol Dodecanol Commercial

wt% 1 1 1 1 1 25 1

Downloaded by [J. C. Pereira] at 12:56 18 October 2011

Figure 4. Height of asphaltene precipitate after 200 hr versus the type of ethoxylated nonylphenol dispersant. The solvent employed as precipitating agent was heptane.

The data in Table 2 suggest that this method is appropriate to evaluate the performance of a dispersing agent. In the present case, the commercial dispersant is the best agent, with a 73% dispersion index. For the ethoxylated nonylphenols (see Figure 4), it is clear that the performance increased with the degree of ethoxylation and that it is signicant only for more polar species (EON > 7). In the case of dodecanol, it is worth noting that for obtained a similar dispersion index is attained only at extremely high concentrations. This result is consistent with the fact that it is much less polar than the NP4 species.

Conclusions
The Turbiscan is a useful and simple tool for the study of asphaltene precipitation when El Furrial crude oil is diluted with n-alkanes. The variation of transmittance versus time brings a clear information on the asphaltene precipitation in excess of several

Effect of Solvent on Asphaltene Precipitation

2439

solvents. With n-alkanes, the initial precipitation is much slower, thus indicating that the formation of the rst aggregates is delayed. This may be explained by the fact that the resins are less soluble in n-alkanes than in cyclohexane and hence are more likely to interact with asphaltenes to keep them as smaller aggregates. Measurement of the precipitate height provides an estimate of the performance of an asphaltene dispersing agent. The commercial dispersant was the best agent, with a 73% dispersion index. For the ethoxylated nonylphenols, the performance increased with the degree of ethoxylation, which is signicant only for more polar species (EON > 7).

Acknowledgment
The authors are grateful to the Council of Scientic and Humanistic Development of the University of Los Andes for nancial support under Project I-1039-07-08-F. P.J.C. thanks the University of Carabobo for a doctoral fellowship.

Downloaded by [J. C. Pereira] at 12:56 18 October 2011

References
Acevedo, S., Escobar, G., Mndez, B., Rojas, A., Layrisee, I., and Rivas, H. (1985). Asphaltenes and resins from the Orinoco basin. Fuel 64:17411747. Al-Sahhaf, T. A., Fahim, M. A., and Amal, S. E. (2002). Retardation of asphaltene precipitation by addition of toluene, resins, deasphalted oil and surfactants. Fluid Phase Equil. 194197:1045 1057. Angle, C. W., Longa, Y., Hamza, H., and Lue, L. (2006). Precipitation of asphaltenes from solventdiluted heavy oil and thermodynamic properties of solvent-diluted heavy oil solutions. Fuel 85:492506. Carnahan, N., Salager, J., L., Antn, R. E., and Davila, A. (1999). Properties of resins extracted from Boscan crude oil and their effect on the stability of asphaltenes in Boscan and Hamaca crude oils. Energ. Fuel. 13:309314. Castillo, J., Hung, J., Fernndez, A., and Mujica, V. (2001). Nonlinear optical evidences of aggregation in asphaltene-toluene solutions. Fuel 80:12391243. Ekulu, G., Nicolas, C., Achard, C., and Rogalski, M. (2005). Characterization of aggregation processes in crude oils using differential scanning calorimetry. Energ. Fuel. 19:12971302. Kraiwattanawong, K., Fogler, H. S., Gharfeh, S., Singh, P., Thomason, W., and Chavadej, S. (2009). Effect of asphaltene dispersants on aggregate size distribution and growth. Energ. Fuel. 23:15751582. Labrador, H., Fernandez, Y., Tovar, J., Muoz, R., and Pereira, J. C. (2007). Ellipsometry study of the adsorption of asphaltenes lms on glass surface. Energ. Fuel. 21:12261230. Marugn, J., Calles, J., Dufour, J., Gimnez-Aguirre, R., Pea, J., and Merino-Garcia, J. (2009). Characterization of the asphaltene onset region by focused-beam laser reectance: a tool for additives screening. Energ. Fuel. 23:11551161. McLean, J., and Kilpatrick, P. (1997). Effects of asphaltene solvency on stability of water-in-crudeoil emulsions. J. Colloid Interface Sci. 189:242253. Ortega-Rodrguez, A., Cruz, S. A., Gil-Villegas, A., Guevara-Rodrguez, F., and Lira-Galeana, C. (2003). Molecular view of the asphaltene aggregation behavior in asphaltene-resin mixtures. Energ. Fuel. 17:11001108. Pereira, J. C. (2000). Structural characterization of asphaltene from Venezuelan crudes oil. M.Sc. Thesis, University Central of Venezuela, Caracas. (In Spanish) Pereira, J. C., Lpez, I., Salas, R., Silva, F., Fernndez, C., Urbina, C., and Lpez, J. C. (2007). Resins: The molecules responsible for the stability/instability phenomena of asphaltenes. Energ. Fuel. 21:13171321. Pfeiffer, J., and Saal, R. (1940). Asphaltic bitumen as colloid system. J. Phys. Chem. 44:139149.

2440

J. C. Pereira et al.

Downloaded by [J. C. Pereira] at 12:56 18 October 2011

Priyanto, G., Mansoori, G., and Suwono, A. (2001). Measurement of property relationships of nano-structure micelles and coacervates of asphaltene in a pure solvent. Chem. Eng. Sci. 17:69336939. Qin, X., Wan, P., Sepehrnoori, K., and Pope, G. (2000). Modeling asphaltene precipitation in reservoir simulation. Ind. Eng. Chem. Res. 39:26442654. Rocha, L. C., Silva, M. F., and da Silva, A. C. (2006). Inhibition of asphaltene precipitation in Brazilian crude oils using new oil soluble amphiphiles. J. Petrol. Sci. Eng. 51:2636. Rogel, E., and Len, O. (2001). Study of the adsorption of alkyl-benzene-derived amphiphiles on an asphaltene surface using molecular dynamics simulations. Energ. Fuel. 15:10771086. Speight, J. (2004). Petroleum asphaltene constituentsPart 1Asphaltene constituents, resins and the structure of petroleum. Oil Gas Sci. Tech. 59:467477. Spiecker, P. M., Gawrys, K. L., and Kilpatrick, P. K. (2003). Aggregation and solubility behavior of asphaltenes and their subfractions. J. Colloid Interface Sci. 267:178193. Wiehe, I. A., and Jermansen, T. G. (2003). Design of synthetic dispersants for asphaltenes. Petrol. Sci. Tech. 21:527536. Yudin, I. K., Nikolaenko, G. L., Gorodetskii, E. E., Kosov, V. I., Melikyan, V. R., Markhashov, E. L., Frot, D., and Briolant, Y. (1998). Mechanisms of asphaltene aggregation in toluene-heptane mixture. J. Petrol. Sci. Eng. 20:297301.