Article pubs.acs.org/EF
Characterization of Carbon-Dioxide-Induced Asphaltene Precipitation Milind Deo*,† and Martha Parra†,‡ †
Department of Chemical Engineering, University of Utah, Salt Lake City, Utah 84112, United States ABSTRACT: Asphaltene precipitation continues to be one of the major problems during the carbon dioxide enhanced oil recovery process. Supercritical carbon dioxide is injected into oil reservoirs to enhance oil recovery in the tertiary phase of production. Development of multiple-contact miscibility is important in the success of a carbon dioxide flood. In this paper, we establish conditions at which solids are formed in carbon dioxide−oil systems and characterize these solids by a variety of analytical techniques. Crude oils containing low and medium amounts of heptane-insoluble asphaltenes were used in the study. A high-pressure thermodynamic system was designed and fabricated, and a number of thermodynamic experiments were performed with dead and live oils. Multiple-contact experiments were first performed by adding a “light−intermediate” cut to the crude oil and later by creating actual multiple contacts. Precipitation onset for dead oils was obtained between 20 and 30 mol % CO2. CO2-induced solids consisted of about 30 wt % pentane insolubles (asphaltenes) and 7 wt % resins. Concentrations of pentane insolubles (asphaltenes) in the liquid fractions obtained after thermodynamic experiments decreased with an increase in the CO2 concentration used in the experiment. Thermodynamic experiments with live oils showed that the precipitate amounts were 3−4 times higher compared to equivalent dead oil solid deposits. For concentrations higher than or equal to 28 mol % CO2, the precipitates had similar characteristics, with about 34 wt % pentane insolubles and 7 wt % resins. Live-oil experiments also showed that concentrations of pentane insolubles (asphaltenes) in the liquid fractions obtained after thermodynamic experiments decreased with an increase in CO2 concentrations. These values were even lower than the values for dead oils. Solids analyses revealed that the CO2-induced precipitates contained shorter alkylated chains than asphaltenes defined as pentane or heptane insolubles. The functional groups of compounds in the solids obtained using different paraffinic solvents and CO2-induced precipitation were similar. The field deposits in general were heavier in comparison to all of the laboratory-generated samples. Scanning electron microscopy (SEM) images quantified using energy-dispersive spectroscopy (EDS) showed a gradual increase in sulfur moieties as the molecular weight of the solvent increased with CO2-induced precipitates falling between C6 and C7. The field deposits contained the highest amount of sulfur. Calculations from a homogeneous molecular thermodynamic model provided good agreement with the experimental results.
■
INTRODUCTION The polyaromatic fraction of crude oil is often characterized by its solubility in normal alkanes, such as pentane, hexane, and heptane. Asphaltenes are the polar, polyaromatic, and highmolecular-weight hydrocarbon fractions that precipitate upon mixing the stabilized crude oil with excess pentane or heptane.1,2 Precipitation of solids believed to be asphaltenes has been reported at various stages of oil production. Precipitation is almost universal in gas flooding applications but is also observed in the primary production of undersaturated crude oils. Asphaltene precipitation during tertiary recovery processes has been reported in the north Aegean Sea, Venezuela, Algeria, California, and Mississippi.3,4 Asphaltene precipitation has been cited as one of the most significant problems in the production of oil by CO2 flooding from the various operations in Colorado and the Permian Basin. Precipitation is typically observed in surface facilities but is seldom reported to occur in the formation.5,6 Deposited asphaltenes can reduce effective oil mobility by blocking pore throats and adsorbing onto the rock, thus changing the formation wettability. In some instances, asphaltenes are known to increase hydrocarbon viscosity by inducing the formation of water in oil emulsions.7 © 2011 American Chemical Society
Asphaltene content in the crude as determined by conventional methods is not a good indicator of asphaltene precipitation problems in gas flooding. It has been reported that a crude oil with more than 17 wt % asphaltenes did not have asphaltene problems, while an oil with only 0.1 wt % had serious asphaltene deposition problems.4 Asphaltenes precipitate because of changes of the pressure, temperature, and composition and the thermodynamic instability that these changes entail. In terms of composition, the stabilizing power of the oil depends upon relative amounts of paraffins, aromatics, and resins. It is important to determine the asphaltene precipitation envelope within which asphaltenes flocculate and precipitate. To this effect, it is necessary to determine the onset of precipitation followed by the measurement of asphaltene precipitation at different conditions. Burke et al.8 obtained experimental asphaltene precipitation data from static precipitation tests, in which several oil and solvent mixtures were evaluated in a high-pressure cell at Special Issue: 12th International Conference on Petroleum Phase Behavior and Fouling Received: September 15, 2011 Revised: December 16, 2011 Published: December 20, 2011 2672
dx.doi.org/10.1021/ef201402v | Energy Fuels 2012, 26, 2672−2679
Energy & Fuels
Article
reservoir temperature and pressure. They showed that precipitation would depend upon the composition of the crude oil, the added solvent, and the concentration of asphaltenes in the oil. Results show that the temperature, pressure, and phase behavior of the system also influence asphaltene precipitation. In static pressure−volume−temperature (PVT) tests, asphaltene flocculation usually occurs, whereas in corefloods, asphaltene precipitation/deposition in the core matrix may not occur.9,10 Graue and Zana11 reported asphaltene precipitation in the extensive laboratory evaluation undertaken to evaluate the performance of a possible CO2 tertiary project at the Rangely Weber Sand Unit in Colorado. Precipitation occurred at a gas concentration of 44 mol % or higher with an injection gas (95 mol % CO2 and 5 mol % CH4), while with a second gas (85 mol % CO2, 5 mol % CH4, and 10 mol % N2), it occurred at a gas concentration as low as 25 mol % (lowest gas concentration studied). The solid precipitation was estimated to be 2−5 vol % of the original reservoir oil. Hervey and Iakovakis6 reported that deposits have been found in pumps, tubing, wellheads, flowlines, and workover equipment, with no evidence of asphaltene precipitation in the formation. Hirschberg et al.12 studied the influence of the temperature and pressure on asphaltene precipitation during natural depletion. They found that, when CO2 was injected into the crude oil, no precipitation was observed without the addition of decane, an intermediate hydrocarbon component. They also concluded that precipitation is highest at the bubble point pressure. Monger and co-workers13,14 provided a comprehensive database on precipitation associated with CO2 recovery. They observed precipitation for mixtures containing between 70 and 96 mol % CO2. They observed that the development of miscibility is associated with the formation of the solid phase. Novosad and Constain15 observed precipitation at 40−50 mol % CO2 concentrations. Srivastava10 indicated that asphaltene began to flocculate at about 42 mol % CO2 concentration, and after that, there was a linear increase in asphaltene flocculation with the CO2 concentration. Verdier et al.16 reported onset points for CO2-induced precipitation in a variety of oils and showed that, as the temperature increased, asphaltene stability decreased. Takahashi et al.17 and Okabe et al.18 studied onset points and deposition under flow conditions for CO2 + oil systems. There is some discrepancy in the amount of CO2 required for inducing precipitation. The primary mechanism through which additional oil is recovered in most CO2 floods is multiple-contact miscibility. As multiple-contact miscibility is induced, the oil downstream becomes enriched in intermediates. In this paper, we test the hypothesis that enriching the crude oil in intermediates leads to a higher degree of precipitation than from the original crude oil. A schematic of possible compositional changes from the CO2 injection well to the producer is shown in Figure 1.
■
Figure 1. Schematic of possible compositional changes in the reservoir as CO2 is injected into formations. The multiple-contact miscibility (MCM) front is enriched with oil components extracted by CO2.
Figure 2. Experimental system used to study solids precipitation with first and multiple-contact experiments. a low-pressure input to feed the required fluids to the vessels, and two positive-displacement pumps connected to the vessels to transfer fluids from each of these vessels. The PVT cells had sight-glass windows on either side, which were useful in visualizing the phase changes and fluids. The cells were connected to a circulating pump, allowing for equilibration of the fluids. The outlet from the PVT cells was connected to a set of three parallel, high-pressure, in-line filters; each line had two filters with pore sizes of 2 and 0.5 μm for trapping precipitates for quantitative analyses of the solid phases. The system pressure was maintained with a special back-pressure regulator (BPR). Commercially designed and built BPRs have serious problems maintaining back-pressure when high-pressure carbon dioxide is the displacement fluid. The BPR employed in all of the studies was specially designed and fabricated. The fluid from experiments was displaced into two interconnected cylinders equipped with a pressure indicator. The sample was received in the first cylinder, and the gas fraction was moved to the second cylinder. The separation procedure, along with pressure and compositional measurements, permitted quantitative analyses of the gas and liquid samples collected.
EXPERIMENTAL SECTION
Experiments in this study were carried out in high-pressure, hightemperature equipment illustrated Figure 2. The system was designed for temperatures up to 300 °F and pressures up to 5000 pounds per square inch (psi), permitting the experiments to be run at reservoir conditions. A constant temperature was maintained by placing the equipment in an insulated wooden box. The fluid displacement section was used to supply all of the required fluids: oil, CO2, and water to the PVT cells. This section was composed of three moveable piston vessels (MPVs) to store the fluids, 2673
dx.doi.org/10.1021/ef201402v | Energy Fuels 2012, 26, 2672−2679
Energy & Fuels
Article
Four different types of thermodynamic experiments were performed, each of which had a definite protocol: (1) carbon dioxide first contact with dead oil, (2) carbon dioxide first contact with live oil, (3) simulated multiple-contact miscibility experiments, and (4) true multiple-contact experiments. All of the experiments were conducted at 160 °F. Details of the experimental protocol are provided by ParraRamirez.20
■
RESULTS Crude Oil Composition. Most experiments were performed with oil RO. The C5−C60 carbon number distribution of the original crude oil as measured by simulated distillation [American Society for Testing and Materials (ASTM) D2887], is shown Figure 3. The oil contained 3.75 wt % C60+. The Figure 5. SARA distributions in the original oil.
First-Contact Experiments. First-contact experiments with carbon dioxide were performed by charging the PVT cell with crude oil and CO2 mixtures in different molar proportions according to the method described by ParraRamirez.20 The mixtures were displaced above the bubble point through online, high-pressure filters at a constant pressure (3000 psia). The amounts of precipitates at various CO2 concentrations are shown in Figure 6. On the basis of these
Figure 3. Carbon number distribution of the original crude oil.
gravity of the oil was about 34° API (American Petroleum Institute). Saturates, aromatics, resins, and asphaltenes (SARA) were obtained using the scheme shown in Figure 4. The SARA
Figure 6. Precipitation amounts for first-contact dead oil experiments at different CO2 concentrations.
results, the precipitation onset is between 20 and 30 mol % CO2, which is consistent with the findings by Graue and Zana.11 Precipitate amounts were less than 1%, even at CO2 concentrations of 76 mol %. Carbon Dioxide First Contact with Live Oil. In the live oil experiments, the gas/oil ratio (GOR) was kept constant at 155 standard cubic feet/stock tank barrel (scf/stb) to reflect average field values. The gas used contained about 91% methane, 4% ethane, and 5% propane. The gas produced in the field had more constituents; the composition was simplified, and only the main constituents were employed. The composition of the live oil was determined from calculations performed using a thermodynamic program (Winprop from Computer Modeling Group, Inc.). The composition of the live oil (measured by ASTM D2887) is shown in Figure 7. Concentrations of CO2 (mole fractions) are reported with
Figure 4. Separation scheme used to obtain SARA fractions from the parent oil.
fractions of the crude are shown in Figure 5. The asphaltene fraction of the oil determined as pentane insolubles was found to be 2.8 wt %, and the heptane insolubles were determined as 1.4 wt %. 2674
dx.doi.org/10.1021/ef201402v | Energy Fuels 2012, 26, 2672−2679
Energy & Fuels
Article
Table 1. Composition of the Precipitates from Live Oil Experiments CO2 (mol %) compound class
0
11
28
38
55
66
saturates aromatics resins asphaltenes
60.42 20.09 5.66 8.83
44.61 18.40 4.50 16.65
36.26 19.84 6.16 34.09
38.65 20.98 7.18 30.27
37.82 21.00 7.11 33.60
38.67 18.32 7.97 35.01
precipitates. From the SARA results of precipitates obtained from thermodynamic experiments, it can be concluded that, when a significant amount of precipitation occurs (28 mol % CO2), the quality of the precipitates is similar. The data also show that CO2-induced precipitates contain about 35 wt % asphaltenes (defined as pentane insolubles). It is observed that, at 80 mol % CO2, the CO2-induced precipitate is 4.45 wt %, while the pentane insolubles in the precipitate are only 1.58%. Simulated Multiple-Contact Experiments. A light− intermediate cut from the same crude oil (up to C15) obtained by the ASTM D86 distillation procedure was used to simulate the process of enrichment of the gas phase through extraction of the light and intermediate fractions of the oil. This was performed to emulate the multiple-contact miscibility process, and the hypothesis was that this would increase CO2-induced precipitation. Cumulative carbon number distribution of the light cut added is shown in Figure 9. The fraction added to the
Figure 7. Composition of the reconstituted live oil.
respect to dead oil instead of live oil. All of the experiments were performed at 3000 psia and 160 °F. Amounts of precipitates at different CO2 concentrations are shown in Figure 8. Pentane insolubles in the solids are also
Figure 8. Precipitation amounts for first contact live oil experiments.
shown in the figure. The precipitate amounts compared to equivalent dead oil experiments were 3−4 times higher. The experiments covered a complete range of CO2 concentrations, including a blank, at 0 mol %, and up to 80 mol %. These results confirm that the precipitation starts between 20 and 30 mol % CO2.11 The onset concentrations for the Weyburn crude oil were in 39−46 mol % CO2.10 The maximum amount of precipitate was about 4%. The data also show that the CO2 concentration with respect to the dead oil required for precipitation of asphaltenes does not change significantly as we go from dead oil to live oil. The addition of paraffinic light gases appears to destabilize asphaltenes, and larger precipitates are obtained with live oils containing about 10% higher carbon number alkanes (ethane and propane). Table 1 shows the solids compositions determined from the SARA analysis. The precipitation in the blank experiment (0 mol % CO2) is 0.22 wt % of the original oil. From the SARA analysis of this precipitate, the concentration of asphaltenes (defined as pentane insolubles) is only 8.83 wt % of the precipitate, which means that the precipitate obtained in the filters, in this experiment, must also correspond to occluded oil. Precipitate from the experiment with 11 mol % CO2 is small (0.45 wt % of the original oil); its asphaltene content is higher than for the blank experiment but low compared to other
Figure 9. Cumulative carbon number distribution of the light cut added to the crude oil to test if this increases the amount of CO2induced asphaltenes.
oil to simulate multiple-contact composition amounted to 30% (molar) with respect to the live oil and 48% with respect to the dead oil. The precipitation with the addition of the light−intermediate fraction to the oil is plotted in Figure 10. With the addition of the “light−intermediate” fraction, the amount of precipitate is approximately twice the amount for an equivalent CO2 concentration without the cut (in comparison to the live oil precipitates). This suggests that multiple-contact mixtures are likely to yield twice the amount of precipitate expected through first-contact experiments. With the addition of the light cut, the precipitation is also more sensitive to CO2 concentrations and increases more rapidly as the CO2 concentration increases. The concentrations of pentane insolubles in the precipitates were 27.49, 43.77, and 42.62% for experiments with 25, 48, and 68 mol % CO2, respectively. These CO2-induced precipitates 2675
dx.doi.org/10.1021/ef201402v | Energy Fuels 2012, 26, 2672−2679
Energy & Fuels
Article
A comparison of the solids obtained in true multiple-contact experiments to single-contact precipitate amounts is shown in Figure 12. The multiple-contact precipitate amounts are about
Figure 10. CO2-induced precipitates with simulated multiple-contact experiments. Pentane insolubles in the precipitated solids are also shown.
Figure 12. Precipitate amounts with first- and true multiple-contact experiments.
contain more pentane insolubles than the precipitates from the dead or live oils, indicating that the solids are heavier. True Multiple-Contact Miscibility. True multiple-contact miscibility is achieved by vaporizing gas drive mechanism (forward contacts), in which the gas phase is enriched through extraction of the light and intermediate fractions of oil. The overall composition of the oil−CO2 mixture for the first contact was 76 mol % CO2 and 24 mol % dead oil. This mixture was equilibrated in one of the PVT cells at a pressure below the bubble point of this mixture. Most of the vapor phase from this mixture was transferred isobarically to the second PVT cell, which already contained fresh crude essential for the second contact. The remaining mixture in the first PVT cell was displaced through the high-pressure filters into the fluid-sampling system. The vapor phase from the second contact was displaced in a similar fashion into a third-contact PVT cell (the first PVT cell was prepared to be the third-contact cell, once all of the sample was displaced from it). Prior to this displacement, this PVT cell was also charged with the necessary crude oil. The remainder of the sample from the second contact was displaced through a separate line of filters into the fluid-sampling system. The third contact yielded miscibility. Solid, liquid, and gas sampling procedures were repeated for the third contact. This procedure is illustrated in Figure 11.
2−3 times the first-contact precipitate amounts for the same CO2 concentrations. Pentane insolubles in the precipitates were about 35% of the precipitate, and resins were about 7%. The basic hypothesis that the process of creating multiple contacts enhances asphaltene precipitation was verified through these experiments.
■
SOLIDS ANALYSIS The solids, liquids, and gases generated in all of the experiments were analyzed. Solids were analyzed using column chromatography, gas chromatography, gas chromatography− mass spectrometry (GC−MS), field ionization mass spectrometry, time-of-flight mass spectrometry, nuclear magnetic resonance, and Fourier transform infrared (FTIR) spectrometry. Only selected results are provided. Simulated distillation (ASTM D2887) results of a few samples are shown in Table 2. Table 2. Simulated Distillation Results of Solid Samples from the Field and Laboratory (wt %)a sample
eluted fraction
non-eluted fraction
C40
field C5-field C5-sol-field RO-lab resins
28.91 4.47 93.54 71.73 14.13
71.09 95.53 6.46 28.27 85.87
25.44 0 83.55 61.11 9.97
The “field” samples were collected from the field undergoing CO2 enhanced oil recovery. C5-field refers to pentane insolubles in the sample, while C5-sol-field refers to the soluble portion of the field sample. RO-lab refers to the CO2-induced precipitate from the dead oil experiment at 30 mol % CO2. Resins are from the parent oil. a
It is expected that the compounds that elute from the field sample would be in the pentane-soluble portion of the same. The RO-lab precipitate has a chromatogram similar to the chromatograms for the field precipitate and the pentane soluble in the field precipitate; 47.83% of the RO-lab sample was pentane-insoluble. GC−MS analysis revealed the presence of occluded oil in all samples, as indicated by the SARA analysis on the solids reported in the previous sections.
Figure 11. Protocol and parameters for the true multiple-contact experiments. 2676
dx.doi.org/10.1021/ef201402v | Energy Fuels 2012, 26, 2672−2679
Energy & Fuels
Article
ratios for the different precipitates indicate shorter substituted alkyl chains for CO2-induced precipitates compared to paraffinic insolubles in the same oil. The aromatic/carbon− oxygen bonding ratios are closer to 1 for the field and labinduced CO2 precipitates, indicating a relatively higher concentration of heteroatomic species. The scanning electron micrographs of solids (and asphaltenes) obtained by various means from the RO oil are presented in Figure 14. The CO2-induced precipitate from
A direct comparison of the SARA distributions of the field solids and CO2-induced laboratory precipitates (dead oil with 30 mol % CO2) is provided in Table 3. Table 3. SARA Distribution in CO2-Induced Solids fraction
field
RO-lab
saturates aromatics resins asphaltenes (C5 insolubles)
12.71 9.96 2.67 74.12
32.14 13.42 4.61 47.83
The analysis reveals that the solids are not all asphaltenes. The laboratory sample contains a significantly higher percentage of saturates. The asphaltenes in the field solid sample in comparison to the laboratory sample are much higher. The exact process by which the field sample was deposited was not known. The laboratory sample (RO-lab) possibly includes occluded oil. Conditions at which the solids are formed in the field and in the laboratory may be vastly different. Once the solids are formed in the field, they may have been subjected to other solvent treatments. Xylene is often used to treat solids. Residual xylene in the wellbore may have dissolved the saturate portion of the precipitate, making it more aromatic/polar. It is also unclear what role “weathering” plays on the composition of the precipitated solids. Thus, there is a variety of reasons why the solids from the field differ in composition from the solids obtained in the laboratory. The FTIR signatures (Figure 13) of the solids were similar, with some remarkable differences. The presence of methyl and Figure 14. Scanning electron micrographs of the various asphaltene samples. The “field” samples were obtained as solid precipitates in a field undergoing carbon dioxide enhanced oil recovery.
thermodynamic experiments was generated at high pressure in-line filters (2 μm), and the field precipitate obtained from the precipitate accumulated in submersible pumps in the field undergoing a carbon-dioxide-enhanced process was tested. The morphology of all of the samples was irregular, and the fractions were amorphous. The pentane insolubles in the field sample appeared cleaner with the dissolution of some of the extraneous smaller particles. The sample that looked the most agglomerated was the sample from the thermodynamic experiment (RO-lab). This was expected because it is an unwashed sample. When this sample was analyzed by SARA analysis, the results showed that only 45% of it was pentaneinsoluble, indicating that it has saturates, naphthenaromatics, and resins that can contribute to its appearance like a solid instead of particles. Scanning electron microscopy (SEM) images quantified using energy-dispersive spectroscopy (EDS) showed a gradual increase in sulfur moieties as the molecular weight of the solvent increased with CO2-induced precipitates falling between C6 and C7. The field deposits contained the highest amount of sulfur.
Figure 13. FTIR signatures of the precipitates obtained using the different indicated solvents (lab referring to CO2) in comparison to the field solids.
methylene groups is common in all of the precipitates. The intensity of these groups is higher for the precipitate from the CO2 experiment possibly because of the presence of occluded oil. Intensity ratios are tabulated in Table 4. CH2/CH3 intensity
■
Table 4. Important Features in the FTIR Signatures of Precipitated Solids sample CO2-induced asphaltenes field asphaltenes pentane asphaltenes from the same oil heptane asphaltenes from the same oil
CH2/CH3 CH2/CO
VALIDATION OF EXPERIMENTAL DATA USING THE HOMOGENEOUS MOLECULAR THERMODYNAMIC MODEL Several recent papers have appeared on modeling CO2-induced precipitation.19 The approach in this paper is based on the Flory−Huggins polymer solution theory. The crude oil is treated as a mixture of two liquid phases. The first phase is pure
CC/CO
1.2 1.1 1.4
10 4 3
1 1 3
1.5
4
5 2677
dx.doi.org/10.1021/ef201402v | Energy Fuels 2012, 26, 2672−2679
Energy & Fuels
Article
asphaltene liquid phase that acts as a solute, and the second phase consists of the remaining components of the crude oil, which act as a solvent. The molar volume (VA) and solubility parameter (δA) of the asphaltene are evaluated by fitting the model with the experimental data. The molar volume (VL) and solubility parameter of liquid phase (δL) are calculated from the Peng−Robinson equation of state (EOS). The model flowchart is shown in Figure 15.
properties of the pseudo-components of the live oil. The pseudo-components were determined using the CMG WinProp Program. The live oil was lumped from 43 components into 11 hypothetical components. A good agreement between experimental results for live oil precipitation experiments and the homogeneous molecular thermodynamic model was observed (Figure 16). Optimized
Figure 16. Comparison of the modeling and experimental results for the precipitation experiments with the live oil.
values of the asphaltene solubility parameter and molar volume were 302 psia0.5 and 0.43 ft3/mol, respectively. This parameter fitting approach is adequate for the systems looked at in this study. A more rigorous fundamental approach described by Gonzalez et al.19 may be necessary to extend the results of this study beyond the parameter space investigated.
■
CONCLUSION Multiple-contact miscibility is the primary mechanism for increasing recovery in CO2 floods for enhanced oil recovery. Multiple contacts enrich the crude oil in the extracted light− intermediate components. The hypothesis in this paper was that the tendency for the formation of CO2-induced precipitates would be enhanced because of the multiple contacts. A high-pressure thermodynamic system was designed and fabricated, and a number of thermodynamic experiments were performed with dead oils (oils with no initially dissolved gas) and live oils (oils with a predetermined quantity of dissolved gas). To test the hypothesis specifically, two types of experiments were conducted: one where a light−intermediate cut (C15) was added to the crude oil, and the second was a true multiple-contact miscibility experiment. Most of the experiments were conducted with a light crude oil with a low original asphaltene content. The onset of precipitation was observed between 20 and 30 mol % CO2 concentrations. Thermodynamic experiments with live oils showed that the precipitate amounts were 3−4 times higher compared to equivalent dead oil solid deposits. With the addition of the “light−intermediate” fraction (simulated multiple contact), the amount of precipitation was approximately twice the amount for an equivalent CO2 concentration without the cut (in comparison to live oil precipitates). The precipitation was also more sensitive to CO2 concentrations and increased more rapidly as CO2 concentrations increased. True multiple-contact experiments were performed with dead oils. The multiple-contact precipitate amounts were about 2−3 times the first-contact precipitate amounts for the same CO2 concentrations.
Figure 15. Flowchart of the modeling procedure using the homogeneous molecular thermodynamic model.
All of the experimental systems were modeled separately. Only the precipitation results of the live oil experiments are discussed in this paper as an example. It was necessary to lump the components into pseudo-components before an EOS could be used efficiently. Table 5 shows the composition and Table 5. Composition and Properties of the Live Oil Components Used in the Modeling pseudocomponents
composition (mol %)
Tc (K)
Pc (atm)
acentric factor
C1 C2 C3 C5 C6 C7 FC8−FC12 FC13−FC17 FC18−FC25 FC26−FC30 FC31−FC40 FC41+
23.110 0.941 1.317 6.224 5.233 2.879 22.701 15.200 11.680 4.064 4.141 2.5115
190.6 305.4 369.8 469.6 507.5 543.3 620.28 714.433 794.19 856.18 904.11 1067.85
45.4 48.2 41.9 33.3 32.46 30.97 25.09 18.45 13.82 10.83 8.98 4.09
0.225 0.098 0.152 0.251 0.275 0.3083 0.4344 0.6368 0.8525 1.036 1.1716 1.6191 2678
dx.doi.org/10.1021/ef201402v | Energy Fuels 2012, 26, 2672−2679
Energy & Fuels
Article
(13) Monger, T. G.; Fu, J. C. The nature of CO2-induced organic deposition. SPE Tech. Pap. 1987, 147−159, DOI: 10.2118/16713-MS. (14) Monger, T. G.; Trujillo, D. E. Organic deposition during CO2 and rich-gas flooding. SPE Reservoir Eng. 1991, 17−24. (15) Novosad, Z.; Constain, T. G. Experimental and modeling studies of asphaltene equilibria for a reservoir under CO2 injection. SPE Tech. Pap. 1990, 599−607, DOI: 10.2118/20530-MS. (16) Verdier, S.; Carrier, H.; Andersen, S. I.; Daridon, J.-L. Study of pressure and temperature effects on asphaltene stability in presence of CO2. Energy Fuels 2006, 20, 1584−1590. (17) Takahashi, S.; Hayashi, Y.; Yazawa, N.; Sharma, H. Characteristics and impact of asphaltene precipitation during CO2 injection in sandstone and carbonate cores: An investigative analysis through laboratory tests and compositional simulation. SPE Tech. Pap. 2003, DOI: 10.2118/84895-MS. (18) Okabe, H.; Takahashi, S.; Mitsuishi, H. Distribution of asphaltene deposition in the rock samples by gas injection. SPE Tech. Pap. 2010, DOI: 10.2118/138765-MS. (19) Gonzalez, D.; Vargas, F.; Hirasaki, G.; Chapman, C. Modeling of CO2-induced asphaltene precipitation. Energy Fuels 2008, 22, 757−762. (20) Parra-Ramirez, M. Precipitation and characterization of asphaltene. Ph.D. Dissertation, University of Utah, Salt Lake City, UT, 2002.
The analysis of precipitates from thermodynamic experiments and solids obtained from the field revealed that the CO2induced precipitates are not all asphaltenes (defined as either pentane- or heptane-insoluble solids). The laboratory sample contained significant fractions of saturates. The asphaltenes in the field solid sample compared to the laboratory sample were higher. The homogeneous molecular thermodynamic model gave a good representation of the precipitation of solids, especially for dead and live oils. The model prediction was in close agreement with the experimental data after regression of asphaltene properties, such as molar volume, solubility parameter, and molecular weight.
■
AUTHOR INFORMATION
Corresponding Author
*E-mail:
[email protected]. Present Address ‡
After a career in Ecopetrol, Martha Parra now resides in Columbia.
■
ACKNOWLEDGMENTS We thank the National Energy Technology Laboratory, U.S. Department of Energy (DOE), for financial support. Academic licenses to the CMG Software from Computer Modeling Group, Calgary, Alberta, Canada, are gratefully acknowledged. Crude oil and field asphaltene samples were provided by our industrial partners. We also acknowledge the help of colleagues from the Petroleum Research Center at the University of Utah.
■
REFERENCES
(1) Speight, J. G. The Chemistry and Technology of Petroleum, 3rd ed.; Marcel Dekker: New York, 1999. (2) Speight, J. G. In Asphaltenes and Asphalts; Yen, T. F., Chilingarian, G. V., Eds.; Elsevier: Amsterdam, The Netherlands, 1994; Vol. 1, p 10. (3) Sharma, G. D. Characterization of Oil and Gas Reservoir Heterogeneity; U.S. Department of Energy (DOE): Washington, D.C., 1994; Final Report DOE/ID/12839-12 (DE94005526), p X1. (4) Kokal, S. L.; Sayegh, S. G. Asphaltenes: The cholesterol of petroleum. SPE Tech. Pap. 1995, 169−181, DOI: 10.2118/29787-MS. (5) Leontaritis, K. J.; Mansoori, G. A. Asphaltene flocculation during oil production and processing: A thermodynamic colloidal model. SPE Tech. Pap. 1987, 149−158, DOI: 10.2118/16258-MS. (6) Hervey, J. R.; Iakovakis, A. C. Performance review of miscible CO2 tertiary project: Rangely Weber Sand Unit, Colorado. SPE Reservoir Eng. 1991, 163−168. (7) Leontaritis, K. J.; Amaefule, J. O.; Charles, R. E. A systematic approach for the prevention and treatment of formation damage caused by asphaltene deposition. SPE Tech. Pap. 1992, 157−164, DOI: 10.2118/23810-MS. (8) Burke, N. E.; Hobbs, R. E.; Kashou, S. F. Measurement and modeling of asphaltene precipitation. J. Pet. Technol. 1990, 1440− 1446. (9) Nighswander, J. A.; Kalra, H.; Majeed, A. Effect of low molecular weight hydrocarbons on the onset concentration and bulk properties of asphaltene at in-situ temperature and pressure conditions. Proceedings of the American Institute of Chemical Engineers (AIChE) Meeting; Houston, TX, March 28−April 1, 1993; Paper 55E. (10) Srivastava, R. K.; Huang, S. S. Asphaltene deposition during CO2 flooding: A laboratory assessment. SPE Tech. Pap. 1997, 617− 626, DOI: 10.2118/37468-MS. (11) Graue, D. J.; Zana, E. T. Study of possible CO2 flood in Rangely field. J. Pet. Technol. 1981, 33 (7), 1312−1318. (12) Hirschberg, A; deJong L.N. Schipper, B. A.; Meijer, J. G. Influence of temperature and pressure on asphaltene flocculation. Soc. Pet. Eng. J. 1984, 283−293. 2679
dx.doi.org/10.1021/ef201402v | Energy Fuels 2012, 26, 2672−2679
