This article was downloaded by: [informa internal users] On: 4 January 2011 Access details: Access Details:

[subscription number 755239602] Publisher Taylor & Francis Informa Ltd Registered in England and Wales Registered Number: 1072954 Registered office: Mortimer House, 3741 Mortimer Street, London W1T 3JH, UK

Petroleum Science and Technology

Publication details, including instructions for authors and subscription information: http://www.informaworld.com/smpp/title~content=t713597288

Thermodynamic Prediction of Paraffin Wax Precipitation in Crude Oil Pipelines

L. Oyekunlea; O. Adeyanjua a Department of Chemical Engineering, University of Lagos, Akoka-Yaba, Lagos, Nigeria Online publication date: 02 December 2010

To cite this Article Oyekunle, L. and Adeyanju, O.(2011) 'Thermodynamic Prediction of Paraffin Wax Precipitation in

Crude Oil Pipelines', Petroleum Science and Technology, 29: 2, 208 217 To link to this Article: DOI: 10.1080/10916460903330163 URL: http://dx.doi.org/10.1080/10916460903330163

PLEASE SCROLL DOWN FOR ARTICLE

Full terms and conditions of use: http://www.informaworld.com/terms-and-conditions-of-access.pdf This article may be used for research, teaching and private study purposes. Any substantial or systematic reproduction, re-distribution, re-selling, loan or sub-licensing, systematic supply or distribution in any form to anyone is expressly forbidden. The publisher does not give any warranty express or implied or make any representation that the contents will be complete or accurate or up to date. The accuracy of any instructions, formulae and drug doses should be independently verified with primary sources. The publisher shall not be liable for any loss, actions, claims, proceedings, demand or costs or damages whatsoever or howsoever caused arising directly or indirectly in connection with or arising out of the use of this material.

Petroleum Science and Technology, 29:208217, 2011 Copyright Taylor & Francis Group, LLC ISSN: 1091-6466 print/1532-2459 online DOI: 10.1080/10916460903330163

Thermodynamic Prediction of Parafn Wax Precipitation in Crude Oil Pipelines

L. OYEKUNLE1 AND O. ADEYANJU1
1

Department of Chemical Engineering, University of Lagos, Akoka-Yaba, Lagos, Nigeria

Abstract Crude oils are generally very complex chemical systems consisting predominantly of hundreds to thousands of hydrocarbon compounds from simple lowmolecular-weight parafns to high-molecular-weight waxes. At low temperatures these parafns precipitate as a wax phase, which may cause the plugging of pipes and many other problems. In this study the cloud point and the amount of the precipitated wax at different temperatures were determined using a modied multi-solid wax model at stock tank conditions and at high pressure for three live crudes and synthetic oil. The model is based on the description of the nonideality of the phases in equilibrium using the Peng-Robinson cubic equation of state for the liquid phase with a modied afunction and the universal quasi chemical (UNIQUAC) equation for the solid phase, which takes into account the effects of temperature, pressure, and composition on the wax precipitation. The model assumes the crystalline nature of wax formed. The experimental data and model predictions reveal that the cloud point decreases with increased pressure and increases when the crude oil is mixed with more light hydrocarbons. The values predicted by the new model show very little deviation from experimental data. Keywords crude petroleum, parafns, thermodynamic modeling, wax deposition

Downloaded By: [informa internal users] At: 14:47 4 January 2011

Deposition of crystalline parafnic waxes is frequently encountered during the production and transportation of crude oil and its products. In the petroleum industry, wax precipitation causes severe operational problems by plugging the valves and the ow lines. One of the most important measurements that characterize the waxing potential of uids is the cloud point. This is the temperature at which cloudiness of a relatively transparent petroleum uid is observed upon cooling. As long as the operating temperature remains above the cloud point, wax precipitation will not occur. Typically, cloud point measurements used in industry are inadequate. The standard ASTM D2500-88 or IP 219/82 cloud point measurement methods are only applicable for clear uids because they rely on visual observation of the wax crystals. For dark crude oils, cloud points are estimated from viscometry and differential scanning calorimetry (DSC). Oftentimes, the viscosity or DSC data are misinterpreted. A more serious problem with these methods is that sufcient wax must crystallize before the cloud point can be detected. A number of engineering models have been proposed for calculating oil-wax equilibria (Hansen et al., 1988; Erickson et al., 1993; Prausnitz et al., 1999; Banki and FiroozAddress correspondence to Layioye Oyekunle, Department of Chemical Engineering, University of Lagos, Akoka-Yaba, Lagos 101017, Nigeria. E-mail: layioyekunle@yahoo.com

208

Prediction of Parafn Wax Precipitation

209

abadi, 2002). Wax precipitation may occur in reservoir formation, wellbores, production tubings, submersible pumps, surface equipment, and surface pipelines. The present study is limited to precipitation in pipelines where the primary mechanism for wax precipitation is thermal cooling. Early models focused on low-pressure uids (Pederson et al., 1991; Coutinho and Stenby, 1996; Lira-Galeana et al., 1996). The need for a model that could describe the wax formation in live oil at reservoir condition promoted the development of the model for high pressures (Ungerer et al., 1995; Pan and Firoozabadi, 1997; Lindeloff et al., 1999).

Model Development
The model of Coutinho and Stenby (1996) for dead oil is the starting point. This model treats the solubility of parafn as ideal and contains no effect of pressure on the solid liquid equilibria. The general solidliquid equilibrium relates, for each compound, the fugacity of both phases in the standard state, f , with the pure component thermophysical properties. KiS f ol Hi D ln os D f f RTi l Cp s R ln T Ti
f f

Downloaded By: [informa internal users] At: 14:47 4 January 2011

Ti T C

1 C !

Hitr RTitr

Titr T

(1)

f Ti

Equation (1) can be expressed in terms of the solid- and liquid-phase compositions .x s ; x l / and the activity coefcients . s ; l / xs ln l x
s l

Hi RTi

Ti T ln

Hitr 1 C RTitr T T C i T
f

Titr T

(2)

l Cp s R
f f

Ti

!!

where Ti and Hi are the freezing point temperature and heat of fusion, respectively. For normal parafns of molecular weight, MWi : Ti;normal D 374:5 C 0:02617 For all other components, Ti;non-normal D Ti;normal Hi D Ci MW i Ti where Ci is the constant (D0.163 cal/g K).
f f f f

MW i

20172 MW i

(3)

90
f

80CNi 100 C CNi

(4) (5)

210 Effect of Pressure

L. Oyekunle and O. Adeyanju

The effect of pressure on the fugacity of a pure condensed phase is given by the equation: si fioL .P Po / L L D Ki D oS exp .vi xi RT fi viS / D KiS exp .P Po / vi RT (6)

where KiS is given in Eq. (1). Previous studies show that for carbon number greater than 10, the liquid and solid molar volumes of normal parafns are a linear function of molecular weight (Nelson et al., 1960). The difference between the liquid and solid molar volumes can then be expressed as in Eq. (7): vi D C2 M Wi
Downloaded By: [informa internal users] At: 14:47 4 January 2011

(7)

Variation of Melting Points with Pressure At the melting point, the solidliquid K value is given as KiL D exp Hi RTi l Cp s R
f

Ti Tm Tm
f Ti

1 C
f Ti

Hitr RTitr !!

Titr Tm .P

D 1:0

(8)

ln

Tm

exp

Po / vi RT

Equation (8) can be simplied to: KiL C1 Mi Ti D exp RTm

f

Ti Tm

C2 Mi .P Po / 1 C RTm

D 1:0

(9)

Solving for the melting temperature, T , and differentiating with respect to pressure and setting C2 D 0:17 cm3 /g, C1 D 0:163 cal/g K and making the unit conversions gives:
d Tm C2 C D D 0:025 dP C1 bar

(10)

Solid-Phase Fugacity The solid-phase fugacity .fi S / is related to the liquid-phase fugacity .fi L / by: " fi L .T; P; x/ fi
ig

f S .T; P / fiL .T; P / D xi i L fi .T; P / fiig

"

(11)

The ratio of the fugacity of component i in the solid phase to the fugacity of component i in the liquid phase is the inverse of the K values dened by Eq. (6).

Prediction of Parafn Wax Precipitation xi

L i P

211 (12)

xi iL P KiL
^L i L i

si D KiL Ki D xi Hitr C RTitr

^L i

^L i L i

Pvi exp RT l Cp s R ln

T
f

Hi exp RT T C i T
f

"

1 !#

T Ti
f

! (13)

Titr T

Ti

Downloaded By: [informa internal users] At: 14:47 4 January 2011

is the fugacity coefcient of component i in the liquid mixture and iL is the fugacity of pure liquid component I ; are evaluated at the system temperature and pressure. Equation (13) is the nal solidliquid K value for the model. All that remains is to choose a suitable equation of state to calculate the fugacity coefcient. The PengRobinson equation of state was chosen because it correlates well with experimental solidliquid equilibria data. Fusion and Transitional Properties The melting point, Ti ; the enthalpy of fusion, Hi ; and the heat capacity of fusion, Cpi , are evaluated for P, N, A species using following correlations (Won, 1989): Ti D 374:5 C 0:02617Mi Hi D 0:1426Mi ti
f f f f f

20172=Mi

(14) (15)

The correlations used for the normal parafns are based on the data by Cavetts (1964) and Marano and Holders (1997) approach. Ti D 421:63
f f

1936412 exp. 7:8945.Ci 0:2376Ci2 7:400Ci

1/0:07194 34:814

(16) (17)
f

Hi D 0:00355Ci3

The correlations proposed by Lira-Galeana et al. (1996) were used to calculate Ti f for naphthenes, iso-parafns, and aromatics species and Hi for naphthenes and isoparafns. Ti D 333:5
f f

419 exp. 0:00855Mi /

f

(18) (19)

Hi D 0:05276Mi Ti

The correlation proposed by Pan and Firoozabadi (1997) was used to estimate the enthalpy of fusion for aromatic species Hi D 11:2Ti
f f

(20)

The correlation of Pedersen et al. (1991) was used to estimate Cpi for all P, N, and A species. Cpi D 0:3033Mi 4:635 10 4 Mi T (21)

212

L. Oyekunle and O. Adeyanju

Critical Properties and Acentric Factor The work of Pan and Firoozabadi (1997) provided a general correlation for critical properties. Pci D A Be
CMi

(22)

For hydrocarbon with molecular weight of more than 300 g/mol coefcients A, B, and C in Eq. (22) for the P, N, and A species were obtained from Pan and Firoozabadi (1997). Riazi and Al-Sahhafs (1995) expression was used to calculate the acentric factor of parafns: !D 0:000185397Ci3 C 0:0448946Ci 0:0520750 (23)

Downloaded By: [informa internal users] At: 14:47 4 January 2011

Model Validation
The developed model was tested using the experimental data from the literature (Pan and Firoozabadi, 1997). Four crude oil samples 14 of different compositions were employed. The modied multi-solid model developed (Eq. (11)) was used to calculate the wax precipitation of the oil samples. The interaction coefcients between CO2 and N2 were set at 0.002, the interaction between CO2 and hydrocarbon and between N2 and hydrocarbon given by Pan and Firoozabadi (1997) was used. The interaction coefcient between C1 and another hydrocarbon, Cj , were estimated from the equation: kcij D 0:0289 C 1:633 10 4 Mj (24)

The interaction between the normal alkane/non(normal alkane and the nonnormal alkane/normal alkane are set to zero. The interaction coefcients between two normal alkanes Ci and Cj as proposed by Pan and Firoozabadi (1997) were calculated by: kij D 6:872 10
2

C 3:6

10 6 Mi2

8:1

10 4 Mi

1:04

10 4 Mj

(25)

where kij D 0:0 (Mi Mj and Mi 255), and Mi and Mj are the molecular weights of light and heavy normal alkanes, respectively. Characterization The two-parameter gamma distribution function (Isehunwa and Falade, 2007) was rst used to generate the molar distribution of the plus fraction of the crude (plus fraction is the last available residue; for sample 1 oil it is C40C fraction). Then the distribution was lumped into three to four pseudocomponents of nearly equal weight. Hydrocarbons from C10 to the plus fraction were also lumped in groups of ve consecutive carbons. The parafns naphthenes aromatics (PNA) distribution of the four oil samples (14) and naphthenes aromatics (NA) distribution for sample 2 oil were estimated. Erickson et al. (1993) have shown that the amount of normal parafn follows an exponential decay with carbon number using high-temperature gas chromatography and the amount of normal alkanes becomes negligible at very high molecular weights. An exponential decay approximation was used to estimate the distribution of normal alkanes of all pseudocomponents for samples 1, 3, and 4. The aromatic distribution of the pseudocomponents for the four

Prediction of Parafn Wax Precipitation

213

crudes is approximated according to the experimental data by Jaubert (1995). It was also assumed that the fraction is resin-like if the molecular weight is between 800 and 950 g/mol and is asphaltene-like if its molecular weight exceeds 950 g/mol. Therefore, the amount of parafn and naphthenes is set at zero for the heavy pseudocomponents with molecular weight exceeding 800 g/mol.

Results and Discussion

The proposed model was tested against experimental data and cloud points were predicted for four oil samples at different pressures. Comparisons between the model predictions and the set of data are presented in Figures 13, and the percentage weight of wax precipitation is shown in Figures 4 and 5. The various plots revealed the fact that there is a signicant effect of pressure on the cloud point and the amount of wax precipitated.
Downloaded By: [informa internal users] At: 14:47 4 January 2011

Figure 1. Comparison of experimental data and model prediction of cloud point for sample 1.

Figure 2. Comparison of experimental data and model prediction of cloud point for sample 2.

Figure 3. Comparison of experimental data and model prediction of cloud point for sample 3.

214

L. Oyekunle and O. Adeyanju

Figure 4. Hydrocarbon composition of oil samples at 1 bar, mol%.

Downloaded By: [informa internal users] At: 14:47 4 January 2011

Figure 5. Plot of wax precipitated against temperature for sample 1 at different pressures.

Effect of Pressure The Figures 13 show that the model predictions agree quite well with the experimental results, with the cloud point decreasing as the pressure increases, for a liquid mixture of xed composition. The model predictions are a little higher than experimental results: an average of 4.5 K for sample 1, 8.2 K for sample 2, and 4.3 K for sample 3. These are probably due to the parafn contents of the oil samples, especially the higher percentage of lighter fraction in sample 2 (51.3% C1 ). The total average absolute deviation percentage (%AAD) between the calculated and experimental data was found to be 1.44%. In Figure 1 a smooth curve was obtained by model prediction while there is a turning point in the experimental results. The reason for this could be due to the uncertainty of the experimental data because no turning point was observed with Figures 2 and 3. In both Figures 5 and 6 it is clearly revealed that wax precipitation decreases with increases in pressure as well as temperature. This is due to the increase in solubility of heavy fractions in the liquid phase at higher pressure. Figure 5 for oil sample 1 shows an almost linear relationship and Figure 6 for oil sample 2 can be represented by two lines of different slopes. These are indications that the wax contents of the oils are of different complexities. The simple explanation is due to the higher content of light fractions (51% C1 ) of sample 2. The plots of wax precipitated against temperature for sample 3 are very similar to those of sample 2. Effect of Composition Figure 4 presents the composition of the four oil samples. The composition mainly differs in the content of methane (C1 ) and heavy C10C components. Sample 1 contains high C10C

Prediction of Parafn Wax Precipitation

215

Figure 6. Plot of wax precipitated against temperature for sample 2 at different pressures.

Downloaded By: [informa internal users] At: 14:47 4 January 2011

Figure 7. Comparison of experimental and model prediction of cloud point for oil sample 4.

components and samples 2 and 3 have high contents of methane (C1 ). Sample 4 oil has very low light fractions but contains the largest quantity of C10C components (>70%). The effect of composition on wax precipitation was examined by using synthetic oil (sample 4) obtained from the literature (Pan and Firoozabadi, 1997). The effect of pure hydrocarbon such as C3 , n-C5 , n-C7 , n-C10 , n-C12, and n-C15 on solubility of solid wax precipitation was investigated by determining the cloud point temperature of different mixture of the synthetic oil and about 30% each of the hydrocarbons C1 to C15. Figure 7 shows the experimental and model-predicted cloud points when various hydrocarbons are mixed. It is noted here that the cloud point initially decreased from C1 to C5 hydrocarbons and then increased rapidly. These results agree with the experimental data. The results are also in line with the solubility of heavy hydrocarbon (>n-C30) in light hydrocarbon solvents (n-C5 -n-C12; Coutinho et al., 2001).

Conclusions
A thermodynamic model has been developed to predict the effects of light ends and high pressure on solid parafn formation in crude oils based solely upon a detailed crude oil compositional analysis. The model prediction was validated against some experimental data from the literature. The model prediction for the combined effects of light ends and pressure and the experimental data agreed to within the experimental uncertainty of the data. The model results also show that the cloud point is strongly inuenced by heavy normal parafns, whereas the amount of precipitated wax depends on the concentration of n-parafns in solution with isoparafns, naphthenes, and aromatics. The inuence of pressure on wax formation is shown to be as important as the inuence of temperature.

216

L. Oyekunle and O. Adeyanju

The developed model can provide an appropriate description of wax precipitation over a wide pressure range.

References
Banki, R., and Firoozabadi, A. (2002). Modelling of wax deposition in pipelines from irreversible thermodynamics. Paper No. SPE 77571, 2002 SPE Annual Technical Conference and Exhibition, San Antonio, TX, October. Cavett, R. H. (1964). Physical data for distillation calculationVapour liquid equilibrium. API Pro. Sec. III, 42. Coutinho, J. A. P., Pauly, J., and Daridon, J. L. (2001). A thermodynamic model to predict wax formation in petroleum uids. Braz. J. Chem. Eng. 18:411422. Coutinho, J. A. P., and Stenby, E. H. (1996). Predictive local composition models for solidliquid and solidsolid equilibrium in n-alkanes: Wilson equation for multicomponent systems. Ind. Eng. Chem. Res. 35:918925. Erickson, D. D., Niesen, V. G., and Brown, T. S. (1993). Paper No. SPE 26604, 1993 SPE Annual Technical Conference and Exhibition, Houston, TX, October 36. Hansen, J. H., Fredenslund, A., Pedersen, K. S., and Rnningsen, H. P. (1988). A thermodynamic model for predicting wax formation in crude oils. AIChE J. 34:19371942. Isehunwa, O. S., and Falade, G. K. (2007). Improved characterization of heptanes-plus fractions of light crudes. Paper No. SPE 111918, 31st Nigeria Annual SPE International Conference and Exhibition, Abuja, Nigeria, August 68. Jaubert, J. N. (1995). Phase equilibrium calculations on an Indonesian crude oil using detailed NMR analysis or a predictive method to assess the properties of the heavy fraction. Ind. Eng. Chem. Res. 34:640655. Lindeloff, N., Heidemann, R., Andersen, S. I., and Stenby, E. H. (1999). Phase-boundary calculations in systems involving more than two phases, with application to hydrocarbon mixtures. Ind. Eng Chem Res. 38:11071113. Lira-Galeana, C., Firoozabadi, A., and Prausnitz, J. M. (1996). Thermodynamics of wax precipitation in petroleum mixtures. AIChE J. 42:239248. Marano, J. J., and Holder, G. D. (1997). General equation for correlating the thermophysical properties of n-parafns, n-olens, and other homologous series. 2. Asymptotic behavior correlations for PVT properties. Ind. Eng. Chem. Res. 36:18951907. Nelson, R. R., Webb, W., and Dixon, J. A. (1960). The solubility of high molecular weight parafns in aromatic or naphthene base. J. Chem. Phys. 33:17561764. Pan, H., and Firoozabadi, A. (1997). Pressure and composition effect on wax precipitation: Experimental data and model results. Paper No. SPE 36740, SPE Annual Technical Conference and Exhibition, Denver, CO, October, pp. 250259. Pedersen, K. S., Schonberg, P., and Ronningsen, H. P. (1991). Wax precipitation from North Sea crude oils. 4. Thermodynamic modelling. Energ. Fuel. 5:924932. Prausnitz, J. M., Lichtenthaler, R. N., and Azevedo, E. G. (1999). Molecular Thermodynamics of Fluid-Phase Equilibria, 3rd ed., Upper Saddle River, NJ: Prentice-Hall. Riazi, M. R., and Al-Sahhaf, T. A. (1995). Physical properties of n-alkanes and n-alkyl hydrocarbons application to petroleum mixtures. Ind. Eng. Chem. Res. 34:41454148. Ungerer, P., Faissat, B., Zhou, H., Leibovici, C., Behar, E., Moracchin, G. I., and Courcy, J. P. (1995). High pressure-high temperature reservoir uids: Investigation of synthetic condensate gases containing a solid hydrocarbon. Fluid Phase Equil. 111:287311. Won, K. W. (1989). Thermodynamic calculation of cloud point temperatures and wax phase positions of rened hydrocarbon mixtures. Fluid Phase Equil. 53:377396.

Downloaded By: [informa internal users] At: 14:47 4 January 2011

Nomenclature
CNi Cp carbon number heat capacity of fusion, cal/Kmol

Prediction of Parafn Wax Precipitation f ig fi L fi fiO;L fiO;S fiS f Hi tr Hi K Kij MWi P Po R Si T f Ti Tm Ttr V v viL viS v x fugacity of pure, condensed phase, bar fugacity of component I as an ideal gas, bar fugacity of component I in liquid phase, bar fugacity of pure component I in liquid phase, bar fugacity of pure component I in solid phase, bar fugacity of component I in solid phase, bar heat of fusion for component i, cal/mol heat of transition for component I, cal/mol solidliquid K value binary interaction parameter for components I and j molecular weight of component I, g mol pressure, bar atmospheric pressure, bar gas constant, J/Kmol mole fraction of component I in solid phase temperature, K freezing/melting temperature for component I , K melting point temperature, K transition point temperature, K volume molar volume molar volume of component i in liquid phase, cm3 /mol molar volume of component i in solid phase, cm3 /mol difference between liquid and solid molar volumes for component i mole fraction

217

Downloaded By: [informa internal users] At: 14:47 4 January 2011

Greek Letters ^L i L i fugacity coefcient of component i in liquid mixture fugacity coefcient of pure liquid component i