Amel A Nimer,, Abdelshakour A. Mohamed and Ali A.

Rabah

NILE BLEND CRUDE OIL: WAX SEPARATION USING MEK-TOLUENE MIXTURES

* Amel A Nimer, Abdelshakour A. Mohamed and Ali A. Rabah
Department of Chemical Engineering, University of Khartoum P.O. Box 321, Khartoum, Sudan

:
, )% 78.52 (. ) .(MEK ) (28-70C )5-03( ) 1:61 1:23( ).(50-100%MEK % 29 75V%MEK 50C 02 1:02 0.17C , , , , , , . )61 (C10-C %6 +54 C %01.

_______________
:*Corresponding Author E-mail: rabahss@hotmail.com ________________________________________________________________________________________________________ 9002 ,21 Paper Received May 14, 2009; Paper Revised September 10, 2009; Paper Accepted October

0102 October

The Arabian Journal for Science and Engineering, Volume 35, Number 2B

71

Amel A Nimer,, Abdelshakour A. Mohamed and Ali A. Rabah

ABSTRACT This paper presents the operational conditions for wax production from Nile Blend; a Sudanese oil, characterized as sweet oil with a high level of wax content (25.87 wt%). The method of solvent extraction was employed using a Methyl Ethyl Ketone (MEK)- Toluene mixture as a solvent. The study covered a wide range of mixing temperatures (28 to 70 oC) and residence time (5 to 30 minutes), solvent to oil ratio (16:1 to 32:1), and solvent compositions (50100%MEK). The result showed that 92% of the wax in the crude can be extracted with a mixture of 75v% MEK at a mixing temperature of 50 oC, residence time of 20 minutes, and solvent to oil ratio of 20:1 at a filtration temperature of -17 oC. The wax produced is characterized by a slightly greasy feel and is soluble in Naphtha, Toluene, Xylene, Kerosene, Benzene, Ethyl Alcohol, Acetone, and Carbon Tetra Chloride. The gas chromography analysis showed that the light products (C10-C16) constitute 6% of the produced wax and C45+ is 10%. Key words: Nile blend, dewaxing, MEK-Toluene mixtures

18

The Arabian Journal for Science and Engineering, Volume 35, Number 2B

October 2010

Amel A Nimer,, Abdelshakour A. Mohamed and Ali A. Rabah

NILE BLEND CRUDE OIL: WAX SEPARATION USING MEK-TOLUENE MIXTURES

1. INTRODUCTION On delivery to the refinery, the storage of waxy crude stream is difficult as it produces massive gel and sludge in the storage tank [1]. This raises pumping expenses and cleaning costs. To overcome this problem, chemical additives, known as pour point depressants (PPD), are state-of-the-art practice. Recent studies have indicated that these chemicals are the prime cause of corrosion in equipment and piping systems of the refinery [2]. If the wax could be removed prior to shipment to the internal refinery process, the problems of accumulation in storage and piping system corrosion caused by chemical additives could be minimized. These are the justifications of this work. Due to the fact that the commercial production of Nile Blend started in the late 1990s, the available literature regarding its thermodynamic characteristics are extremely limited. On the other hand, the dewaxing process is mostly aimed at the feedstock of lube oil, the bottom product of atmospheric or vacuum distillate. Hence, the available literature on dewaxing is mostly concerned with dewaxing of lube oil feedstock. There are mainly two different methods for wax removal: solvent and catalytic. Solvent dewaxing is more selective for removing both heavier normal and non-normal hydrocarbons. Catalytic dewaxing removes the normal paraffins more evenly over the boiling range, while the light non-normal hydrocarbon are removed more selectively. Oil obtained from solvent dewaxing has higher yield and viscosity index than that of catalytic dewaxing [3]. The removal of wax from oil by means of solvent extraction involves mixing crude oil with solvent to form a solution followed by cooling (chilling) the mixture (oil + solvent + wax) to form wax-crystal-lattice and, finally, separating the crystal lattice from the solution. The dewaxing is influenced by many parameters such as type of solvents, cooling rate, and temperature and solvent to oil ratio. Several pure solvents or mixtures of solvents have been in use for solvent dewaxing. These include Methylisobutylketone (MIBK), Dichloromethane, Trichlorethylene, and a mixture of Methyl Ethyl Ketone (MEK) and Toluene and MEK/Benzene. Pure Toluene is an excellent oil solvent, and, has good solvent power for wax as well. If it is used alone, a tight lattice of wax will form that hinders filtration [4]. In contrast, MEK shows low solvent power to paraffinic compound (i.e., low selectivity) and as such it precipitates the wax very well. Previous studies on Arabian light oil indicated that optimum wax separation occurs at MEK-toluene solvent composition ranges from 4075 v% MEK [4, 5]. Dewaxing of West Siberian crude with MEK-Toluene mixtures indicated that with increasing MEK in the mixture, the filtration time increases (i.e., the filtration viscosity increases) and the oil yield decreases. With increasing MEK, the viscosity of the oil index increases, but the solid point remains constant [6]. In solvent dewaxing, most of the energy consumption goes into pumping and regeneration of solvent. Therefore, the current trend is to use lower solvent dilution ratios and smaller amounts of solvent in washing the precipitate (wax cake) on filters. At the other hand, an increase in solvent to oil ratio will enhance the filtration rate because solvent prompts the crystal growth and lowers the viscosity of the mixture (oil + solvent). It also increases the dewaxed oil yield and decreases the oil content in the wax yield. The previous literature indicates that the solvent to oil ratio is between 16:1 to 32:1 on mass basis [7]. The rate of cooling (high or low) has a strong influence on crystal formation and particle characteristic. For example, a high cooling rate is found to promote the formation of small crystals but they are of the needle type that clogs the filter. Previous experiments showed that the optimum cooling rate varies between 0.564.4 oC per minute depending on the type of crude oil. The chilling or filtration temperature is usually in the order of -20 oC [8]. Due to the fact that crude oil differs from place to place and as such the operation conditions for wax extraction varies from crude to crude, the aim of this work is to study the operational conditions for wax production from Nile Blend. The influence of solvent composition, mixing temperature, residence time, and solvent to oil ratio on wax extraction will be studied. 2. MATERIALS AND METHODS It has to be mentioned that all the experimental analyses were performed at the Central Petroleum Laboratory (CPL), Ministry of Energy and Mining, Sudan. The level of accuracy of the equipment and apparatus herein mentioned are supplied by CPL. 2.1. Materials Two liters of Nile Blend, a grade of Sudanese crude oil, was provided by the Ministry of Energy and Mining. Table 1 shows the result of the analysis of this sample. The analysis was performed in CPL using standard methods

October 2010

The Arabian Journal for Science and Engineering, Volume 35, Number 2B

19

Amel A Nimer,, Abdelshakour A. Mohamed and Ali A. Rabah

as indicated in the said table. Table 2 shows the properties of MEK and Toluene used in this work. These are the manufacturers data. The manufacturer for both chemicals is Fischer Scientific, UK. Table 1. Properties of Nile Blend Property Test Method Salt Content ASTM D 3230 Density at 15 oC ASTM D 5002 Wax Content UOP 46 Flash Point IP 170 Pour Point ASTM D97 Kinematic Viscosity at 50 oC ASTM D 445 Kinematic Viscosity at 80 oC ASTM D 445 Water Content by Distillation ASTM D 4006 Ash content ASTM D 482 Asphaltene Content IP 143 Sulfur Content ASTM D4294

Unit mgNaCl/L kg/m3 wt% o C o C mm2/s mm2/s v% wt% wt% wt%

Value 9.12 850.4 25.87 21.5 33.0 22.64 8.884 0.09 0.043 0.11 0.051

Table 2. Properties of MEK and Toluene (Manufacturer 's Data) Property Molecular Formula Molecular Weight Boiling Point oC Melting Point oC Flash Point oC Vapor Pressure at 20 oC (mmHg) Vapor Density 15 oC Lower Explosive Limit % Upper Explosive Limit % Solubility in Water at 20 oC Manufacturer Purity w% 2.2. Experimental Apparatus Figure 1 shows a schematic presentation of the dewaxing processes. These include mixing and heating, cooling (chilling), filtration, and evaporation. The mixing and heating were carried out using a magnetic stirrer with a hotplate. It was type Heidolph Hei-Tec Magnetic Stirrers manufactured by Brinkmann and Eppendorf. It has a speed up to 1000 rpm. The level of uncertainty of the rotational speed is 1% of the maximum speed. The chiller is a thermostat type FP50 (HD) manufactured by Julabo (Germany). The thermostat employes ethylene glycol and is capable of producing cooling and heating levels in the range of -25 to 50 oC. Its highly accurate PID temperature controller with digital readout has an accuracy of 0.02K. The filtration system consists of a Gooch crucible placed on a flask and the flask is connected to a vacuum pump. On top of the Gooch crucible glass, wool was placed. The vacuum pump is type N 035 AN and is capable of producing a vacuum pressure of 10-4 mbar. A rotary evaporator type RE 100 (BIBBY) was used to evaporate the solvent and Naphtha from wax. The heating is provided using water bath type RE 100B.
Naphtha
Naphtha + traces of solvent

MEK C4H8O 72 79.6 -86 -6 77.5 2.42 2 10 27.5% Fischer Scientific, UK 98

Toluene C7H8 92.1 111 95 4 22 3.14 1.2 77 Nearly Insoluble Fischer Scientific, UK 98

Solvent Crude oil Mixer and heater Chiller Filtration Oil + solvent + traces (wax) Evaporator Wax + traces (oil)

Figure 1: Schematic of experimental setup

20

The Arabian Journal for Science and Engineering, Volume 35, Number 2B

October 2010

Amel A Nimer,, Abdelshakour A. Mohamed and Ali A. Rabah

Besides these apparatuses, a four decimal digital balance type TOLEDO AB204S manufactured by METTLER of Switzerland was used. It has a range of 10mg to 210 g with an accuracy of 1.0 mg. A potable platinum thermometer is used to monitor the temperature during heating and cooling (chilling) of the mixture. The thermometer has a level of accuracy of 0.5K. Gas Chromatography (flame ionization detector) model Chrompac CP 9001 was used for wax analysis. The GC has a column-type wall coated open tube WCOT fused silica. The column length and inside diameter are 25m and 0.25mm, respectively. It has a retention time repeatability within 0.1 minute. The tests were performed in accordance with ASTM D5442. 2.3. Experimental Design and Procedure In the present work, there are eight parameters to be investigated. These are three temperatures (heating, cooling, and evaporation), two retention times (mixing and chilling), one mixing speed, one solvent to oil ratio, and one mixture concentration. If each parameter is varied twice in accordance with a 2k factorial method, there will be 256 experiments. With a minimum replicate of three, this will be infeasible due to high cost and time. An alternative method of fractional factorial designs was used. Here, only the parameters that best describe the problem are studied. In this process, the mixing speed, chiller and evaporator temperatures, and chiller resident time are kept constant in this work. Each of the rest of the parameters is replicated at least three times. The number of experiments and their procedures are given as: 1. Material preparation: 720 samples each of 2 g of crude oil are prepared. MEK - Toluene mixtures of different concentrations (50, 75, and 100 %MEK) are prepared; a known amount of MEK is added to a known amount of Toluene. 2. Heating and mixing: The desired amount of solvent is added to the 2 g of oil. The flask is tightly closed and manually shaken then placed in the magmatic stirrer (the stirring ball is placed inside the flask). The mixing is done gradually; starting at low speed and increasing towards full speed. The mixing time is varied from 530 minutes with 5 minute intervals (6 replicates) at a temperatures range of 28 to 70 oC [28, 40, 50, 60, and 70 oC]. 3. Chilling: After mixing and heating, the flask with its content is transferred to the chiller. It is placed in the ethylene glycol bath of the thermostat. Since the desired chilling temperature is -17 oC, the thermostat is set at -20 oC. This is made to reduce the cooling time (increase the cooling rate). When the bulk temperature reaches -17 oC, the thermostat temperature is adjusted to and left for an additional 30 minutes at -17 oC. It was observed that the bulk temperature reaches the desired level within 30 minutes. This is equivalent to a cooling rate of up to 2.25 oC per minute. 4. Filtration: The wax formed in the mother solution was filtered under vacuum through glass wool in a Gooch crucible. The flask containing the mother solution was rinsed with 20 ml of fresh solvent at -17 oC to ensure that all components have been removed. The wax collected on the glass wool was washed with 100 ml of Naphtha. 5. Evaporation: The Naphtha and the traces of solvent were driven off under vacuum evaporation. The water bath temperature is maintained at 95 oC. The wax in the flask is then left for 24 hours in a desiccator containing P2O5. This was made to dry the wax from the traces of Naphtha and solvent. The produced wax is then weighed and analyzed using GC. The test method ASTM D 5442 was used. The produced wax was subjected to touch inspection. In addition, its solubility in Naphtha, Toluene, Xylene, Kerosene, Benzene, Ethyl Alcohol, Acetone, and Carbon Tetra Chloride was checked. 3. RESULTS AND DISCUSSIONS The extracted wax, here called yield X, is determined as a percentage of the total wax in crude oil as
X = Wi 100% WT

Wi is extracted wax in grams for experiment i, and WT is the total wax in the crude oil sample in grams. In the present work, the total wax is 25.87 w% of the sample. 3.1. Influence of Mixing Time and Temperature Figures 2 (a) and (b) show the wax yield at various mixing temperatures and residence times. 1. The wax yield increases along the line of isotherm with time till it reaches a point beyond which it asymptotes. It begins to asymptote at residence time of about 20 minutes. It can be concluded that for all lines of isotherm at different solvent to oil ratios, the residence time of 20 minutes gives the highest yield. Although there are few cases in which the maximum yield occurs at retention time of about 25 minutes, the increase in wax yield

October 2010

The Arabian Journal for Science and Engineering, Volume 35, Number 2B

21

Amel A Nimer,, Abdelshakour A. Mohamed and Ali A. Rabah

between 20 and 25 minutes does exceed 1 %. In turn, this does not justify the energy consumed to maintain the mixture at the desired temperatures for the said difference in the residence times. A comprehensive economical analysis for this study is found in [9]. 2. The wax yield increases with temperature at constant residence time. This is expected since heating decreases viscosity and enhances solvent distribution in the mixture. It can also be seen that, at residence time of 20 minutes, the difference in the wax yield at 50 oC and that at 70 oC does exceed 5%. Similar to retention time, this relatively small increment does not justify the energy required to raise the mixture (oil, wax, and solvent) temperature by 20 oC. It should be remembered that the mixture will be cooled again, which requires additional energy for this purpose.

100 95 90

a) 75% MEK

100 95 90 85 28 40 80 75 70 65 60

b) 50% MEK

Yield %

85 80 75 70 65 60 5 10 15 20 25 30

28

Temperature oC

50 60 70

Temperature

oC

40 50 60 70

10

15

20

25

30

Residence Time (min.)

Residence Time (min.)

Figure 2: Influence of temperature on dewaxing at solvent to oil ratio of 16:1

3.2. Influence of Solvent to Oil Ratio Figures 3(a) and (b) show the influence of the solvent to oil ratio on wax yield. As mentioned earlier, the increase in solvent will enhance the filtration rate because solvent prompts the crystal growth and lowers the viscosity of the mixture (oil + solvent). The result clearly confirms these facts. However, the effect becomes almost asymptotic at a solvent to oil ratio of 20:1.
100 95 90 100

a) 75% MEK

95 90

a) 50% MEK

Yield %

85 80 75 70 65 60 5 10 15 20 25 30

Yield %

85 80 75 70 65 60 5 10 15 20 25 30

Solvent to Oil Ratio

16.1 20:1 28:1 32:1

Solvent to Oil Ratio

16.1 20:1 28:1 32:1

Residence Time (min.)

Residence Time (min.)

Figure 3: Influence of solvent to oil ratio on dewaxing at a mixing temperature of 50 oC

3.3. Influence of Solvent Compositions Figure 4 shows the influence of solvent compositions in wax extraction. As expected, an increase in the ratio of MEK in the solvent enhances the wax yield. Increasing MEK in the solvent from 50% to 75% enhanced wax recovery by 3.5 to 5.0% at the same operational conditions. This is because MEK prompts the crystallization of wax and enhances the filtration rate. At 100 wt% MEK, high wax recovery is obtained. It was observed during the experimental run that the filtration rate is slower than that with MEK-Toluene mixtures. This is expected since pure MEK is known to form a third phase in addition to the existing solvent/oil phase, wax/solvent phases. The new phase consists of molecules which are insoluble in MEK that clogs the filter [4].

22

The Arabian Journal for Science and Engineering, Volume 35, Number 2B

October 2010

Amel A Nimer,, Abdelshakour A. Mohamed and Ali A. Rabah

100 95 90

Yield %

85 80 75 70 65 60 5 10 15 20 25 30

MEK content in the solvent

100%v 75%v 50%v

Residence Time (min.)

Figure 4: Influence of solvent composition on wax removal at a mixing temperature of 50oC and solvent to ratio of 20:1

3.4. Wax Characteristics The wax produced at the operational conditions of 75% MEK solvent composition, 20:1 solvent to oil ratio, 50 C mixing temperature, and 20 minute residence time at -17 oC filtration temperature is inspected and found to be slightly greasy to the touch. It is also found to be soluble in Naphtha, Toluene, Xylene, Kerosene, Benzene, Ethyl Alcohol, hot Acetone, and Carbon Tetra Chloride. Table 3 shows the result of GC. The light hydrocarbon (nC10nC16) constitutes about 6% and C45+ is 10 %. The light n-paraffins may originate from the oil entrapped within the wax.
o

Table 3. GC Analysis for Produced Wax Carbon Number nC10 nC11 nC12 nC13 nC14 nC15 nC16 nC17 nC18 nC19 nC20 nC21 nC22 nC23 nC24 nC25 nC26 nC27 Mass% 0.225 0.633 0.994 1.257 1.380 1.474 1.575 1.986 2.190 2.968 4.354 6.279 7.104 7.805 7.115 6.853 6.232 6.045 Total Carbon Number nC28 nC29 nC30 nC31 nC32 nC33 nC34 nC35 nC36 nC37 nC38 nC39 nC40 nC41 nC42 nC43 nC44 C45+ Mass % 4.957 4.445 3.349 2.705 1.945 1.493 1.073 0.810 0.650 0.496 0.398 0.299 0.228 0.174 0.184 0.165 0.137 10.023 100.000

4. CONCLUSION This study provides experimental data on dewaxing of Nile Blend. The operational conditions for wax production from Nile Blend are 75% MEK, 20 minute retention time, and 20:1 solvent to oil ratio at -17 oC cooling and a 30 minute cooling time. Under these conditions, the wax contains 6% of light n-paraffins (nC10-nC16) and 10% of C45+. ACKNOWLEDGMENT The authors acknowledge the support of the Central Petroleum Laboratory (CPL) at the Ministry of Energy and Mining, Sudan.

October 2010

The Arabian Journal for Science and Engineering, Volume 35, Number 2B

23

Amel A Nimer,, Abdelshakour A. Mohamed and Ali A. Rabah

REFERENCES
[1] [2] [3] [4] [5] [6] [7] [8] [9] F. G. Ruben, T. P. Visintina, R. L. Lockharta, and D. Paolo, Structure of Waxy Crude Oil Emulsion Gels, Journal of Non-Newtonian Fluid Mechanics, 149(13)(2008), pp. 3439. H. V. Doctor and H. D., Mustafa, Crude Oil Cleaning With Solvent, Hydrocarbon Asia, 1(2003), pp. 6264. J. T. Robert and J. M. Arthur, Study of Solvent and Catalytic Lube Oil Dewaxing by Analysis of Feedstocks and Products, Ind. Eng. Chem. Res., 31(7)(1992), pp. 17311738. J. P. Wauquier, Petroleum Refining: Separation Process Part (2). Edition Technip. Paris, 2000. E. B. Evans, Modern Petroleum Technology. London: Institute of Petroleum, 1962. B. V. Gryaznov, V. B. Krylov, and T. V. Chernyaeva, Influence of Solvent Composition on Deep Dewaxing of Raffinates, Chemistry and Technology of Fuels and Oils, 18(9)(1982), pp. 439441. J. J. Meketta, Petroleum Processing Handbook. New York, 1992. H. Lei and Z. Jin-Jun, Viscoelasticity of Gelled Waxy Crude Oil, Journal of Central South University of Technology, 14(1)(2007), pp. 414417. A. A. Nimer, Production of Petroleum Wax from Nile Blend, PhD thesis, Department of Chemical Engineering, University of Khartoum, 2008.

24

The Arabian Journal for Science and Engineering, Volume 35, Number 2B

October 2010