Journal of Industrial and Engineering Chemistry 19 (2013) 536–539

Contents lists available at SciVerse ScienceDirect

Journal of Industrial and Engineering Chemistry

journal homepage: www.elsevier.com/locate/jiec

Used lubricating oil regeneration by various solvent extraction techniques

Saeed M. Al-Zahrani *, Meilana Dharma Putra
Department of Chemical Engineering, King Saud University, P.O. Box 800, Riyadh 11421, Saudi Arabia

A R T I C L E I N F O A B S T R A C T

Article history: The used lubricating oil regeneration using extraction technique has been widely known as one of the
Received 1 May 2012 cheapest and most competent processes. The best performance of solvent was shown by R-11 as the
Accepted 5 September 2012 lowest percent oil losses, followed by CFC-113, MEK, 1-butanol, and 2-propanol. This performance
Available online 13 September 2012
sequence could be also evaluated by method of solubility parameter difference (Es) using Hildebrand and
Peng–Robinson EoS approaches. The effectiveness of extraction parameters was also studied. The percent
Keywords: oil losses decreased with the extraction temperature. The values of optimum ratio of solvent to oil
Solvent
obtained by the Es method gave reasonable results compared to the experimental method.
Used lubricating oil
ß 2012 The Korean Society of Industrial and Engineering Chemistry. Published by Elsevier B.V. All rights
Recycle
Extraction reserved.
Temperature
Peng–Robinson EoS
Hildebrand

1. Introduction using appropriate solvent extraction. Other parameters i.e. ratio of

solvent to oil, extraction condition and type of solvents are also
Lubricating oil that has good characteristics shows a very studied. The solvents of methyl-ethyl-ketone, 2-propanol and 1-
important role in the industry. Lubricants used in internal butanol have been used in system of used oil recycling [4], while in
combustion engines could be injured by various strains, depending the recent contribution the extraction process has been conducted
on operating conditions, fuel quality, ambient conditions and using those solvents followed by adsorption on solids [6]. An
operating parameters [1]. Without the appropriate treatment, the optimum extraction using solvents, namely alcohols, ketones, and
used oil will tend to be discarded that degrades the environment hydrocarbons, has been attempted by different methods
[2]. Different techniques have been proposed for recycling of used [2,5,7,8,12,13]. These contributions are further related to the
lubricating oils. Amongst the alternative processes projected in optimization of the process based on the ability of solvent to
recent years, solvent extraction process has received substantial separate sludge from used oil.
interest [3–11]. The solvent extraction is a simple process by The experimental testing and/or the key properties such as
mixing used oil and solvent with desired proportions, which of solubility are necessary to evaluate the solvent performance [8]. The
course, ensure complete miscibility of the base oil in the solvent. correlation between the removal of percent sludge from used oil
The additives and carbonaceous impurities that are normally with the difference of solubility (solvent and polymer compound)
present in used oil, should be rejected by the extracting solvent. has been conveyed in those literatures [7,11]. An increase of
These impurities settle and flocculate based on gravity weight. The solubility parameter difference, further enhances the solvent
solvent is then recovered by distillation for recycle reason [3]. capability to extract the additives and impurities from used oil
The simple description of recycling of used oil process is shown [4]. The anti-solvency energy (Es) has been proposed [8] to elucidate
in Fig. 1 as more detailed diagram was presented elsewhere [4]. the effect of solvent and polymer volumes, as well as to account for
The contaminants of the used oil about 10–14% can be removed by the effect of extraction temperature on the solubility of both
solvent extraction process [5]. This system discards the maximum compounds. Further, the Es was defined as a difference of solubility
amount of sludge particles from used oil and may also lose the parameter between solvent and used lubricating oil. The values
minimum amount of base oil in the sludge phase [4]. The essential which resulted from the correlation between Es and either ratio of
contribution of this research is to minimize the percent oil losses solvent to oil or percent sludge removal, were well-fitted with the
experimental results [14] and further could be used for the
parameters of the solvent extraction. In this investigation, the
* Corresponding author.
difference of solubility parameter (Es) between solvents and oil
E-mail address: [email protected] (S.M. Al-Zahrani). would be correlated with the percentage of oil losses in sludge phase.

1226-086X/$ – see front matter ß 2012 The Korean Society of Industrial and Engineering Chemistry. Published by Elsevier B.V. All rights reserved.
http://dx.doi.org/10.1016/j.jiec.2012.09.007
 S.M. Al-Zahrani, M.D. Putra / Journal of Industrial and Engineering Chemistry 19 (2013) 536–539 537

Solvent Clay The prediction of solubility parameter difference calculated by

Recovery Treatment both methods above would be used to evaluate the optimum ratio
Solvent Clay Sludge of solvent to oil as compared with the experimental method. The
molar weight and molar volume of the base oil have been predicted
Filtration by determining density and API degree of the base oil experimen-
Extraction
tally.
Sludge
2. Experimental
Used Oil Recovered Oil
Settlement Five extracting solvents, i.e. 2-propanol, 1-butanol, methyl-
ethyl-ketone (MEK), trichloro-trifluoroethane (CFC-113) and tri-
chloro-fluoromethane (R-11) were used in process of recycling of
Water & used lubricating oil. Nimir [12] has described the experimental
Sediment
technique and the conditions of separation of the sludge phase
Fig. 1. Schematic diagram of used lubricating oil recycling using solvent extraction. (additive, impurities, and carbonaceous particles) from solvent:oil
phase. The different ratios of solvents to oil were varied at 0.8, 1,
1.2, 2, 3, 4, and 5. Both solvent (msol) and used oil sample
The solubility parameter difference (Es) was calculated using (moil = 10 g) were mixed. Then, the mixture was stirred at 300 rpm
the following equation: for 20 min and at extraction temperature (23, 30, 40 and 50 8C).
  This step was carefully conducted to ensure no oil loss in the flask
Es ¼ jd1  d2 j (1)
with proper mixing and appropriate vessel system. The final
  mixture was left for 48 h to settle under gravity action. The black
where d and d are the solubility parameters of solvent and base
1 2
oil, respectively. The solubility parameter of solvent and base oil particle, i.e. the settled sludge observed at the bottom of the flask
(d*) in terms of energy unit (J1/2) was calculated using the following was weighted as ms. The solvent–oil mixture was separated from
equation: the sludge, namely the wet sludge. The wet sludge was then
pffiffiffiffi washed by 2-propanol and n-hexane [7] to anticipate removing
d ¼ d V (2) about 95% of interstitial oil content present in the sludge phase [4].
The washed sludge was heated for 5 min at 100 8C to evaporate
where V is the volume of the base oil and/or solvent in cm3.
excess solvents. The final sludge obtained namely dry sludge was
There are many correlations for predicting the cohesiveness and
reported as md. The percentage of oil losses was calculated based
values of solubility parameter [15]. The solubility parameter d (J/
on the following formula:
cm3)1/2 of the solvents and base oil used in this work would be
calculated by the following equations. ms  md
Percent oil losses ð%Þ ¼  100 (8)
moil
1. The equation of Hildebrand
Hildebrand [16] has proposed an equation for predicting the
3. Results and discussion
solubility parameter (d) of liquid. The prediction of solubility
using this equation has been applied in those literatures [4,15].
Fig. 2 shows the dependence of percent oil losses on the ratio of
The equation is exhibited in Table 1 (Eq. (3)), where DHv is the
solvent to oil for R-11–oil system at various extraction tempera-
vaporization enthalpy (J/mol), vm is the molar volume (cm3/
tures. Four other solvents show similar trends as this R-11–oil
mol), R is the gas constant (8.315 J/mol K), and T (K) is the
system. Percent oil losses decreased with increasing ratio of
temperature. The vaporization enthalpy of the base oil was
solvent to oil. Percent oil losses further decreased with extraction
determined from available physical property charts of lube oil
temperature. On the other hand, the higher extraction temperature
[16].
led to the higher yield of extraction [20]. Thus, more contaminant
2. The equation state of Peng–Robinson (PR EoS)
could be removed. Hence, the extraction process in high
The PR EoS has shown a great contribution in gas and liquid
properties, and has even been applied in many industries. On
other hand, the PR EoS has been used to predict the solubility
parameter of various liquids, hydrocarbon and oil [17–19]. The
equations are shown in Table 1 (Eqs. (4)–(7)), where Tc is critical
temperature (K), Pc is critical pressure (bar) and v is acentric
factor.

Table 1
The equation solubility parameter of Hildebrand and Peng–Robinson EoS.

Method Equation
qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
Hildebrand d ¼ DHnvmRT (3)

   1=2 1=2
aðTÞ
Peng–Robinson EoS d¼ vm ðvm þbÞþbðvm bÞ 1 þ k  TTc (4)

b ¼ 0:0778RT
Pc
c
(5)
    1=2 2
0:45724R2 Tc2 T
aðTÞ ¼ Pc 1þk 1 Tc (6)
Fig. 2. Dependence of percent oil losses on the ratio of solvent to oil for R-11–oil
k ¼ 0:37464 þ 1:54226v  0:26992v2 (7) system at various extraction temperatures.
538 S.M. Al-Zahrani, M.D. Putra / Journal of Industrial and Engineering Chemistry 19 (2013) 536–539

Fig. 3. Dependence of the percent oil losses on the ratio of solvent to oil for R-11–oil
system at 23 8C. Fig. 4. Dependence of percent oil losses on the ratio of solvent to oil for the five
solvents of extraction at 50 8C.

temperature resulted in high sludge removal [12] and low oil cm3)1/2, respectively. This finding is in accordance to the
losses. Further, the higher percent oil losses in the sludge phase led hypothesis [21], in which lower solubility difference between oil
to the lower yield of the extraction process [4]. The yield can be and solvent ((d1  d2), (J/cm3)1/2) led to higher oil miscibility, i.e.
improved with the increase of the number of extraction stages or lower percent oil losses.
designing a system of counter current extraction [9]. It means that Fig. 5 exhibits variation of percent oil losses with Es based on
the cost of the process will increase. However, the optimum Hildebrand and Peng–Robinson methods. At high percent oil
temperature associated with the yield of extraction should be losses, Peng–Robinson and Hildebrand methods showed good
simultaneously evaluated based on the maximum sludge removal, agreement of Es values, though, at low percent oil losses the
minimum oil losses and minimum process cost. In addition, the agreement of Es values became less. However, both curve trends
type of solvent is also involved in the selection of the appropriate are similar to the curve of dependence of percent oil losses on ratio
extraction temperature. It is not a simple step to optimize the of solvent to oil. Further, the percent oil losses which decreased
extraction temperature, hence the designers should simultaneous- with the Es showed an agreement to that theory [21].
ly reevaluate these problems by the techno-economic evaluation The relation between the Es and the ratio of solvent to oil is
analysis [4]. shown in Figs. 6 and 7 for Hildebrand and Peng–Robinson methods,
Fig. 3 shows the dependence of percent oil losses on the ratio of respectively. Further, the Es calculated from experimental data
solvent to oil for R-11–oil system at 23 8C. An increase of ratio of shows the dependence on the solvent to oil ratio. The optimum
solvent to oil gave a decrease of the percent oil losses. The high ratio could be evaluated from both these curves as similar
decrement of percent oil losses occurred at low ratio of oil to technique applied in Fig. 3. Though, Hildebrand and Peng–
solvent i.e. 0.8–1.2, while at high ratio from 2.5 to 5, percent oil Robinson gave the reasonable values of Es as compared in Fig. 5,
losses decreased slowly with the ratio. Further, the region between both showed the identical values of CCR. In addition, the values of
a sharp drop and a gradual decrease led to an optimum ratio in CCR for all solvents and at various extraction temperatures are
which the minimum oil losses in sludge phase were reached while presented in Table 2. Both methods provided the reasonable values
the less effect of solvent to oil ratio occurred. The optimum ratio in of CCR as compared with the values obtained from experimental
this region was called the critical clarifying ratio (CCR). It is clear method (curve of percent oil losses versus ratio of solvent to oil).
that increasing the ratio of solvent to oil enhanced the mutual Therefore, this finding reveals that the optimum ratio of solvent to
solubility of oil in a medium of solvent, thereby decreasing the oil oil obtained through both the methods of Es techniques can be used
losses in the sludge phase [4]. The CCR point could be determined as early prediction in designing extraction process.
by graphical technique on percent oil loss curve against ratio of
solvent to oil. As shown in Fig. 3, the vertical line (evaluated by
using linear regression method based on data at high percent oil
losses) that represents the minimum ratio of solvent to oil, is cut by
the horizontal line (evaluated by using linear regression method
based on data at high ratio of solvent to oil) that represents the
minimum percent oil losses. The CCR is obtained by projecting
vertically the intersection point between the two lines.
The percent of oil loss curves on various ratios of solvent to oil
for five solvents at 50 8C are presented in Fig. 4. At low ratio of
solvent to oil, percent oil losses for both 1-butanol and 2-propanol
were much higher than those of other solvents. However, all
solvents showed the same trend as the percent oil losses decreased
with the ratio of solvent to oil. R-11 provided the lowest percent oil
losses, followed by CFC-113, MEK, 1-butanol, and 2-propanol.
Further, both methods of Hildebrand and Peng–Robinson showed
similar, though not identical, values of the solubility parameter
difference. R-11 has the lowest solubility difference of 0.08 (J/
cm3)1/2, followed by CFC-113, MEK, 1-butanol and 2-propanol Fig. 5. Variation of the percent oil losses with Es for MEK–oil system at 23 8C based
which are 0.8 (J/cm3)1/2, 2.9 (J/cm3)1/2, 5.9 (J/cm3)1/2 and 6.9 (J/ on Hildebrand and Peng–Robinson method.
 S.M. Al-Zahrani, M.D. Putra / Journal of Industrial and Engineering Chemistry 19 (2013) 536–539 539

4. Conclusions

The lubricating oil is a very important material used in

industry. Therefore, the extraction process for recycling of used
lubricating oil as an efficient method is urgently needed for more
study. By this work, the performance of solvent was evaluated
based on effective extraction parameters, i.e. ratio of solvent to oil
and extraction temperature. R-11 amongst other solvents
provided the lowest percent oil losses and further solvent
conducted at higher extraction temperature gave lower percent
oil losses. Optimization of extraction temperature should
consider either maximum sludge removal or minimum oil losses.
The optimum value of ratio solvent to oil associated with the
minimum oil losses was evaluated graphically based on experi-
mental results. The optimum ratio could also be determined by a
Fig. 6. Variation of Es with solvent to oil ratio for MEK–oil system at 23 8C based on method of solubility parameter. The solubility parameter has
Hildebrand method. been used in many industries such as paint manufacturing.
Further, the need of using the solubility parameter is to identify
effective solvent for the extraction process [15]. Hence, both
methods of Hildebrand and Peng–Robinson which evaluate the
solubility parameter and then the optimum ratio of solvent to oil
can be proposed to be applied in the extraction process of
recycling of used lubricating oil.

Acknowledgments

The authors extend their appreciation to the Deanship

of Scientific Research and the Research Center of the College
of Engineering at Kind Saud University for supporting this
work.

References

[1] Y.G. Ko, C.H. Kim, Journal of Colloid and Interface Science 301 (2006) 27.
Fig. 7. Variation of Es with solvent to oil ratio for MEK–oil system at 23 8C based on [2] A. Hamad, E. Al-Zubaidy, M.E. Fayed, Journal of Environmental Management 74
Peng–Robinson method. (2005) 153.
[3] M. Alves dos Reis, Industrial & Engineering Chemistry Research 30 (1991) 2449.
[4] N.O. Elbashir, S.M. Al-Zahrani, M.I. Abdul Mutalib, A.E. Abasaeed, Chemical
Engineering and Processing: Process Intensification 41 (2002) 765.
Table 2 [5] J.H. Sherman, N.D. Danielson, R.T. Taylor, J.R. Marsh, D.T. Esterline, Environmental
The CCR evaluated graphically based on experimental results, Hildebrand and Technology 14 (1993) 1097.
Peng–Robinson methods. [6] J.L. Assunção Filho, L.G.M. Moura, A.C.S. Ramos, Brazilian Journal of Chemical
Engineering 27 (2010) 687.
Solvent Temperature (8C) Experimental Hildebrand Peng–Robinson [7] M.A. Dos Reis, M.S. Jeronimo, Industrial & Engineering Chemistry Research 27
2-Propanol 23 2.04 2.11 2.06 (1988) 1222.
30 1.92 2.00 2.01 [8] N.O. Elbashir, The Effect of Design Parameters on the Solvent Extraction Perfor-
mance for Recycling of Used Oil, M. Eng. Thesis, University of Technology
40 1.77 1.83 1.83
Malaysia, 1998.
50 1.72 1.75 1.76
[9] H.-H. Lai, Minimization of Waste Oil in Oily Sludge by Solvent Extraction, MES
1-Butanol 23 1.95 1.94 1.94 Dissertation, Lamar University, 1989.
30 1.84 1.85 1.85 [10] J. Saunders, Lubricants World 6 (1996) 20.
40 1.75 1.73 1.72 [11] M.A. Reis, M.S. Jeronimo, Industrial & Engineering Chemistry Research 29 (1990)
432.
50 1.71 1.68 1.69
[12] O.M. Nimir, M.I. Abdul Mutalib, R. Adnan, in: Proceedings of RSCE97, Johor,
MEK 23 1.87 1.94 1.94 Malaysia, (1997), p. 420.
30 1.82 1.79 1.80 [13] F. Bendebane, L. Bouziane, F. Ismail, Journal of Industrial and Engineering Chem-
40 1.78 1.72 1.73 istry 16 (2010) 314.
[14] N.O. Elbashir, M.I. Abdul-Mutalib, M. Sahriff, R. Adnan, in: Proceedings of RSCE98,
50 1.68 1.65 1.66
Manila, Philippines, (1998), p. 110.
CFC-113 23 1.79 1.72 1.70 [15] A.F.M. Barton, Solubility Parameters and Other Cohesion Parameters, CRC Press,
30 1.45 1.56 1.55 Boca Raton/London, 1991.
40 1.27 1.52 1.53 [16] J.M.P.J.H. Hildebrand, R.L. Scott, Regular and Related Solutions: The Solubility of
Gases, Liquids and Solid, Van Nostrand Reinhold, NY, 1970.
50 1.25 1.48 1.46
[17] E.K. Goharshadi, M. Hesabi, Journal of Molecular Liquids 113 (2004) 125.
R-11 23 1.76 1.63 1.60 [18] J.Y. Zuo, O.C. Mullins, D. Freed, D. Zhang, Journal of Chemical & Engineering Data
30 1.42 1.56 1.52 55 (2010) 2964.
40 1.28 1.46 1.44 [19] M. Jamshidnezhad, Journal of the Japan Petroleum Institute 51 (2008) 217.
[20] J. Rincón, P. Cañizares, M.T. Garcı́a, Journal of Supercritical Fluids 39 (2007) 315.
50 1.22 1.39 1.40
[21] H.G. Elias, Macromolecules, first ed., Plenum, NY, 1977.