Energy & Fuels 1994,8, 788-792
788
Utilization of Refinery Sludge for Lighter Oils and Industrial Bitumen A. P. Kuriakose* and S. Kochu Baby Manjoorant Department of Polymer Science and Rubber Technology, Cochin University of Science and Technology, Kochi 682 022, India Received November 24, 1993. Revised Manuscript Received February 28, 1994”
This paper reports the data obtained in an attempt to utilize the waste sludge of the Cochin Refineries Ltd., Kochi, India. About 17% of a lighter oil fraction can be recovered from this sludge and the characteristics of the lighter oil are such that it can be used as diesel fuel by blending with other appropriate refinery streams. An attempt was also made to convert the residue left after the removal of the lighter oils into different grades of industrial bitumen. This residue, obtained after vacuum distillation, was heat treated without and with different catalysts. The change in softening point:penetration ratio when heat treated without any catalyst was not enough to meet the specifications of any industrial bitumen. Catalysts like sulfur, FeC13, and P205were able to bring down the penetration sharply but failed to increase the softening point which is a requirement for the different grades of industrial bitumen. AlC13 was found to bring about the different reactions required in the vacuum residue and converted it into some useful grades of industrial bitumen, viz., 65125,75130,85125,and 90115. Possible reaction mechanism involved is also postulated. The optimum conditions of temperature, heating duration, and amount of catalyst required for these different grades was determined.
Introduction In petroleum refineries, a lot of sludge accumulates a t the bottom of tanks where crude oil is stored. This is taken out during periodic tank cleaning and dumped separately in ponds. Also, the bottom portion left behind in furnace oil tanks, LSHS (low sulfur heavy stock) tanks, asphalt tanks, etc. is also taken out at the time of their periodic cleaning and dumped in the above-mentioned ponds. Whatever heavy oil spillages occur during the operation of this petroleum refinery is also dumped into the ponds. About 8000-10000 tonnes of oily sludge is accumulated in the quarry pond of Cochin Refineries Ltd. a t Ambalamugal in Kerala, India. This sludge is an accumulation from the last 20 years and has been exposed to the atmosphere in all seasons. It is likely to increase further by 500-1000 tonnes per annum. A study of the constituents of this sludge has shown’ that it contains approximately 25 9% water, 5 % inorganic solids, and about 70 % hydrocarbons. The hydrocarbon part is reported to contain 7.8 wt % of asphaltenes and has a gross calorific value of about 10 300 kcallkg. The ash content is 4.8% and percentage weight of the different elements in the ash is Fe 23.49; A1 10.57; Ca 1.64; Na 0.57; K 0.46; Ni 0.12; V 0.23; Mg 0.65; Zn 0.21; Ti 0.53; and Mn 0.10, The different methods for the disposal of the sludge considered are (1) burning in a rotary incinerator, (2) burning in a step furnace type incinerator, (3) microbial treatment to convert the hydrocarbons to combustible gases, (4) using in a delayed coker, and (5) separation of water and sediments at elevated temperature using diluents and emulsifiers and subsequent burning.
* Author to whom correspondence should be addressed.
Address for communication: Dept. of Quality Control, Cochin Refineries Ltd., Kochi, India. Abstract published in Advance ACS Abstracts, April 1, 1994. (1)“Report on sludge disposal at Cochin Refinery”. Indian Oil Corporation R&D Report, May 1988, No. 88042. t
0887-0624/94/2508-0788$04.50/0
The above methods help mainly by disposing of the sludge and not by effective utilization. So it was thought worthwhile to see whether this sludge can be utilized as a source of light oils and industrial bitumen. The raw materials used a t present for industrial bitumen are somewhat costly. Industrial bitumen of various grades are manufactured in India2 from the vacuum residue of some imported crudes like Arab mixed, Suez Blend, etc. by air blowing in the presence of catalysts at temperatures of 200-275 O C . It has been shown3 that dehydrogenation and polymerization are involved in air blowing and that only a minor amount of oxygen is added to the asphalt. It is also reported that naphthene aromatics are converted into polar aromatics and then to a~phaltenes.~The following are reported to be the reactions occurring during air blowing of the raw material.5
-
RH
+ 0,
+ R’H R’R’H R’R’H + R”H R’
-
R’
+ HOO’
(R’H = unsaturated compound)
-
disproportionation
R’R’HR’’H
stable products
As a result of the comparatively low concentration of hydrocarbon radicals, there is small probability of their recombination (2R’ R-R) and the interaction of the radicals with oxygen takes place to a smaller extent as follows:
-
(2) Joseph Francis, D.; Antony, T. P. Ind. J. Technol. 1988,26,579-
582. ( 3 ) Corbett, L. W.; Swarbrick, R. E. Proc. Assoc. Asphalt Pauing Technol. 1960, 29, 104. ( 4 ) Rossi, Albert0 Manufacture of blown asphalts-their physical and chemical variations. Bol. Inform. Petr. (Buenos Aires) 1942,19, 37. (5) Antony, T. P. Ph.D. Thesis, Cochin University of Science and Technology, 1989.
0 1994 American Chemical Society
Utilization of Refinery Sludge for Lighter Oils R' ROO'
- + - + - + - + - +
+ 0,
ROO'
+ R'H
ROOH
ROOH
RO'
+ *OH RH + HOO' R"H
H,O, R'H
+ *OH
R'*
'OH
R"'
H,O
R'
H,O,
2 'OH R"
H,O
Various catalysts and oxidizing agents have been proposed for augmenting the air-blowing process of the vacuum residue to give a product having a higher penetration for a given softening point.6 They include ~ u l f u rP , ~z O and ~ ~ F e c l ~ .Since ~ the agents used here cannot be recovered as such, technically they might better be termed chemical reactants than catalysts. In any event, the general effect is to reduce blowing time as well as to change the softening point-penetration relationship. Reduction of blowing time is an economic incentive, whereas the change in the flow properties permits the manufacture to specifications. In his studies, Grunderm a n d o has shown that metal chlorides act as catalysts a t relatively low temperatures and without air blowing, causing condensation and polymerization reactions similar to those obtained in air blowing. He has shown that the best catalyst is AlC13 which converted naphthenic aromatic asphalts by treatment a t 150 "C for 3 h into asphalts of medium to high hardness. In view of the fact that the refinery sludge mentioned earlier contains many useful hydrocarbons and that it accumulates in large quantities in the refinery creating a disposal problem, it was thought worthwhile to study the possibility of converting this sludge into some useful raw material like industrial bitumen without the air-blowing process. In the present study, an attempt was also made to separate the lighter oil fractions from the sludge, characterize them, and see whether they can be blended with appropriate refinery streams. Keeping this view in mind, the sludge obtained was first purified and dehydrated. This was then subjected to vacuum distillation to separate the light oils. About 17% of the dehydrated sludge was recovered as light oil. The residue left was treated with varying amounts of catalysts like sulfur,FeCl3, P2O5, and a t different temperatures ranging from 200 to 275 "C, for time periods varying from 1to 3 h. The products obtained were tested for different parameters and the results compared with different grades of industrial bitumen.
Experimental Procedure Sludgewas collected from the quarry pond of CochinRefineries Ltd., Ambalamugal, Kerala, India. Sulfur, FeC13, P&,, AlC13 (anhydrous),and carbon disulfide used in the study were all of (6) Hoiberg, A. J. Catalysts for use in blowing asphalts. R o c . Assoc. Asphalt Paving Technol. 1950,19, 225. (7) Brooks,B. T. The oxidation of mineral oils by air. Ind.Eng. Chem. 1917,9, 746.
( 8 ) Shearon,W. H. Catalytic asphalt-PhosphorousPentoxideasphalts. Ind. Eng. Chem. 1953,45, 2122. (9)Hampton, W. H.US Patent, August 16 (1949),No. 2479235. (10)Crundermann, Erich Deut. Akad. Wiss.,Leipzig, GerErdoel Kohle 1965, 18(10)780-7.
Energy & Fuels, Vol. 8, No. 3, 1994 789 Table 1: Characteristics of Dehydrated Sludge and Vacuum Rssidue of Sludge vacuum test dehydrated residue characteristic method sludge of sludge specific gravity at 27 O C BS 2000182 1.014 1.017 softening point ("C) IP 58/65 46 52 penetration (1/10nm) IP 49/76 230 41 ductility (cm) IP 32/55 32.5 48 flash point ("C) IP 34/75 >200 >200 solubility in CS2 (wt %) B.S. 4600 99.81 99.78 loss of heating (wt %) IP 45/58 0.93 0.10 kinematic viscosity IP 71/60 474 1500 at 100 O C (cS) total sulfur (wt W ) IP 61/65 3.43 2.1 Table 2 Data of Vacuum Distillation temp on conversionto press. atm press. temp recovery in volume (%) ("C) (mm) (760mm) ( O C P 5 10 15 20 25 30 4
149 193 216 235 249 270
0.6 0.9 0.8 0.8 0.8 1.75
347 397 430 455 414 481 (cracking starts)
initial boiling point = 295 "C.
L.R. grade. To removewater, inorganic materials, etc. the sludge (150 kg) was heat treated in a barrel of 200 L capacity fitted inside with steam coils. It was maintained at 110 f 10 OC for 12 h at which time the sludge was fully dehydrated (tested as per IP 291173). The remaining hot oil was then passed through strainers (60and 40 mesh) to remove solid impurities. The oil thus obtained was highly viscous and solid at room temperature. Its characteristics are given in Table 1. A 163-gsample of this dehydrated sludge was taken in a round-bottomed flask and subjected to vacuum distillation. Hot water was circulatedround the condenser and the receiver so that the waxy distillate coming out as vapour did not stick to the sides of the condenser. The lighter oil fraction thus recovered from the sludge amounts to 17% (see Table 2 for data of vacuum distillation). The residue obtained after vacuum distillation was tested for different parameters. These results are also given in Table 1. According to Bureau of Indian Standards (IS 702-1961)there are 10 different grades of industrial bitumen depending upon the softening point-penetration relationship. They are 65/25, 75/15,75/30,85/25,85/40,90115,105120, 115115, 135110,and 15516. The first figure represents the softening point and the second one penetration. A grade 65/25should have the softening point between 59.5 and 70.5 and penetration between 21 and 29. Attempts to convert the vacuum residue of the sludge to some of the above grades of industrial bitumen were carried out as follows. A 250-g portion of this vacuum residue was heated without any catalyst in a cylindrical can (16.5cm height and 9 cm dia) at 250 "C for 3 h with periodical stirring. The sample was then taken out and tested for the different parameters (Table 3).The above experiment was repeated adding 2 5% each of sulfur, FeCls, Pz05and AlC13. From the results (Table 3) it is seen that AlC13 can bring about appreciable variation in the softening pointpenetration ratio while the other catalysts used were not able to bring about such a significant variation. To determine the optimum concentration of AlC13 and the optimum time and temperature required, the experimentwas further repeated with different percentages of AlCb ranging from 1 to 2.75% and temperatures ranging from 200 to 275 OC for periods varying from 1-3 h (see Figures 1-4). The samples were taken out at definite intervals and tested. The softening point reported in this paper was determined by the ring and ball method according to IP 58/65. Here a steel ball
Kuriakose and Manjooran
790 Energy & Fuels, Vol. 8, No. 3, 1994
Table 3 Data on Heat Treatment of Vacuum Residue at 250 OC without and with Different Catalysts for 3 h catalyst specific softening penetration ductility flash matter soluble in loss on (2%) gravity point ( O C ) (1/10 mm) (cm) point ( O C ) carbon disulfide (wt %) heating (wt %) >300 99.76 1.020 55.0 31 20.5 0.07 Nil 1.028 58.5 23 13.0 >300 99.70 Sulfur 0.05 FeC13 1.024 58.0 26 17.5 >300 99.72 0.05 P2O6
AlC13
56.0 83.0
1.021 1.030
29 20
>300 >300
19.0 5.0
99.74 99.64
0.06 0.04
,
\
0 2030c e 225'C A 2 50*C A 275'C
0 1 % AlCg
e
~*/.AIcI~ A 2.5%AIC13 A 275%AlC13
20
25 PENETRATION, 1 /lOmm
30
90
15
20
PENETRATION. 1/10"
25
30
Figure 1. Effect of different percentages of aluminum chloride on properties of vacuum residue of sludge.
Figure 2. Effect of temperature on properties of vacuum residue of sludge containing 2.5 % aluminum chloride.
(9.5 mm diameter) of specified weight (3.5 g) is placed upon a disk of sample contained within a metal ring of depth 6.4 mm; inside diameter at bottom and top 15.9and 17.5mm, respectively, and outside diameter 20.6 mm. The assembly is heated at a constant rate and the softening point is taken as the temperature at which the sample becomes soft enough to allow the ball enveloped in bitumen to fall the specified distance (25 mm). To determine softening point below 80 O C , a water bath was used for heating while for those above 80 O C a glycerinebath was used. Penetration was determined as per IP 49/76. A penetrometer made by Precision Scientific Co., USA, was used for the purpose. The experiment was conducted at 25 O C for 5 s with a total moving weight of 100 g. The distance in tenths of a millimeter that a standard needle (50 mm long and 1.02 mm diameter) vertically penetrates the sample is reported as penetration. Ductility was determined as per IP 32/55 at 27 "C and at a rate of pull of 50 mm/min. A ductility meter manufactured by Humboldt Manufacturing Co., U.S.A.,was used. The density of water in the bath was adjusted by adding sodium chloride so that the bitumenous thread formed during the test did not touch the bottom of the bath at any time during the test. The distance in centimeters by which a standard briquet having the following dimensionscan be elongated before the thread breaks is reported as ductility (total length = 75 mm, distance between clips = 30 mm, width at mouth of clip = 20 mm; and width at minimum cross section = 10 mm). Flash point was determined by the Pensky-Martens closed method as per IP 34/75, at a heating rate of 5 "C/min and with a stirrer speed of 60 rpm. The temperature at which the vapor above the sample can ignite with
a distinct flash inside the cup on the application of the test flame is reported as flash point. Solubility in carbon disulfide was determined as per IP 47/74 using 2 g of the dry material and 100 mL of carbon disulfide. Loss on heating was determined as per IP 45/58 in a stabiltherm oven (BLUE M Electric Co., USA). A 50-g portion of the sample in the sample container was placed near the circumference of the revolving shelf which is made to rotate at a rate of 5-6 rpm, the temperature being maintained at 163 "C for 5 h after the sample was introduced. Density of the samples was determined as per IP 160/68 usinga hydrometer of range 0.85490g/mL. Recovery was determined as per IP 123/78. A 100-mLvolume of the sample was distilled, and the total volume of the distillate collected in the receiver at 366 O C was recorded as the recovery. Kinematic viscosity was determined as per IP 71/60. The time was measured for a fixed volume of oil to flowthrough the capillaryof a calibrated glass viscometer at 38 "C. The kinematic viscosity of the oil was then calculated from the measured flow time and the calibration constant of the viscometer obtained using freshly distilled water as the primary standard. The diesel index was determined as per IP 21/53. It was calculated using the formula GA/100 where G is the API gravity and A is the aniline point in O F . The aniline point was determined as per IP 2/78. Pure aniline (5 mL) and sample (5 mL) were placed in a tube and mixed mechanically. The mixture was heated at a controlled rate until the two phases became miscible. The mixture was then cooled at a controlled rate and the temperature at which the two phases separated was recorded as the aniline point. Ramsbottom carbon residue was determined as per IP
Energy & Fuels, Vol. 8, No. 3, 1994 791
Utilization of Refinery Sludge for Lighter Oils
Chart 1 density at 15 "C (g/mL) (IP I60/68) recovery at 366 O C (vol % (mL))(IP 123/78) flash point (PMC)("C)(IP 34/75) kinematic viscosity at 38 O C (cS) (IP 71/60) diesel index (IP21/53) carbon residue (Ramsbottom) (wt % ) (IP 14/65) aniline point (OC)(IP 2/78) total sulfur (wt % ) (IP63/65) pour point ("C)(IP 15/67)
A 1% Atcis 0 2 % AlClj
A 25%A C l j 0 Z?T/.AlC13
56
64
72 SOFTENING POINT'C
I
I
80
88
Figure 3. Effect of duration of heat treatment with varying
percentages of aluminum chloride. A 2dC 0 225Oc
A 256C 27dC
0.8923 73 >80 10.6 45 0.29 86.6 1.3 + 21
over CaClz for 20 min and weighed again. The percentage weight of carbon residue was then calculated and reported as Ramsbottom carbon residue. Total sulfur was determined as per IP 61/65. A0.6-g portion of the sample was subjected to combustion using a firing wire of length 100 mm in a bomb of capacity 300 mL containing oxygen at 35 atm pressure. The interior of the bomb and the cup were then washed with distilled water and the washings were collected. The washings were then heated to boiling and 10 mL of barium chloride solution was then added dropwise. Boiling was continued for 5 min more and the sample was then cooled. The supernatant liquid was then filtered through a filter paper (Whatman No. 40) and the precipitate was washed until free from chloride. The paper and the precipitate was then transferred into a weighed crucible and ignited until the residue was white in colour. The crucible was then allowed to cool to room temperature and weighed. The percentage weight of total sulfur was then calculated using the formula 13.73(A/B), where A is weight in grams of barium sulfate and E is weight in grams of the sample taken for test. Pour point was determined as per IP 15/67. The sample was heated in a water bath without stirring to a temperature of 45 OC. The test jar containing the sample was then placed in a vertical position in the cooling bath. At each multiple of 3 O C , the test jar was taken out carefully from the cooling bath and tilted to ascertain whether there is a movement of the oil in the test jar. The complete operation of removal and replacement was done within 3 s. The test was continued until the oil in the test jar showed no movement when the test jar was held in a horizontal position for exactly 5 s. The reading of the test thermometer was recorded as the pour point. Results and Discussion
75
79
63 SOFTENING POINT * C
87
Figure 4. Effect of duration of heat treatment at varying temperatures with 2.5% aluminum chloride. 14/65. A 5-g portion of the sample was introduced into the cocking bulb by means of a hypodermic syringe and the bulb was reweighed. The coking bulb was then placed in the furnace at 550 "C for 20 min. It was then taken out and placed in a desiccator
The initial part of the study demonstrates that about 17% of lighter oils can be isolated from this refinery sludge. Characteristics of the recovered oil (Chart 1)show that it can be blended with other refinery streams to make it useful as high-speed diesel. The catalytic effect on the heat treatment of the vacuum residue of the sludge was also investigated. The results of the action of sulfur, FeCl3, PzOS, and AlC13 on the heat treatment a t 250 "C and a t a catalyst ratio of 2 % for 3 h are shown in Table 3. It is seen that sulfur, FeCl3, and Pz05 are successful in bringing down the penetration sharply but fail to bring up the softening point. But AlC13 not only brought down the penetration sharply but was also able to bring up the softening point to the required level. The pronounced catalytic effect of AlC13 in such polymerization reactions involving olefins can be explained by means of the following ionic mechanism. Since AlC13 has an incomplete octet, it, when added to the olefin, polarizes it to such an extent that it is capable of adding further monomers.l1 At high temperatures, AlC13 can also (11)Rieche, Alfred Outline of Industrial Organic Chemistry;
Butterworth: London, 1964;Chapter 11, p 394.
Kuriakose and Manjooran
792 Energy & Fuels, Vol. 8, No. 3, 1994
Table 4: Data on Heat Treatment of Vacuum Residue at 250 OC with Different Percentages of AlClJ for Varying Durations amount of duration of heat specific softening flash matter soluble loss on AlC13 (5%)
treatment (h)
gravity
point
penetration
1.0 1.5 2.0 2.5 3.0 1.0 1.5 2.0 2.5 3.0 1.0 1.5 2.0 2.5 3.0 1.0 1.5 2.0 2.5 3.0
1.023 1.023 1.024 1.026 1.026 1.025 1.026 1.028 1.030 1.030 1.028 1.028 1.029 1.032 1.034 1.028 1.030 1.031 1.032 1.033
56 59 62.5 64 65 69 75 78 81 83 76 81 84 88 89 80 84 86 88 89
30 29 27 25 25 25 23 22 21 20 24 22 20 17 16 22 21 20 18 17
1
2
2.5
2.75
ductility 11 9
6.5 6.3 6.25 6.0 5.75 5.5 5.0 5.0 5.3 5.0 4.5 3.75 3.5 5.1 4.5 4.25 4.0 4.0
point >300 >300 >300 >300 >300 >300 >300 >300 >300 >300 >300 >300 >300 >300 >300 >300 >300 >300 >300 >300
in CS2 (wt % ) 99.76 99.72 99.70 99.68 99.65 99.74 99.72 99.70 99.67 99.64 99.72 99.69 99.64 99.60 99.59 99.68 99.65 99.62 99.60 99.60
heating (wt 5%) 0.08 0.06 0.05 0.03 0.03 0.06 0.06 0.05 0.05 0.04 0.05 0.04 0.04 0.03 0.03 0.04 0.04 0.03 0.03 0.03
Table 5: Data on Heat Treatment with 2.5% AlC13 at Varying Temperatures and Duration temp PC)
duration of heat treatment (h)
specific eravitv
softening Doint ("C)
penetration (1/10 mm)
ductility (cm)
flash Doint ("C)
matter soluble in CS2 (wt % )
loss on heating (wt % )
275
1.0 1.5 2.0 2.5 3.0 1.0 1.5 2.0 2.5 3.0 1.0 1.5 2.0 2.5 3.0
1.025 1.026 1.028 1.030 1.032 1.026 1.028 1.029 1.034 1.036 1.025 1.026 1.026 1.028 1.028
78 80 83 86 87 76 83 86 89.5 90 75 79 81 83.5 84.5
25 23 21 19 18 25 22 18 15 14 29 27 26 25 25
6.0 5.75 5.5 5.0 4.0 5.4 4.75 4.25 3.5 3.0 6.0 5.8 5.75 5.0 5.0
>300 >300 >300 >300
99.74 99.72 99.70 99.64 99.61 99.70 99.67 99.65 99.59 99.56 99.72 99.70 99.67 99.64 99.63
0.06 0.06 0.05 0.04 0.04 0.05 0.04 0.03 0.02 0.02 0.06 0.06 0.05 0.05 0.05
225
200
bring about Friedel-Crafts arylation (Sholl reaction).'* Intramolecular Sholl reaction can also take place. Table 4 gives the effect of different percentages (ranging from 1to 2.75%) of A1C13a t 250 "C on the heat treatment of the vacuum residue of sludge for periods ranging from 1 to 3 h (Figure 1). The results show considerable improvement in the softening point-penetration relationship upto a catalyst level of 2.5% and duration of 2.5 h. With higher percentages, the improvement is not significant. Table 5 reports the data obtained when the heat treatment was carried out a t other different temperatures ranging from 200 to 275 "C for varying periods keeping the catalyst level at 2.5% (Figure 2). The results show that a high temperature of 275 "C as well as a lower temperature of 200 "C did not give a better softening point(12)March, Jerry Advanced Organic Chemistry-Reactions, Mechanisms and Structure; International Student edition; McGraw-Hill Kogakusha Ltd. Tokyo, 1968; Vol. 11, p 412.
>300
>300 >300 >300 >300 >300
>300 >300 >300
>300 >300
penetration relationship. The best result was obtained at the temperature of 225 "C and duration of 2.5 h. This can be taken to be the optimum conditions for preparing grades of industrial bitumen of lower penetration and higher softening point like 90115. The results show that only four of the 10 different grades of industrial bitumen can be prepared by the methods used in the present study. Heat treatment of the vacuum residue a t 250 "C for 2.5 h with 1% AlC13 is sufficient for preparing the 65/25 grade. The 75/30 and 85/25 grades can be obtained by heat treatment at 200 "C with 2.5% AlC13 for 1 and 3 h, respectively. Similarly, 90115 grade can be prepared by heat treatment at 225 "C for 2.5 h with 2.5 % AlC13. For the remaining grades, the softeningpointpenetration specifications were found difficult to meet by the methods used in this study, probably due to the low asphaltene content in the vacuum residue of sludge.
