Energy & Fuels 2006, 20, 1647-1655
1647
Delayed Coking of Decant Oil and Coal in a Laboratory-Scale Coking Unit O ¨ mer Gu¨l, Leslie R. Rudnick,* and Harold H. Schobert The Energy Institute, C205 Coal Utilization Laboratory, The PennsylVania State UniVersity Park, PennsylVania 16802 ReceiVed NoVember 18, 2005. ReVised Manuscript ReceiVed April 21, 2006
In this paper, we describe the development of a laboratory-scale delayed coker and present results of an investigation on the recovered liquid from the coking of decant oil and decant oil/coal mixtures. Using quantitative gas chromatography/mass spectroscopy (GC/MS) and 1H and 13C NMR, a study was made of the chemical composition of the distillate liquids isolated from the overheads collected during the coking and co-coking process. 1H and 13C NMR analyses of combined liquids from coking and co-coking did not show any substantial differences. These NMR results of coking and co-coking liquids agree with those of GC/MS. In these studies, it was observed that co-coking with coal resulted in a decrease in the paraffins contents of the liquid. The percentage of cycloparaffins, indenes, naphthalenes, and tetralins did not change significantly. In contrast, alkyl benzenes and polycyclic aromatic hydrocarbons in the distillate were higher in the co-coking experiments which may have resulted from the distillation of thermally cracked coal macromolecules and the contribution of these molecules to the overall liquid composition.
1. Introduction High-performance jet aircraft flying at high Mach speed use fuel as a primary coolant for onboard heat sources such as the engine lubrication oil, hydraulic fluid, environmental control system, avionics and electrical systems, and the airframe.1 High speeds that result in higher air stagnation temperatures prevent the use of air as a coolant and increase the aircraft cooling requirements. As a result, the fuel has been used as a heat sink in some parts of the aircraft. Under such conditions, a conventional jet fuel (JP-8) could be stressed to temperatures above its thermal stability. This therefore leads to fuel degradation, followed by gas formation and solid deposition in the fuel lines and burner nozzles.2,3 Due to this problem, attempts have been made to produce a fuel that would not degrade at operating temperatures as high as 900 °F. Results show that the most important factor controlling the high-temperature thermal stability of jet fuels is their chemical composition.4 Gas and deposit formation follows a dual mechanism of liquid-phase auto-oxidation at low temperatures (e260 °C) and gas-phase or supercritical pyrolysis at high temperatures (g400 °C), plus a combination of these two in the transition region (260-400 °C).5 Hydroaromatics and cycloaliphatics derived from coal, such as tetralin and decalin, have been reported to stabilize freeradical processes and have been used to improve various thermal * To whom correspondence should be addressed.
[email protected]. (1) Edwards, T. USAF supercritical hydrocarbon fuels interests. Presented at the 31st Aerospace Sciences Meeting and Exhibit, Reno, NV, 1993; AIAA 93-0807, pp 1-11. (2) Song, C.; Eser, S.; Schobert, H. H.; Hatcher, P. G. Energy Fuels 1993, 7, 234-243. (3) Heneghan, S. P.; Zabarnick, S.; Ballal, D. R.; Harrison, W. E., III J. Energy Res. Technol. 1996, 118, 170-179. (4) Gu¨l, O ¨ .; Rudnick, L. R.; Schobert H. H. The Effect of Chemical Composition of Coal-Based Jet Fuels on the Deposit Tendency and Morphology, manuscript in preparation. (5) Hazlett, R. N. Free Radical Reactions to Fuel Research. In Frontiers of Free Radical Chemistry; Pryor, W. A., Ed.; Academic Press: New York, 1980; pp 195-223.
processes.6-10 The application of these and other compounds has been shown to significantly stabilize free radicals that form during the pyrolytic degradation of the fuel and results in the enhancement of the thermal stability of jet fuel.11 Therefore, materials such as coal-derived liquids that are rich in aromatic compounds, like naphthalene and methylnaphthalenes, can be converted into hydroaromatic and cycloaliphatic compounds by hydrotreatment and saturation processes. Coal-derived liquids are thus ideal candidates to be upgraded into thermally stable jet fuel. The coking of hydrocarbon fluids is of wide practical importance within the petroleum industry worldwide. Furthermore, the industries that consume coke (fuel and electrodegrade) realize the critical nature of the composition and morphology of coke and its availability. The petroleum industry has an extensive investment in coking in the United States. In correlations of feedstock properties with coking propensity under hydroprocessing conditions, it was reported that high Conradson carbon residue and pentane insolubles indicated severe coking propensity.12 (6) Rudnick, L. R.; Whitehurst, D. D. The effect of Solvent Composition of the Liquefaction Behavior of Western Sub-Bituminous Coal. Presented at the EPRI Contractor’s Meeting, Palo Alto, CA, May 7-8, 1980. (7) Derbyshire, F. J.; Odoerfer, G. A.; Rudnick, L. R.; Varghese, P.; Whitehurst, D. D. Fundamental studies in the conversion of coals to fuels of increased hydrogen content. Volume 1: The chemistry and mechanics of coal conversion to clean fuel. Volume 2: Appendixes. EPRI Report AP2117 under Research Project 1655-1, EPRI: Pleasant Hill, CA, November 1981; Vols 1 and 2. (8) Whitehurst, D. D.; Buttrill, S. E., Jr.; Derbyshire, F. J.; Farcasiu, M.; Odoerfer, G. A.; Rudnick, L. R. Fuel 1982, 61, 994-1005. (9) Rudnick, L. R.; Whitehurst, D. D. New Approaches in Coal Chemistry; Blaustein, B. D., Bockrath, B. C., Friedman, S., Eds.; ACS symposium series 169; American Chemical Society: Washington, D.C., 1981; pp 153-171. (10) Rudnick, L. R.; Whitehurst D. D.; Derbyshire, F. J. Solvent Compositional Changes During SCT Coal Conversion. Presented at the AIChE Meeting, Orlando, FL, February 28-March 3, 1982. (11) Coleman, M. M.; Sobkowiak, M.; Fearnley, S. P.; Song, C. ACS DiV. Pet. Chem. 1998, 43 (3), 353-356.
10.1021/ef0503820 CCC: $33.50 © 2006 American Chemical Society Published on Web 05/18/2006
1648 Energy & Fuels, Vol. 20, No. 4, 2006
Delayed coking is characterized by long-residence-time heating of bottoms residues from crude oil fractionators, thermal cracking tars, decant oil from catalytic cracking, and aromatic tars from ethylene manufacture. These materials are fed to the coke drum after passing through a heater unit. Fractionation recovers liquids produced by dealkylation and cracking reactions. The fact that coke quality varies with the chemical composition of the precursor feedstock creates a significant incentive to examine the process of coking and how it relates to the composition of the feedstock.13,14 Delayed coking is a process practiced at many refineries in the US and globally. A variety of processes and processing conditions have been reported in the patent and technical literature.15-25 The process of co-coking involves the simultaneous cocarbonization of coal and petroleum-derived liquids under delayed coking conditions. In our studies, co-coking is used to produce coal-based liquids that would ultimately be converted into prototype thermally stable jet fuel, while relying on domestic fossil fuel resources. In addition, the co-coking process produces a solid “coke” byproduct. Not only should the coke have a high market value, but it also needs high-tonnage markets. These potential markets would be anodes for aluminum smelting, graphite for arc furnace electrodes, metallurgical coke, activated carbon, or carbon black. The fact that the chemical composition of a recycle solvent or reacting hydrocarbon liquid controls the selectivity and overall behavior of coal upgrading liquefaction processes has been known for many years.26,27 In these processes, the solvent governs product selectivity by controlling the free radical chemistry involved in these processes.28 These materials, called hydrogen donors, such as tetralins and decalins, work by stabilizing the free radicals produced during thermal treatment. By capping the free radicals, retrogressive reactions are minimized. The presence and concentration of free radicals produced during thermal treatment of coal have been previously (12) Kriz, J. F. Prepr. Pap.sAm. Chem Soc., DiV. Fuel Chem. 1989, 34 (3), 405-408. (13) Rudnick L. R.; Galya, L. G. Energy Fuels 1991, 5, 733-738. (14) Rudnick L. R.; Galya, L. G. A Study of Structural and Chemical Changes of Liquid Products During Coking of Petroleum Residua. Presented at the 21st Biennial Conference on Carbon, Buffalo, NY, June 13-18, 1993; pp 338-339. (15) Nirell, K. G. Delayed coking process and method of formulating delayed coking feed charge. U.S. Patent 6,048,448, 2000. (16) Goyal, S. K.; Faagau, G. S. Coking process with decant oil addition to reduce coke yield. U.S. Patent 4,832,823, 1989. (17) Varghese, P. Coking of coal with petroleum residua. U.S. Patent 4,427,532, 1984. (18) Wynne, F. E., Jr.; Lopez, J.; Zaborowsky, E. J. Coke from coal and petroleum. U.S. Patent 4,259,178, 1981. (19) Li, K. W. Manufacture of petroleum coke with fines recycling. U.S. Patent 4,082,650, 1978. (20) Chen, N. Y.; Walsh, D. E. Delayed coking process. U.S. Patent 4,302,324, 1981. (21) York, E. D.; Rustam, K. F.; Hall, R. D. Delayed coking and dedusting process. U.S. Patent 4,421,629, 1983. (22) Allan, D. E. Delayed coking process with split fresh feed. U.S. Patent 4,492,625, 1985. (23) Calderon, J. L.; Betancourt, H. Process and facility for making coke suitable for metallurgical purposes. U.S. Patent 4,551,232, 1985. (24) Becraft, L. G. Method for producing isotropic coke. U.S. Patent 5,174,891, 1992. (25) Rudnick, L. R. Process for upgrading heavy petroleum feed stock. U.S. Patent 4,642,175, 1987. (26) Ruberto, R. G.; Cronauer, D. C.; Jewell, D. M.; Seshadri, K. S. Fuel 1977, 56 (1), 25-32. (27) Curran, G. P.; Struck, R. T.; Gorin, E. Ind. Eng. Chem. Process Des. DeV. 1967, 6, 166-173. (28) Whitehurst, D. D.; Farcasiu, M.; Mitchell, T. O.; Dickert, J. J., Jr. Nature and origin of asphaltenes in processed coals. EPRI Report AF-1298 Final Report Under Project RP-410, EPRI: Pleasant Hill, CA, December 1979.
Gu¨l et al.
studied by matrix-isolation and in situ using electron spin resonance spectroscopy.29-31 Further studies were reported on free radicals formed and terminated during the process of coking.32,33 Co-coking at Penn State grew from work performed during coal liquefaction studies.34 Development continued with the need to produce coal-based liquids needed for the jet fuels program. Direct liquefaction was not likely to be economically viable soon, and commercialization of direct liquefaction would require construction of a “greenfield” plant, whereas co-coking would require only retrofitting of existing refinery infrastructure. Therefore, it was decided to investigate whether it would be possible to introduce coal or coal-derived liquids (coal tar) into existing petroleum refinery operations. This would ensure that both the desired infrastructure would be in place and that the US would be totally self-sufficient when coal-based jet fuels are to be brought into full-scale production. One option under consideration was the introduction of coal into the delayed coking process. Coal would be fed in slurry form with a petroleum resid into a coker where the volatile components of the coal and resid would be vaporized and subsequently condensed. The resulting coal-based liquid would then go on to be refined and converted to a thermally stable jet fuel, other transportation fuels, and specialty chemicals and feedstocks. The solid coke product would also be utilized either as a feedstock for specialty carbon products (anodes or electrodes) or as fuel. Earlier work in these laboratories on the co-coking process concentrated on small-scale batch experiments.35-38 Reactions in sealed tubing reactors did not reproduce the delayed coking process closely enough. During delayed coking, volatiles are evolved and vented from the reactor and collected. In a sealed system, the volatiles were trapped and thus subject to further decomposition reactions. This led to a decrease in liquid yield and a subsequent increase in solid product. Furthermore, the quantity of products produced was small (initial feedstock charge was approximately 10 g), thus limiting detailed and comprehensive characterization and further processing of the products. For these reasons, a small pilot-scale delayed coker unit was constructed. (29) Rudnick, L. R.; Tueting, D. R. Fuel 1984, 63 (2), 153-157. (30) Rudnick, L. R.; Tueting, D. R.; Sinclair, J. L. ESR Investigations Related to Carbonization Tendency. In Extended Abstracts of the 17th Biennial Conference on Carbon, June 16-21, 1985; American Carbon Society, Washington, D.C., 1985; pp 403-404. (31) Lewis, I. C.; Singer, L. S. Chemistry and Physics of Carbon; Walker, P. L., Jr., Thrower, P. A., Eds.; Marcel Dekker: New York, 1981; Vol 17, pp 1-88. (32) Rudnick, L. R.; Sinclair, J. L. In Symposium on Magnetic Resonance of HeaVy Ends, Proceedings of the 189th ACS National Meeting, Division of Petroleum, April 28-May 3, 1985; American Chemical Society, Washington, D.C., 1985; Vol. 30, No. 2, pp 237-246. (33) Rudnick, L. R.; Sinclair, J. L. Variable Temperature ESR Studies of Commercial Coker Feeds. Presented at the 190th ACS National Meeting, Division of Petroleum, September 8-13, 1985; Vol. 30, No. 4, p 720. (34) Martin S. C.; Tomic J.; Schobert H. H. Coal-Resid Co-Processing by Simulated Delayed Coking. Proc. 9th Int. Conf. Coal Sci. 1997, 3, 15571560. (35) Badger, M. W.; Fickinger, A. E.; Martin, S. C.; Mitchell, G. D.; Schobert, H. H. Novel Co-Processing Applications for Production of CoalDerived Jet Fuels. In AIE 8th Australian Coal Science Conference Preprints, Sydney, Australia, December 1998; p 245. (36) Fickinger, A. E.; Badger, M. W.; Mitchell, G. D.; Schobert, H. H. ACS DiV. Fuel Chem. 1999, 44 (1), 106-109. (37) Fickinger, A. E.; Badger, M. W.; Mitchell, G. D.; Schobert, H. H. ACS DiV. Fuel Chem. 1999, 44 (3), 684-687. (38) Fickinger, A. E.; Badger, M. W.; Mitchell, G. D.; Schobert, H. H. ACS DiV. Fuel Chem. 2000, 45 (2), 299-303. (39) Rodriguez, J.; Tierney, J. W.; Wender, I. Fuel 1994, 73 (12), 18701875.
Delayed Coking in a Laboratory-Scale Coking Unit
Energy & Fuels, Vol. 20, No. 4, 2006 1649
Table 1. Proximate and Ultimate Analysis of the Feeds Used in This Study EI-106 Analysis coal ash (%) volatile matter (%) fixed carbon (%) carbon (%) hydrogen (%) nitrogen (%) sulfur (%) oxygen (by diff) (%)
decant oil
Proximate Analysisa 8.1 27.3 64.6 Ultimate Analysisa 80.9 4.6 1.3 nd nd
0.0
89.7 9.3 0.2 0.8
Fluidity Datab fluid temperature range 88 maximum fluidity (ddpm) 7002 softening temperature (°C) 397 temperature at maximum (°C) 446 Organic Petrography: Volume % Mineral Matter Free total vitrinte (%) 86.5 total liptinite (%) 1.4 total inertinite (%) 12.1 a Values are reported on a dry basis. b Determined using a Geisler plastometer.
This study addresses the liquid products isolated from the process of delayed coking. The liquid products have been distilled into typical refinery cuts and characterized. A description of solid products from these studies will be reported in a subsequent publication. 2. Experimental Section 2.1. Materials. The petroleum-based decant oil used in this study represents a typical decant oil with low sulfur content. The coal used in this study was a Powellton Eagle (high volatile A bituminous) coal. Proximate and ultimate analyses for these feedstocks are shown in Table 1. 2.2. Apparatus. The delayed coking unit was designed after a unit developed at PARC Technical Services, Harmarville, PA. The pilot-scale laboratory coker (PSLC) consisted of a 7.5 cm i.d. × 102.5 cm cylindrical reactor unit having an internal volume of approximately 4.5 L (Figures 1 and 2).
The PSLC at Penn State was modified from the original design to facilitate the collection of liquid products and the isolation of solid from the coke drum. While this paper deals with liquid products isolated during co-coking, the modifications to the PSLC affecting both liquid and solid products are mentioned here to completely describe the reaction system. The first approach to mitigate the difficult physical effort needed to remove coke from the reactor was to machine a small taper into the coking reactor wall from the bottom edge where the preheater enters over about 30.48 cm (12 in.). This gradual conical shape, 0.318 cm (1/8 in.) taper over 30.48 cm (12 in.), ensures that once the coke artifact is freed from its adhesion to the reactor walls, it would be contained in a space larger than the artifact itself and, therefore, be easily pushed from the reactor. To facilitate this, the second approach was to apply a commercially available, high-temperature stable, ceramic mold release to the interior reactor walls to the thermocouple assembly and to the bottom end cap of the reactor prior to assembly of the coking reactor for each experiment. This sufficiently reduced the friction and adhesion of coke to these reactor parts and allowed for the smooth removal of a coke artifact from the reactor. The system pressure, temperature, and flow rates are monitored by a number of computer-controlled devices, and data from these devices is recorded throughout the experiment. In this study, the unit was run without continuous steam injection and with a steam strip period after the completion of the hydrocarbon feedstock introduction to the coke drum, so that the effect of steam stripping on product yields could be measured. Steam injection did not change the slurry feed rate. The slurry feed rate in these experiments was continuous and constant and was measured gravimetrically with time. The preheater was a 2.5 cm o.d. × 51 cm stainless steel tube fitted directly to the bottom of the reactor (Figure 1). This was fed by a 0.953 cm (3/8 in.) o.d. feed line that was outside the furnace and was heated to 120 °C using heating tape. This design configuration allowed for essentially trouble-free pumping of the coal-decant oil slurry over a wide temperature range. The temperature gradient through this 51 cm preheater was on the order of 200 °C, with an outlet temperature of 420-460 °C. This was connected to a 0.635 cm (1/4 in.) o.d. line that carried feedstocks from the feed pump. 2.3. Reaction Procedures. The following operating conditions were used: coke drum inlet temperature 465 °C, coke drum pressure 25 psig, oil or slurry feed rate 16.7 g/min, and feed introduction to
Figure 1. Coking reactor schematic: (1) heated feedstock tank; (2) feedstock pump; (3) balance; (4) metering pump; (5) superpreheater; (6) preheater; (7) coker drum; (8) thermocouple well; (9) DP cell; (10) back pressure regulator; (11) receiver tank; (12) mass flow meter.
1650 Energy & Fuels, Vol. 20, No. 4, 2006
Gu¨l et al.
Figure 2. Photos of the coking reactor, a coke artifact extraction from the coker, and an isolated piece of coke.
the coker 360 min. Variation of the coke drum inlet temperature and coke drum pressure will be the subject of future studies. In the steam stripping experiments, water was added at a steam feed rate of 1 vol % water relative to the feed rate. Steam was added over a period of 60 min. At the conclusion of each experiment, the coke drum was maintained at temperature for an additional 360 min to ensure carbonization of nonvolatile components. The pilot-scale laboratory coker and associated equipment is shown in Figure 1. In the co-coking experiments, coal was fed in slurry form with the decant oil into the coker where the volatile components of the coal and oil were vaporized and subsequently condensed. This pilotscale vented reactor system was used for coking or co-coking experiments. The vented reactor system allowed for flash vaporization of the volatiles and subsequent carbonization of the heavy petroleum fraction and coal. In the delayed coking process, feedstock is pumped (16.7 g/min) into the coker drum where reactions between the coke and the liquid lead to the formation of light desirable liquids and carbonaceous solid. The feed is initially charged to a heated feedstock vessel that was continuously mixed throughout the coking or co-coking experiment to achieve and maintain homogeneity. In these experiments, the temperature was kept at 66 °C. The feedstock was transferred to a smaller feed vessel (700 mL capacity) by using a rough pump. The small feed vessel was placed on a balance for monitoring the feeding rate and kept at 66 °C. The feed in the small feed vessel was transferred to the coking unit using a metering pump. The feed was incrementally heated along the feed line to the preheater. Feed was heated in the lines prior to the preheater to about 180 °C and, then, to about 420 °C in the preheater. Heated feedstock from the preheater was fed to the coker drum. Thermocouples attached at different positions along the coke drum were used to measure and to control the temperature during the experiment. Light hydrocarbons that vaporized exit from the top of coker drum and pass through a series of condensers. Gases went through a mass flow meter and were either collected or sent to vent. Process control variables, such as temperature, pressure, and feed rate, will be reported in future studies. In the experiments reported here, the liquid products from the reactions were passed through a series of condensers and valves that facilitate the isolation of liquid product as a function of reaction time. During each run, liquid samples were taken at selected intervals. At the conclusion of the experiment, the mass of the liquid condensate was measured. In addition, the carbonaceous solid was removed from the coke drum and weighed.
Table 2. Conditions and Product Distributions for Coking and Co-coking Experiments A
B
DO
DO
C
D
feedstock (h) steam strip at 500 °C (h) hold at 500 °C (h) feed rate (g/min) preheater inlet (°C) preheater outlet (°C) coke drum inlet (°C) coke drum lower/middle (°C) coke drum top (°C) material fed to reactor (g)
6 0 6 16.7 181 417 446 493
6 0 6 16.7 214 464 479 493
6 0 6 16.7 188 419 474 481
6 1 5 16.7 179 457 n/a 486
458 6028
471 5701
466 6054
471 6012
Coke liquid gas (by difference) coke + liquid product liquid/coke (ratio) % coke % liquid product % gas (by difference)
860 4800 368 5660 5.58 14.27 79.63 6.10
Product 794 4418 489 5212 5.56 13.93 77.50 8.57
1917 3989 148 5906 2.08 31.67 65.89 2.44
1770 3838 404 5608 2.17 29.44 63.84 6.72
DO/coal (4:1) DO/coal (4:1)/SS
2.4. Analytical Procedures. Gas chromatography/mass spectroscopy (GC/MS) analysis, using a Shimadzu QP5000 spectrometer, was performed on liquid samples to study whether the composition changed as a function of time. The temperature was held at 40 °C for 5 min, programmed from 40 to 270 °C at 6 °C/ min, and then held at 270 °C for an additional 20 min. An XTI-5 ((Restek; 30 m × 0.25 mm × 0.25 µm) column was used for the GC/MS analyses. Collected liquid samples were then combined and were used for simulated distillation and 1H and 13C NMR analysis. Samples were analyzed on a Bruker AMX 360 NMR operating at 9.4 T. Samples were dissolved in a 1/1 volume ratio in CDCl3 containing 1 vol % of tetramethylsilane (TMS) as a standard. A pulse width of 5 µs and a pulse delay of 5 s for 1H with a 90° tip angle and a 5 µs pulse width and a pulse delay of 45 s for 13C with a 70° tip angle were used to ensure quantitative results. In 13C analyses, 20 mg Cr(AcAc)3 was used for 2 mL of an overhead liquid/CDCl3 mixture. Regions of the spectra were integrated and peaks were assigned based on literature chemical shift values for
Delayed Coking in a Laboratory-Scale Coking Unit
Energy & Fuels, Vol. 20, No. 4, 2006 1651
Table 3. Decant Oil Coking Liquid Composition as a Function of Reaction Time by GC/MS name
60 min
120 min
180 min
240 min
300 min
360 min
mean
std dev
1-octene 1-nonene 1-decene naphthalene decane 2-methylnaphthalene undecane (isomer) undecane (isomer) tridecane 2-methylphenanthrene 1-methylanthracene 9,10-dimethylphenanthrene 2,5-dimethylphenanthrene hexadecane pyrene trimethylphenanthrene (isomer) trimethylphenanthrene (isomer) 6,9-dimethyltetradecane 2-methylpyrene 3,4,5,6-tetramethylphenanthrene 4-methylpyrene 1-methylpyrene octadecane 2,4,5,7-tetramethylphenanthrene 5,12-dihydronaphthacene o-terphenyl 1,3-dimethylpyrene benz[a]anthracene triphenylene eicosane 5-methylterphenyl 6-methylchrysene
1.00 1.21 1.46 2.26 3.03 1.29 9.36 6.92 1.10 1.27 1.13 1.58 1.03 1.00 1.68 1.11 1.09 1.35 2.54 1.19 3.82 1.95 1.43 1.15 1.73 2.24 3.09 1.28 1.65 2.37 1.13 2.42
0.68 0.81 0.89 1.66 1.84 0.86 5.05 4.21 0.78 1.02 0.79 1.16 0.74 0.93 1.70 1.35 0.92 1.15 2.13 1.09 2.93 1.44 1.17 1.13 1.50 1.57 2.45 1.29 1.56 0.75 0.64 2.29
0.73 0.92 1.00 2.35 3.07 1.29 6.26 4.77 1.10 1.44 1.18 1.60 0.93 1.04 2.02 1.49 0.99 1.32 2.50 0.96 3.27 1.31 1.36 0.85 1.38 1.49 2.15 1.07 1.70 0.50 0.96 2.32
0.70 0.87 0.95 2.22 2.29 1.20 6.15 4.94 1.06 1.38 1.14 1.45 0.81 0.94 2.00 0.81 0.85 1.20 2.42 0.83 1.48 1.25 1.28 0.89 1.65 1.81 2.72 1.52 1.85 0.75 0.87 2.75
0.48 0.67 0.73 1.67 1.89 1.02 5.27 3.80 0.86 1.10 0.91 1.28 0.73 0.86 1.86 0.83 0.79 1.21 2.46 1.03 2.92 1.40 nd 1.31 1.70 2.49 2.61 1.50 1.78 0.81 0.99 2.61
0.59 0.80 0.83 2.03 2.05 1.07 5.83 4.42 0.89 1.24 1.09 1.33 0.82 0.95 1.92 0.96 0.85 1.26 2.48 1.06 2.69 1.43 1.42 0.61 1.77 2.68 2.91 1.53 1.89 1.06 1.05 2.63
0.70 0.88 0.98 2.03 2.36 1.12 6.32 4.84 0.97 1.24 1.04 1.40 0.84 0.95 1.86 1.09 0.92 1.25 2.42 1.03 2.85 1.46 1.33 0.99 1.62 2.05 2.66 1.37 1.74 1.04 0.94 2.50
0.16 0.17 0.23 0.28 0.51 0.16 1.43 1.00 0.13 0.15 0.14 0.16 0.11 0.06 0.13 0.26 0.10 0.07 0.14 0.11 0.71 0.23 0.50 0.23 0.14 0.45 0.30 0.17 0.11 0.62 0.16 0.17
Table 4. Decant Oil-Coal (20%) Co-coking Liquid Composition as a Function of Reaction Time by GC/MS name
60 min
120 min
180 min
240 min
300 min
360 min
mean
std dev
1-octene 1-nonene naphthalene decane 2-methylnaphthalene undecane (isomer) undecane (isomer) tridecane 2-methylphenanthrene 1-methylanthracene 3,6-dimethylphenanthrene 9,10-dimethylphenanthrene 2,5-dimethylphenanthrene hexadecane pyrene trimethylphenanthrene (isomer) trimethylphenanthrene (isomer) 6,9-dimethyltetradecane 2-methylpyrene 3,4,5,6-tetramethylphenanthrene 4-methylpyrene 1-methylpyrene octadecane 2,4,5,7-tetramethylphenanthrene 5,12-dihydronaphthacene terphenyl (isomer) terphenyl (isomer) 1,3-dimethylpyrene triphenylene eicosane (isomer) 6-methylchrysene 4-methylchrysene eicosane (isomer)
1.23 1.48 2.65 3.18 1.35 9.65 7.19 1.13 1.35 1.02 1.07 1.72 1.09 1.10 1.45 1.68 1.06 1.40 2.30 0.97 3.20 1.72 1.54 1.07 1.39 2.19 1.45 2.51 1.29 2.47 2.05 3.11 1.65
1.20 1.39 2.40 2.87 1.31 8.19 5.91 1.13 1.32 0.97 1.02 1.53 0.97 1.10 1.55 1.52 0.84 1.40 2.32 0.84 2.73 1.55 1.45 0.84 1.45 1.88 1.62 2.99 1.44 2.59 0.98 2.99 1.89
0.71 0.93 2.07 2.02 1.02 5.28 4.01 0.90 1.03 0.72 0.80 1.28 0.77 0.90 1.66 1.31 0.82 1.26 2.21 0.53 2.50 1.48 1.35 1.01 1.50 2.26 1.60 2.80 1.70 2.41 2.56 3.10 1.92
0.88 1.01 2.65 2.42 1.43 6.63 4.99 1.31 1.59 1.21 1.19 1.75 1.11 1.31 2.02 1.76 1.12 1.68 2.85 1.06 3.14 1.74 1.86 1.19 1.90 2.37 1.97 3.51 1.87 3.60 1.69 3.90 2.53
0.69 0.84 2.31 2.14 1.21 5.54 4.28 1.05 1.41 1.02 0.96 1.44 0.83 1.15 1.59 1.46 1.02 1.53 2.34 0.85 2.46 1.49 1.67 1.12 1.63 2.59 1.55 3.28 2.02 3.18 2.60 3.66 2.29
0.57 0.81 2.02 2.06 1.15 5.37 4.15 1.02 1.38 1.00 0.88 1.29 0.89 1.12 1.83 1.38 1.00 1.54 2.18 0.88 2.33 1.55 1.62 1.05 1.65 1.83 1.73 3.34 1.84 3.51 2.54 3.56 2.29
0.88 1.08 2.35 2.45 1.25 6.78 5.09 1.09 1.35 0.99 0.99 1.50 0.94 1.11 1.68 1.52 0.98 1.47 2.37 0.86 2.73 1.59 1.58 1.05 1.59 2.19 1.65 3.07 1.69 2.96 2.07 3.39 2.10
0.25 0.26 0.25 0.44 0.14 1.63 1.14 0.13 0.17 0.14 0.13 0.19 0.13 0.12 0.19 0.16 0.11 0.13 0.22 0.16 0.34 0.10 0.16 0.11 0.17 0.27 0.16 0.34 0.25 0.49 0.59 0.34 0.30
1H
and 13C.40 Simulated distillation gas chromatography was performed on these samples to determine the boiling point distribution. The simulated distillation measurements were made according to the ASTM 2887 method by using an HP 5890 GC-FID fitted with an MXT-500 simulated distillation column (10 m, 0.53 mm i.d., and 2.65 µm; Restek). The carrier gas flow rate was adjusted to 13 mL/min for sim-dist analysis, and SimDis Expert 6.3 software was used to calculate the percentage of fractions.
(40) Rodriguez J.; Tierney J. W.; Wender I. Fuel 1994, 73 (12), 18701875.
Finally, the combined distillate liquids from each coking or cocoking experiment were vacuum distilled into refinery cuts corresponding to gasoline, jet fuel, diesel, and fuel oil.
3. Results and Discussion 3.1. Product recovery. The objective of this study was to compare the results of coking of a decant oil and the co-coking of the same decant oil with a coal. In co-coking experiments, the coal was used at 20 wt % and the slurry was continuously
1652 Energy & Fuels, Vol. 20, No. 4, 2006
Gu¨l et al.
Table 5. Decant Oil-Coal (20%) Co-coking Liquid Composition as a Function of Reaction Time by GC/MS (with Steam Stripping) name
60 min
120 min
180 min
240 min
300 min
360 min
mean
std dev
1-octene 1-nonene 1-decene 1-undecene naphthalene decane 2-methylnaphthalene undecane (isomer) undecane (isomer) tridecane 2-methylphenanthrene 1-methylanthracene 9,10-dimethylphenanthrene 2,5-dimethylphenanthrene hexadecane pyrene trimethylphenanthrene (isomer) trimethylphenanthrene (isomer) 6,9-dimethyltetradecane 2-methylpyrene 4-methylpyrene 1-methylpyrene octadecane 5,12-dihydronaphthacene terphenyl (isomer) terphenyl (isomer) 1,3-dimethylpyrene benz[a]anthracene triphenylene eicosane 6-methylchrysene 4-methylchrysene
1.42 1.60 1.66 1.21 2.73 5.86 1.39 19.30 13.86 0.93 1.22 0.85 1.51 0.89 0.90 1.26 1.51 1.00 1.34 1.86 1.26 1.21 1.23 1.35 1.98 1.58 2.72 0.69 0.68 2.24 1.87 2.14
0.88 1.04 1.09 0.65 2.18 2.61 1.22 7.74 5.52 0.90 1.20 0.87 1.21 0.77 1.03 1.39 1.35 0.86 1.34 2.08 2.31 1.41 1.44 1.51 2.47 1.61 2.72 1.47 1.70 2.16 2.55 2.57
1.29 1.49 1.58 0.95 3.23 3.79 1.84 10.66 7.74 1.14 1.72 1.32 1.78 1.04 1.29 1.98 1.64 0.92 1.62 2.81 2.99 1.52 1.73 1.38 1.96 1.22 2.63 1.16 1.67 3.20 2.73 2.94
0.94 1.05 1.13 0.78 2.46 2.82 1.45 8.21 5.94 1.02 1.26 0.92 1.51 0.89 1.10 1.60 1.50 1.06 1.55 2.48 2.52 1.48 1.71 1.77 2.77 1.65 3.30 1.49 1.92 2.66 1.70 3.00
0.94 1.10 1.19 0.91 3.09 3.19 1.58 8.99 6.71 1.20 1.58 1.25 1.78 1.07 1.35 2.02 1.19 1.06 1.72 2.72 2.84 1.57 1.81 1.77 2.04 1.64 3.42 1.25 1.74 3.32 2.36 2.81
0.99 0.79 0.92 0.79 2.89 2.69 1.50 7.68 5.83 1.27 1.67 1.30 1.75 1.05 1.33 1.85 1.60 1.01 1.70 2.47 2.45 1.42 1.77 1.94 2.70 1.85 3.61 1.50 2.00 4.05 3.03 3.35
1.08 1.18 1.26 0.88 2.76 3.49 1.50 10.43 7.60 1.08 1.44 1.09 1.59 0.95 1.17 1.68 1.47 0.99 1.55 2.40 2.40 1.44 1.62 1.62 2.32 1.59 3.07 1.26 1.62 2.94 2.37 2.80
0.20 0.28 0.27 0.18 0.36 1.13 0.19 4.09 2.89 0.14 0.22 0.21 0.21 0.11 0.17 0.29 0.15 0.07 0.16 0.34 0.56 0.11 0.21 0.22 0.34 0.19 0.39 0.29 0.44 0.66 0.46 0.38
Table 6. Boiling Point Distributions by Simulated Distillation Gas Chromatography
feeds
IBP-180 °C IBP-356 °F gasoline
180-270 °C 356-518 °F jet fuel
270-332 °C 518-630 °F diesel
332-FBP °C 630-FBP °F fuel oil
DO DO/coal DO/coal/SS mean
5.57 4.99 5.75 5.44
9.83 9.75 10.62 10.07
10.25 9.85 10.28 10.13
74.34 75.41 73.35 74.37
Table 7. Product Distributions by Weight from Vacuum Distillation
feeds
IBP-180 °C IBP-356 °F gasoline
180-270 °C 356-518 °F jet fuel
270-332 °C 518-630 °F diesel
332-FBP °C 630-FBP °F fuel oil
DO DO/coal DO/coal/SS mean
6.07 6.39 7.35 6.60
11.55 10.41 11.06 11.01
9.34 9.35 10.32 9.67
73.05 73.84 71.27 72.72
heated (66 °C) and stirred to ensure homogeneity of the slurry during introduction to the coking reactor. The properties of the decant oil and coal used in the co-coking experiments are described in Table 1. Table 2 shows the conditions used in each of the coking and co-coking experiments. Using similar conditions for each of the experiments, it can be seen that addition of 20% coal results in approximately twice the carbon yield in the co-coking experiments whether or not steam stripping was used. Gas yields were low in all experiments, and reported values are within experimental error of each other. One of the principal objectives was to determine if the process was reproducible in terms of the yields of green coke and liquids isolated from experiments. The reproducibility of the reaction of decant oil in two separate experiments was shown to be good (Table 2, compare A and B). Liquids are obtained in suitable quantity for detailed chemical characterization, recombination, and distillation into refinery cuts for evaluation.
Table 8. Simulated Distillation Boiling Point Distributions of Coker Distillates BP (°C) % off
DO
DO/coal (20%)
DO/coal (20%)/SS
IBP 5.00 10.00 15.00 20.00 25.00 30.00 35.00 40.00 45.00 50.00 55.00 60.00 65.00 70.00 75.00 80.00 85.00 90.00 95.00 FBP
74.2 157.2 243.9 267.7 291.7 330.4 373.0 395.9 406.5 414.2 421.9 427.6 434.3 440.7 447.2 455.1 463.1 472.7 484.7 502.1 537.6
76.6 171.0 251.0 269.8 297.5 342.7 384.9 399.9 409.7 416.8 424.6 430.8 437.4 444.2 450.7 458.3 467.1 477.3 490.5 510.6 559.1
72.5 153.3 235.6 263.2 284.7 317.4 366.7 394.0 405.6 413.9 421.8 427.7 434.5 441.0 447.6 455.5 463.8 473.5 485.7 503.4 537.9
3.2. Composition of Liquid Product as a Function of Reaction Time. Reported in Tables 3-5 are compounds observed at a 1% or greater amount by GC/MS analysis; the mean and the standard deviation of each monitored component are also calculated. This cut of 1% results in not reporting all members of a homologous series. These GC/MS analyses have been performed to assess compositional changes during the 6 h run time. Table 3-5 show approximately 30 of the most prevalent compounds found in the distillates obtained from the coking and co-coking experiments. These analyses showed that, although there were observed increases or decreases in the area percent of a particular component, the standard deviation was low. More quantitative GC/MS data will be discussed below.
Delayed Coking in a Laboratory-Scale Coking Unit
Energy & Fuels, Vol. 20, No. 4, 2006 1653
Table 9. Group Classification by Quantitative GC/MSa of Major Component Types in Distillates (wt %) classification
DO
DO/20% coal
DO/20% coal/SS
paraffins cycloparaffins alkyl benzenes indenes naphthalenes tetralins PAH
37.30 5.98 12.49 2.40 9.97 1.19 30.67
33.31 6.24 15.70 1.72 9.24 1.17 32.63
33.58 6.38 14.10 2.29 8.06 1.16 34.42
a
Calculated using an external standard.
These data suggest that the light fraction of decant oil was distilled during the course of the decant oil delayed-coking experiment, and that is what contributes to the liquid product. In the case of co-coking with coal, both decant oil light fraction and coal-derived light hydrocarbons were codistilled.
For each coking run, the distillate liquid samples taken during each experiment, as described above (Tables 3-5), were combined. The combined oils were evaluated using simulated distillation GC, and they were also distilled into conventional refinery boiling ranges by vacuum distillation. Simulated distillation analysis of the combined distillate liquids from coking experiments with decant oil and decant oil/coal are shown in Table 6. Distillates include material boiling in the gasoline, jet fuel, diesel, and fuel oil ranges. The percentage of all fractions obtained using simulated distillation GC were almost the same for the coking and co-coking experiments, and their average values from the vacuum distillation were 5.44%, 10.07%, 10.13%, and 74.37% for gasoline, jet fuel, diesel, and fuel oil, respectively (Table 6). Table 7 shows actual distillation yields as a weight percent of the liquid product during the coking or co-coking process. The percentages of all determined fractions
Table 10. Distribution of 1H NMR Signals of Combined Oils from the Coking and Co-coking Experiments assignment
band (ppm)
DO only
DO/coal
DO/coal/SS
CH3 γ and, further, some naphthenic CH and CH2 CH2 β and, further, some β CH3 most CH, CH2 β hydroaromatic R to olefinic CH3 R to aromatic carbons CH, CH2 R to aromatic carbons CH2 bridge (diphenylmethane) olefinic aliphatic (total) single ring aromatic diaromatic and most of tri- and tetraromatic some tri- and tetraromatic rings some tetraromatic rings aromatic (total)
0.5-1.0 1.0-1.7 1.7-1.9 1.9-2.1 2.1-2.4 2.4-3.5 3.5-4.5 4.5-6.0
17.8 39.2 1.9 2.5 8.1 9.2 0.2 0.4 79.3 5.3 14.0 1.3 0.0 20.7
17.2 40.6 2.1 2.6 7.8 9.7 0.3 0.4 80.8 5.2 12.8 1.2 0.0 19.2
17.9 39.7 2.0 2.7 7.9 9.6 0.1 0.4 80.3 4.8 13.4 1.4 0.0 19.7
Table 11. Distribution of
13C
6.0-7.2 7.2-8.3 8.3-8.9 8.9-9.3
NMR Signals of Combined Oils from the Coking and Co-coking Experiments
assignment CH3 γ and further from aromatic ring CH3 in ethyl substituted cyclohexane CH3 γ and further from aromatic ring CH3 R shielded by two adjacent rings or groups CH3 R shielded by one adjacent ring or group some CH3 a hydroaromatic and naphthenic CH2 CH3 R shielded by one adjacent rings or group some CH3 a hydroaromatic and naphthenic CH2 CH3 not shielded by adjacent rings or groups some CH3 a hydroaromatic and naphthenic CH2 CH2 γ and further adjacent to terminal CH3 CH2 β in unsubstituted tetralin structures some CH2 naphthenic CH2 R not shielded CH2 β in propyl and indan groups CH3 β in isopropyl CH2 not adjacent to CH in alkyl groups CH2 adjacent to alkyl CH in some CH2 a and CH2 adjacent to terminal CH3 in alkyl substituents with more than four carbons CH2 in ring joining ethylene groups some CH2 naphthenic some ring joining methylene (32-43 ppm) CH alkyl groups (except isoalkyl) CH naphthenic CH2 alkyl groups adjacent to CH some ring joining methylene (32-43 ppm) aliphatics (total) some olefinic (others spread through aromatic region) protonated aromatic some internal (quaternary) aromatic most internal aromatic methyl substituted aromatic naphthenic substituted aromatic alkyl (other than methyl) substituted aromatic heteroatom (N, O, S) aromatic aromatics (total)
band (ppm)
DO only
DO/coal
DO/coal/SS
11.0-12.5
0.4
0.4
0.6
12.5-15.0
3.2
3.2
3.5
15.0-18.0
1.4
1.4
1.6
18.0-20.5
3.8
3.8
4.2
20.5-22.5
3.8
3.7
4.0
22.5-24.0
3.1
3.1
3.1
24.0-27.5
3.7
3.6
3.9
27.5-37.0
22.6
23.6
22.8
37.0-60.0
8.6
9.3
8.2
108.0-118.0 118.0-129.5
50.7 0.6 30.9
52.1 0.6 28.8
51.8 1.5 30.5
129.5-133.0 133.0-135.0 135.0-138.0 138.0-160.0
6.6 2.5 8.7 0.4
6.6 2.6 9.3 0.4
6.3 2.4 7.5 0.6
49.3
47.9
48.2
1654 Energy & Fuels, Vol. 20, No. 4, 2006
Gu¨l et al.
Table 12. Distribution of 1H NMR Signals as a Function of Time of Delayed Co-coking of Pittsburgh Seam Coal with Decant Oil (4:1 Ratio) Assignment
band (ppm)
60 min
180 min
360 min
mean
std dev
CH3 γ and, further, some naphthenic CH and CH2 CH2 β and, further, some β CH3 most CH, CH2 β hydroaromatic R to olefinic CH3 R to aromatic carbons CH, CH2 R to aromatic carbons CH2 bridge (diphenylmethane) olefinic aliphatic (total) single ring aromatic diaromatic and most of tri- and tetraromatic some tri- and tetraromatic rings some tetraromatic rings aromatic (total)
0.5-1.0
6.2
6.0
6.7
6.3
0.4
1.0-1.7 1.7-1.9 1.9-2.1 2.1-2.4 2.4-3.5 3.5-4.5 4.5-6.0
15.4 1.0 3.3 19.2 13.7 0.4 0.1 59.2 12.0 26.5 1.5 0.7 40.8
13.5 1.4 5.0 20.5 9.9 0.2 0.1 56.5 14.5 27.4 0.9 0.6 43.5
13.0 1.6 6.2 21.2 7.6 0.1 0.4 56.8 15.2 26.6 0.7 0.6 43.2
13.9 1.3 4.8 20.3 10.4 0.3 0.2 57.5 13.9 26.9 1.1 0.7 42.5
1.3 0.3 1.4 1.0 3.1 0.2 0.2 1.5 1.6 0.5 0.4 0.1 1.5
Table 13. Distribution of
13C
6.0-7.2 7.2-8.3 8.3-8.9 8.9-9.3
NMR Signals as a Function of Time of Delayed Co-coking of Pittsburgh Seam Coal with Decant Oil (4:1 Ratio)
Assignment CH3 γ and further from aromatic ring CH3 in ethyl substituted cyclohexane CH3 γ and further from aromatic ring CH3 R shielded by two adjacent rings or groups CH3 R shielded by one adjacent ring or group some CH3 a hydroaromatic and naphthenic CH2 CH3 R shielded by one adjacent rings or group Some CH3 a hydroaromatic and naphthenic CH2 CH3 not shielded by adjacent rings or groups some CH3 a hydroaromatic and naphthenic CH2 CH2 γ and further adjacent to terminal CH3 CH2 β in unsubstituted tetralin structures some CH2 naphthenic CH2 R not shielded CH2 β in propyl and indan groups CH3 β in isopropyl CH2 not adjacent to CH in alkyl groups CH2 adjacent to alkyl CH in some CH2 a and CH2 adjacent to terminal CH3 in alkyl substituents with more than four carbons CH2 in ring joining ethylene groups some CH2 naphthenic some ring joining methylene (32-43 ppm) CH alkyl groups (except isoalkyl) CH naphthenic CH2 alkyl groups adjacent to CH some ring joining methylene (32-43 ppm) aliphatics (total) some olefinic (others spread through aromatic region) protonated aromatic some internal (quaternary) aromatic most internal aromatic methyl substituted aromatic naphthenic substituted aromatic alkyl (other than methyl) substituted aromatic heteroatom (N, O, S) aromatic aromatics (total)
60 min
180 min
360 min
mean
std dev
11.0-12.5
band (ppm)
0.2
0.1
0.2
0.16
0.06
12.5-15.0
1.7
1.5
1.4
1.6
0.1
15.0-18.0
1.5
1.5
1.4
1.5
0.1
18.0-20.5
4.7
4.4
4.4
4.5
0.2
20.5-22.5
4.5
4.3
4.4
4.4
0.1
22.5-24.0
1.1
0.8
0.9
0.9
0.1
24.0-27.5
1.1
1.0
1.1
1.0
0.1
27.5-37.0
7.4
6.5
6.8
6.9
0.4
37.0-60.0
2.3
2.1
2.7
2.4
0.3
108.0-118.0 118.0-129.5
24.4 1.1 47.7
22.3 1.0 48.9
23.3 1.5 50.2
23.3 1.2 48.9
1.1 0.3 1.3
129.5-133.0 133.0-135.0 135.0-138.0 138.0-160.0
10.9 3.9 12.0 0.2
11.1 3.9 12.8 0.1
10.6 3.8 10.5 0.2
10.9 3.9 11.8 0.16
0.3 0.1 1.1 0.06
75.6
77.7
76.7
76.7
1.1
using vacuum distillation were almost the same for the coking and co-coking experiments, and their average values from the vacuum distillation were 6.60%, 11.01%, 9.67%, and 72.72% for gasoline, jet fuel, diesel, and fuel oil, respectively (Table 7). There is excellent agreement between the results obtained by simulated distillation GC and the actual isolated yields of the fractions from the vacuum distillation. The various product fractions are largely independent of whether coal was used. One can conclude that during delayed coking mainly decant oil (DO) distilled and that coal components contributed to the distilled liquid in small amounts. The simulated distillations of the combined liquids from each experiment are compared in Table 8. The liquid isolated from the steam-stripped (SS) co-coking experiment had a lower boiling point distribution than did the non-steam-stripped liquids. The 50% point in the boiling point distribution is virtually the
same for all three experiments. The difference in the DO/coal/ SS occurs only below 35% off. At 35% off and above, the boiling point distributions are all virtually the same. One can conclude that steam stripping removed additional light hydrocarbons that were physically trapped in the carbonaceous solid and increased the concentration of lower-boiling-point (
