-
AbstractAlumina with (8–18 wt.%) carbon black composite (AMAC) supports was prepared as bimodal extrudates, containing 11–20% of total porevolume as macropores (i.e. >1000 A ˚ ). These supports, in spite of containing carbon black and macropores, showed good side crushing strength(0.67–1.19 kg/mm) after pyrolysis in 6% O2/N2. AMAC-catalysts were obtained after impregnating these alumina–carbon black supports with Niand Mo, to obtain 3.5 wt.% NiO and 15 wt.% MoO3. These catalysts were evaluated for about 700 h in the hydroconversion of a Mexican vacuumresidue (538 8C+) at 415 8C, 200 kg/cm2,H2/HC = 6000 ft3/barrel in a pilot plant equipped with a Robinson–Mahoney reactor. In comparison witha commercial bimodal alumina-based catalyst (ComCat), AMAC catalysts showed much fewer sediments and less Conradson carbon formation.Initial HDS in AMAC containing macropores can be as high as 92%, while that in a ComCat is 86%. On average, yields of naphtha and kerosenewere 2.6 and 1.34 times higher with AMAC catalysts than those with ComCat, while diesel yields were similar.# 2005 Elsevier B.V. All rights reserved.Keywords: Hydroconversion; Hydrodesulfurization; Hydrodenitrogenation; Vacuum residue; H-oil; Carbon-black; Alumina support; Macropores 1. IntroductionPetroleum residues are the heavy fraction remaining afterdistilling petroleumcrudes at atmospheric pressure (atmosphericresidue) or at reduced pressure of 25–100 mmHg (vacuumresidue). Residues have high molecular weight ( 1000 amunumber average for vacuum residue) and contain polynucleararomatics,34167
- 上一篇：高压泵反渗透设备列车英文文献和中文翻译
- 下一篇：工作流技术英文文献和中文翻译
also termed asphaltenes. Because of their highthermal stability, asphaltenes having 3–4 rings provide thegreatest limitation to the conversion of residue. In addition, thehigh concentrations of heteroatoms (sulfur, nitrogen, vanadium,and nickel) in petroleum residue act along with asphaltenes topoison catalysts. No matter which type of process is used, asubstantial fraction of residue molecules can be cracked off asfragments to become liquids in the transportation fuel andvacuum gasoil boiling ranges. However, one should not try tooverly convert residues because asphaltene content can force theselectivity to go to the thermodynamically favored, but lowervalued, products: coke and/or sediments and/or hydrocarbon gases.One of the processes designed to convert heavy oil residueto lighter fractions is the so-called H-OilTM, which typicallyoperates at 410–420 8C, 120 kg/cm2, LHSV: 0.5 h 1,H2/HC:3500 ft3/barrel, and uses an ebullated bed reactor.Maya crude oil is classified as heavy oil, and it accounts formore than 50% of the total crude oil production in Mexico. Interms of environmental problems, catalytic hydrodesulfurization(HDS) and hydrocracking (HC) technologies are important toproduce low sulfur fuel oils by upgrading residues fromMexicancrude oil which contain not only high concentrations of sulfurcompounds (around 4.5 wt.%), and catalyst-deactivating vana-dium compounds (around 400 ppm), but also asphaltenes. Atpresent,Mexican refineries operate with crude oil blends, that is,55–60% Isthmus (e.g., the Mexican light blend), and 45–40%Maya (e.g., theMexican heavy blend) crude oils; but since heavyoils tend to become relatively abundant, an unavoidable situationfor the Mexican oil industry in the near future will be to refineMaya crude oil richer blends. Two H-OilTM plants, which havebeen operating for more than 30 years in the Salamanca, Gto.refinery (18,500 BPD designed capacity), and more recently inthe Tula, Hgo. refinery (50,000 BPD designed capacity),
transform Mexican heavy vacuum residue into more valuableoil fractions, such as gasoil and diesel, removing great amounts of polluting sulfur compounds. Nonetheless, the H-OilTMprocess is frequently hampered by the formation of high-molecular weight compounds, called sediments, which causemany downstream operational problems, especially by blockingvalves, pumps, flow-control devices, and hot and cold separators.Sediments are polynuclear aromatics which result fromcomplexcracking, recombination, and nucleation reactions amongasphaltenes [1]. H-OilTM plants cannot operate at highconversion levels since sediment yields rise, particularly whenusing heavy residue from heavy crude oils [2]. Thus, sedimentyields above 1 wt.% render H-OilTM plant operation anywherefrom troublesome to inoperable. Taking into account thesedrawbacks, new catalyst designs must offer: (1) accessibility tolarge asphaltene agglomerates in order to increase their diffusionand transformation, and (2) catalytic active sites or supportadsorption sites that are less favorable to sediment formation.In the first approach, and to overcome the difficulty ofasphaltene diffusion through small pores, an appropriate porousstructure can be obviously rationalized in terms of creatingmacropores (>1000 A ˚ ) in the shaped support or catalyst.However, not thatmanymacropores (ca. above 30%of total porevolume) should be created, since shaped particles could turnfragile. The formation of macropores in extrudates is a directconsequence of support or catalyst formulation when makingextrudates. In fact, new generation commercial catalystsdeveloped for the H-OilTM process contain macropores [3].The second approach deals with the catalyst support’s intrinsicacid sites (e.g., g-Al2O3) [4], which are believed to participate insediment formation.In this study, an alumina–carbon black composite supportwas developed with the aim of diminishing either the strength,and/or number of acid sites. Carbon black: (i) is cheap, (ii) hasa low tendency to form coke-related compounds (e.g., veryweak acid sites), (iii) has an affinity for Vand Ni porphyrine-like compounds, which could be refrained from attackingcatalytic active sites, (iv) has high hydrodesulfurization(HDS) [5,6] and hydrodenitrogenation (HDN) [6–8] activity,and (v) has no micropores below 30 A ˚ whichcouldbeeasilyoccluded, therefore turning active Ni and Mo metals intoinactive ones. In earlier works, microporous activated carbonwas used as a support, despite the fact that micropores are oflittle use for catalytic reactions involving large molecules [9],and it is more expensive than carbon black. Unfortunately,carbon or carbon black supports are known for their lowmechanical resistance; this physical property being of theutmost importance since shaped catalyst particles will becolliding among themselves and against reactor walls in theebullated bed reactor. On the other hand, pyrolyzed carbon, orpyrolyzed carbon black, has good mechanical properties.Very little work has been reported on the hydroconvertion ofheavy residues on NiMo- or CoMo- catalysts based on carbonsupports. For instance, Segawa et al. used NiMo on activecarbon as a catalyst for the hydrogenation of an Arabianvacuum residue, and they found that coke formation wasconsiderably suppressed [10].One of the challenges of this study is to achieve an intimateand homogeneous mixture of Al2O3 and carbon black in anextrudable paste, which could eventually result in macropore-containing, mechanically resistant, Al2O3–carbon black extru-dates. In order to record deactivation, using near real-operationconditions, selected catalysts, obtained from impregnatingthese prospective supports simultaneously with Ni and Mo,were evaluated and monitored continually for about 700 h onthe run. A bimodal alumina-based NiMo commercial catalystwas used as a comparison.2. Experimental2.1. Preparation of catalysts2.1.1. Support preparationCatalysts were prepared by closely following the preparationprocedure shown in Fig. 1.Commercial boehmite-phase alumina(325 m2/g; acid dispersability index: 30%; loose bulk density:192–224 kg/m3;Na2O < 0.01 wt.%), dried at 110 8C, was usedthroughout the study. This alumina (200 g) was kneaded with 1–1.5vol.%acetic acid (280–320 cm3) for about 30 min. Then, 15 gof carbon black #970 (particle size: 16 nm, specific surface area:260 m2/g), fromMitsubishi Chemicals, and sucrose (15 g), usedas a binder, were poured in, after which kneading continued forabout 15 min. The black wet paste obtained was kept in a closedplastic vessel for 12 h at room temperature, in order to maintainmoisture and allow homogenization of the mixture. Then, thepaste was extruded in a Brabender mechanical extruder at aconstant speed (30 rpm), using 1/16 in. dice.Wet extrudateswereFig. 1. Preparation scheme of AMAC supports and catalysts. allowed to dry at room temperature overnight, after which theywere dried at 100 8C for 3 h, and 110 8C for 1 h in a static oven.After this, the cylindrical-shaped particles were transferred to atubular oven and were heat-treated in flowing 6 vol.% O2/N2,where the temperature was increased 5 8C/min and thenremained at 600 8C for a period of 2 h. The activated blackextrudates were then impregnated with Ni and Mo. Hereinafter,AMAC, which stands for alumina modified with amorphouscarbon black, will be the term used to refer to these supports.2.1.2. Ni and Mo impregnationFirst, an aqueous ammonia (NH4OH + H2O 3:1) solution of[Ni5(CO3)2 (OH)6 4H2O] and MoO3 was prepared, in order toobtain 3.5 wt.% NiO and 15 wt.% MoO3 on an 81.5 wt.% driedAMAC support. The amount of solution was calculated as 80%of the total pore volume, determined by water absorption, i.e.,the incipient wetness technique. This metals solution wasevenly sprayed by means of an atomizer, onto the shapedsupport in a rotary impregnation vessel. Then, the impregnatedextrudates were dried for 12 h at room temperature, at 100 8Cfor 3 h, 110 8C for 1 h, and finally heat-treaded in flowing N2where the temperature was increased 5 8C/min and thenremained at 600 8C for a period of 2 h. These catalysts will bereferred to as AMAC catalysts.2.2. CharacterizationPore-size distribution and pore volume of support andcatalysts were determined by a Quantachrome Autoscan Hg-porosimeter, using samples previously dried at 120 8C for 4 h.Total pore volume, defined as water-holding capacity, wasmeasured as maximum water absorbed into a dry support orcatalyst. The size (i.e., length and diameter) of the support andcatalysts was measured using an electronic digital clipper, andthe side crushing strength (SCS) was automatically measuredusing a Seiko-Kyowa KA-300PB precise Rheorobot, closelyfollowing theASTMDD4179-01method. The amount of carbonblack incorporated into a support was determined via thegravimetric method by burning at 600 8C, 2 h in air; sampleswere previously heat-treated in 6% O2/N2, as indicated in thesupport preparation section. Loss-on-attrition was measured asfollows: after stirring 50 g of extruded catalyst in 500 g of dieselfor 8 h at 100 rpm and room temperature in a stainless steelbeaker, all the particles that passed through mesh #20 werequantified and recorded as wt.%.2.3. EvaluationFeedstock was a vacuum residue (ca. 80%) made up of amixture of 60% Isthmus and 40%Maya crude oils, and vacuumgasoil cutback (ca. 20%) characterized by physico-chemicalproperties shown in Table 1.