Parviz Rahimi

The effect of carbon additives on the mesophase induction period of Athabasca bitumen

Abstract The ability of solid carbonaceous material to retard the formation of coke during thermal cracking and hydrocracking of heavy hydrocarbons is well known. In this study, we used in-situ microscopy (hot-stage) to obtain additional mechanistic information on whether fine coke and fullerene soot particles retard the growth of mesophase during thermal cracking of Athabasca bitumen, thus reducing the possibility of fouling in preheaters and furnaces. The findings from this study could also have application in other non-catalytic thermal processes such as visbreaking and coking. In the absence of additives in the Athabasca bitumen feed, the formation of mesophase occurred after 61 and 67 min (measured from room temperature) at reaction temperatures of 450°C and 440°C, respectively. The addition of solid coke (ca. 5 wt.%) from a commercial delayed coking operation shortened the mesophase formation time to almost 45–50 min under similar conditions. The coke, having surface area of only 1.65 m 2 /g, resulted in enhanced bitumen fluidity and large-textured mesophase. These observations were rationalized based on the ability of delayed coker coke to release hydrocarbons into the bulk fluid during thermal cracking. Light hydrocarbons released from coke may have changed the solvating power of the liquid phase in bitumen and promoted phase separation, resulting in a shorter induction period. In contrast, adding small amounts of fine fullerene soot (ca. 1 and 5 wt.%) delayed the appearance of mesophase significantly under similar conditions. The ability of fullerene soot to physically absorb the mesophase precursors into its pore structure led to an increase in the apparent viscosity of the bulk phase, which is known to reduce mesophase size and prolong the induction period. Consistent with this, the induction period was prolonged an additional 10 min when the soot surface area was increased from 152 to 208 m 2 /g. The increase in induction period is significant with respect to reaction times and suggests that these fullerene soot materials could be effective in allowing for increased severity and liquid products yield from visbreaking, with less likelihood of fouling in the preheater tubes and furnace walls.

Thermal hydrocracking of Cold Lake vacuum bottoms asphaltenes and their subcomponents

The thermal chemistry of bitumen and heavy oils is extremely complicated, mostly because of the complex nature and the unknown molecular structure of different components that are present in these materials. In order to understand the molecular transformation of bitumen during upgrading better, the following approach was adopted: separate the bitumen into different components (asphaltenes and maltenes), characterize them in detail and react each fraction separately. In the present work, the results of thermal reaction of asphaltenes and its subfractions (A1–A4) separated by column chromatography are presented. Thermal hydrocracking of asphaltenes derived from Cold Lake vacuum bottoms (CLVB) was investigated, neat and in a hexadecane solvent at 440 °C (30 min and 13.8 MPa H2). The amount of coke measured as methylene chloride insolubles remained relatively constant at about 25 wt.%, whether the asphaltenes reacted neat or in hexadecane. Increasing asphaltenes concentration in hexadecane had little effect on coke formation. Results from the present study were compared with those reported previously from our laboratory for the same CLVB asphaltenes, which were thermal hydrocracked in CLVB maltenes and in decalin, and produced significantly less coke than in hexadecane. Unlike the reaction in hexadecane, changing the concentration of asphaltenes in maltenes had a significant effect on coke formation. The results from the previous study were rationalized in terms of competing reactions between maltenes and asphaltene-derived radicals for hydrogen, solvating effects of the media on coke formation, and viscosity effects on the radical–radical combination to form coke. Hexadecane as a solvent was relatively inert when compared to maltenes under the same severity conditions and did not stabilize the reactive radicals or prevent retrogressive reactions. The thermal hydrocracking of CLVB asphaltenes subcomponents (A1–A4) was also investigated in hexadecane. A better correlation was obtained between the coke yield and the molecular weight than either microcarbon residue or the aromaticity of the subcomponents. Based on the lack of additivity in coke yield, it was concluded that a synergistic effect among the subcomponents was most likely responsible for the reduced coke yield obtained from processing the asphaltenes feed.

Preliminary phase diagrams for ABVB+ n-dodecane + hydrogen

Abstract The phase behaviour of athabasca bitumen vacuum bottoms (ABVB) + n -dodecane + hydrogen mixtures was investigated at temperatures and pressures ranging up to 725 K and 7 MPa using an X-ray view cell. Preliminary phase diagrams are presented. The most complex phase behaviour observed is tentatively identified as “S”LLV, although the nature of the “solid” is unclear. From the slope of the “S”LV to LV phase boundary, ΔH fusion is less than 7 kJ/kg. Thus it may be an amorphous solid, a viscous liquid, or micellular in nature. Observed multiphase behaviour is also compared with operating conditions for well known heavy oil upgrading processes. Most processes appear to skirt multiphase zones. A preliminary four component model for the phase behaviour of the ABVB + hydrogen system has been devised using CMG-PROP and HYSIM. An example calculation comparing modelled and experimental phase behaviour is also presented.

Coking propensity of Athabasca bitumen vacuum bottoms in the presence of H-donors — formation and dissolution of mesophase from a hydrotreated petroleum stream (H-donor)

Abstract The effect of H-donors (hydrogenated aromatic hydrocarbons — HHAP) derived from a petroleum stream on mesophase formation during thermal hydrocracking of Athabasca bitumen vacuum bottoms (ABVB, +525°C) was investigated using real time high pressure and temperature microscopy (hot-stage) under a nitrogen atmosphere. When ABVB was treated alone, mesophase formed at a relatively fast rate after only 66–70 min (at 440°C). When HHAP was treated alone under the same conditions, it showed a highly fluid behavior. This material did not develop any mesophase even after a prolonged first cycle of heating (>140 min), but eventually formed mesophase during a second cycle of prolonged heating. Two distinct mesophase types were observed during thermal treatment of HHAP. One type of mesophase dissolved upon heating at moderate temperatures and re-appeared upon cooling, an indication of “thermotropic transformation” typical of true liquid crystalline material derived from low molecular weight components. The other type of mesophase that was most likely derived from high molecular weight components did not dissolve but formed bulk mesophase. ABVB was then mixed with 5 wt.% HHAP to observe the effect of an H-donor on mesophase formation from bitumen. In the presence of HHAP, both the rate and the amount of mesophase formed from ABVB were reduced significantly. Furthermore, the mesophase appearance time (induction period) was prolonged by as much as 20 min. This paper demonstrates the usefulness of hot-stage microscopy as a tool for screening potential additives in visbreaking and fouling operations.

Phase splitting of complex hydrocarbon mixtures

Abstract Significant fluctuations in the light oil yield obtained from laboratory and pilot scale models of heavy oil upgrading, coal/oil coprocessing and direct coal liquefaction processes, arising from apparently minor perturbations in operating conditions, are frequently attributed to synergism. Some aspects of observed synergistic behaviour are explained in terms of relevant liquid-liquid and fluid-fluid equilibria. Partial phase diagrams for simple and complex hydrocarbon mixtures including model solvent mixtures: pyrene/naphthalene/tetralin, and bitumen/heavy oil/anthracene oil mixtures are presented. As the projected or actual operating conditions for hydrogenation processes and the temperature and pressure regions over which the mixtures can experience phase splitting overlap, previously reported synergism is related to phase splitting and compounding factors such as catalyst wetting and localized reagent depletion.

PRELIMINARY PHASE DIAGRAMS FOR BITUMEN/ HEAVY OILS AND RELATED MIXTURES

The phase behaviour of heavy oil mixtures was studied using an X-ray imaging system. Batch phase experiments were performed at temperatures and pressures up to 725 K and 7 MPa. Complex phase behaviour such as liquid-liquid-vapour and solid-liquid-liquid-vapour were observed and preliminary experimental phase diagrams were constructed. This observed complex phase behaviour is consistent with thermodynamic theory and such phase behaviour can be modelled using the Peng-Robinson equation of state and the tangent plane criterion. These latter points are illustrated through the phase diagram for a model reservoir fluid (ethane + propane + n-butane + phenanthrene) which exhibits solid-liquid-liquid-vapour phase behaviour. Example predicted and experimental phase diagrams are presented for this model system.

Coprocessing : elemental and molecular weight distributions in unconverted vacuum residues

The difference between molecules that are converted to distillates during coprocessing (Cold Lake vacuum bottoms and Forestburg subbituminous coal) and those that are not has been studied. The following procedure was used. The fraction of the liquid product boiling above 525 °C was separated into five subfractions by preparative scale gel permeation chromatography (GPC). Each subfraction was weighed. Then the following measurements were made on each subfraction ; molecular weight, elemental analyses (C, H, N, S, V, and Ni), and carbon-13 nuclear magnetic resonance. Five significant observations were made. (1) The molecules that were converted had larger H/C atomic ratios and smaller N/C atomic ratios than the feedstock molecules. (2a) S, V, and Ni heteroatoms could be removed without causing much change in molecular weight. (2b) It was not possible to increase the H/C atomic ratio or decrease the N/C atomic ratio without decreasing the molecular weight. (3) Generally, the unconverted +525 °C residue molecules became smaller as the processing severity increased. (4) The unconverted molecules retained their side chains at mild processing severities (425 °C). When the processing severity increased (450 °C), side chain carbon began to be removed. (5) If there was insufficient reaction time to provide enough hydrogen, then molecules which were larger than the feedstock molecules were formed. Feedstock molecules that lost their hydrogen rich fragments (by cracking or by hydrogen transfer) without contacting enough hydrogen to remove their nitrogen or metal heteroatoms may have oligomerized. If the reaction time was increased to allow more contact time with hydrogen, the oligomerized molecules which were larger than the feedstock molecules disintegrated. Finally, conversion of residue molecules to distillate molecules appeared to be limited by hydrogen addition. For conversion, hydrogen was required either to hydrogenate aromatic rings or to remove nitrogen heteroatoms. There are other important requirements for hydrogen which are not primary steps in the conversion of large molecules to small ones. They include capping pyrolysis fragments and the removal of other heteroatoms (sulfur, metals).