SECOND (48 MONTH) PROGRESS REPORT
INDUSTRIAL RESEARCH CHAIR ON ADVANCED UPGRADING OF BITUMEN
WITH SYNCRUDE CANADA LTD.
February, 2004
Chair Holder: Murray R Gray
Department of Chemical and Materials Engineering
University of Alberta, Edmonton, AB T6G 2G6
Phone: (780) 492-7965 FAX: (780) 492-2881
1. RESEARCH PROGRAM
The scope of the chair program is research on the conversion of oilsands bitumen into value-added products, with active participation in the research by our partners at Syncrude and the National Centre for Upgrading Technology. The major thrust of the research effort is to understand how thermal cracking processes determine the conversion of the feed to products, and the yield and quality of those products. Complementary projects are examining more selective transformation of nitrogen compounds in cracked products, by using chemical and biological catalysts. The projects feature extensive collaborations with Syncrude and university researchers, as described in more detail in Sections 4 and 5. The research projects under the chair are classified in five areas: coking and thermal cracking (most active), bioprocessing, hydrotreating, fundamentals of bitumen chemistry and chlorides and inorganic components. Two new partners have joined Syncrude, NSERC and NCUT in supporting the chair program for 2003/2004; Champion Technologies Ltd and Alberta Energy Research Institute.
The major accomplishments of the Chair program to date include:
· Measurement of fluid viscosity and surface tension cracking conditions, a world-wide first, through collaboration of a large multitalented team from the University of Alberta and Syncrude. The insight from these results, along with parallel studies on the fundamentals of coke formation, is important in improving reactor technology for processing oilsands bitumen. This research has also given us new capabilities for visualizing the reactions of bitumen at typical process conditions.
· Establishing new capabilities in catalysis research at University of Alberta, through the hiring of Alan Nelson and through the establishment of the Alberta Centre for Surface Engineering and Science (ACSES), with Gray as principal investigator and Nelson as one of 9 co-investigators. With $13 million in support for equipment and facilities from CFI and ASRIP, ACSES will be a superb facility for surface analysis and modification for the research community in Western Canada.
· Successful representation of the heaviest components of bitumen as molecular mixtures, which opens new methods for analyzing the reactions of bitumen and predicting its thermodynamic and physical properties. This success was due to collaboration with William McCaffrey and Heather Dettman of NCUT.
1.1 Accomplishments to Date
1.1.1 Coking and Thermal Cracking of Oilsands Bitumen
This research area has produced the most significant breakthrough in the chair the chair program; the measurement of bitumen fluid properties during thermal cracking and coke formation. A collaborative team of up to 7 people developed a completely new approach to measuring the surface tension and the viscosity of bitumen as it cracks and forms coke at temperatures of up to 530 °C. Due to the rapid evolution of vapor from cracking reactions, no existing rheological methods were suitable. The reaction kinetics also dictated that the liquid phase be heated to a fixed temperature in a matter of seconds, and that the measurements of fluid properties be achieved in seconds. Our team (William McCaffrey, Janet Elliott, Zhenghe Xu, Iftikhar Huq (Syncrude), Jason Zhang (now at U Ottawa) and graduate student Moruf Aminu) met this challenge by using thin films of bitumen on rods Curie-point alloy that were rapidly heated in an induction furnace to a fixed temperature. When two rods were touched, a liquid bridge was formed. The forces between the rods during the elongation of the liquid bridge gave estimates of the surface tension and viscosity. By measuring the bridge forces as a function of reaction time, we determined the time for the liquid feed to become a dry solid, and the surface tension and viscosity as a function of time. Surface tensions were low, in the range of 6 mN/m, and insensitive to the extent of reaction. In contrast, the viscosity increased by four orders of magnitude to order 10,000 mPa.s. This information was very significant for Syncrude in designing new methods for injecting liquid feed into coking reactors, and in understanding and modeling the formation of coke deposits within the reactors. Beyond Syncrude’s immediate interest, these results are important for any new upgrading technologies based on thermal cracking of bitumen at low pressure.
This novel approach to measuring fluid properties also allowed us to watch bitumen change and react during thermal processing. By observing thin films of asphaltenes on rods and strips of Curie point alloy, we were able to determine the melting point of this important fraction of bitumen. Previously, many researchers thought that asphaltenes would decompose before melting, but we have now observed that asphaltenes form a liquid melt at 200-240 °C. This observation has opened new directions for research on bitumen fractions.
Three projects were completed on key fundamentals of cracking of bitumen to form distillate products and coke. The first project, by post-doctoral fellow Richard Dutta with William McCaffrey and Karlis Muehlenbachs, developed a novel isotopic tracer technique for determining the factors that control the formation of solid coke, which accounts for approximately 20% of the bitumen processed by Syncrude. 13C labeled styrene was polymerized with toluene to give dimers and oligomers with a range of boiling points in the range of desired distillate products. The isotope tracer was then mixed with bitumen and reacted in liquid films of different thickness, from 15-150 mm. The yield of the coke increased with liquid film thickness, consistent with coupling between the diffusion of products in the liquid phase and reverse reactions to form coke. This result was confirmed by isotopic analysis of the coke, which showed enrichment by the 13C tracer, in proportion to the thickness of the liquid film. This study was the first definitive demonstration of the role of mass transfer in coke formation in bitumens. A follow-up study by Naras Srinivasan with William McCaffrey and Keng Chung showed that unsubstituted aromatic compounds contributed very little to coke formation. This result suggested that polymerization of aromatic rings through attached side chains the most likely mechanism for the formation of coke.
The final study was completed by master’s student Shakir Japanwala on cracking and coking of residue fractions during recycle, in collaboration with Keng Chung (Syncrude) and Heather Dettman (NCUT). Commercial reactors recycle the unconverted residue to extinction, unlike laboratory studies which involve a single pass through the cracking reactor. The recycle operation was simulated by recovering the remaining residue (524 °C + fraction) and subjecting it to the same coking reaction, and again for a third time until almost all of the residue was converted. This approach showed that 99% of the feed residue fraction was converted to coke of volatile products after three reaction stages at 530 °C. The detailed characterization of the products by nuclear magnetic resonance and liquid chromatography provided significant insight into the quality of the liquid products as a function of increasing conversion of the heavy feed fractions.
The experience gained from the studies under the chair program has led to a major reassessment of the reaction mechanisms that underlie the upgrading of the residue fractions. The result was a comprehensive scheme to link the roles of thermal reactions and catalysts in upgrading, using well established chemical principles (Gray and McCaffrey, 2002), which provides a consistent framework for observed reactor behavior to the underlying chemical fundamentals.
1.1.2 Biocatalysis for Removing Nitrogen Compounds
Research on biocatalytic removal of nitrogen compounds was initiated during 2001 in collaboration with Philip Fedorak, Julia Foght and Mike Pickard, and from Syncrude Craig McKnight and Vince Nowlan. We focused on the biocatalytic conversion of the nitrogen compound carbazole to more easily treated organo-nitrogen compounds. Carbazole compounds are one of the most resistant contaminants in products from the oilsands, so novel approaches to removal are a major priority. Working from a knowledge of the pathways for attack on aromatic compounds by bacteria, post-doctoral fellow David Bressler tested the catalytic hydrogenation of a range of potential products from carbazole. His work confirmed that biological pre-processing can render carbazole compounds more easily treated by conventional catalysts. He also showed that the most desirable products would have the nitrogen completely removed. This information was used in the next stage of the project, which was to select organisms for use as biocatalysts. Master’s student Leah Kirkpatrick isolated a number of new bacterial strains that remove nitrogen from carbazole. Unfortunately, none of these strains were able to attack the nitrogen atom selectively. While some strains were more efficient in converting carbazole than other aromatic compounds, none of them were able to remove the nitrogen and leave the rest of the carbon backbone. In every case, nitrogen was only removed after much of the carbon in the carbazole had been converted to carbon dioxide. This work suggests that bacterial enzymes are not capable of selectively attacking nitrogen compounds, in the presence of other aromatics, therefore, a bioprocessing route to removal of carbazole compounds is unlikely to give better performance than existing hydrotreating technology.
The experience from this study, and other related work on my group, led to a reassessment of the role of transport processes in bioprocessing of petroleum components. Published in 2003 (Bressler and Gray, 2003), our review of this area suggests that bacterial processing of residue fractions is very unlikely to succeed, but that the lighter distillated continue to be an excellent target for innovative process pathways.
1.1.3 Catalysis for Removing Nitrogen Compounds
One of the challenges in processing oilsands products is that they are more resistant to catalytic hydrotreating than conventional petroleum fractions. New catalysts which give enhanced performance with other feeds usually fail to give better results with oilsands distillates. Master’s student Will Kanda completed a study of inhibition of hydrogenation of the model compound quinoline by narrow boiling fractions of Athabasca distillates at 330-380 °C, in collaboration with John Adjaye, Vince Nowlan and Keng Chung from Syncrude. Using a novel microreactor technique, Kanda measured the inhibition due to a variety of nitrogen-bearing fractions. His results showed significant inhibition of quinoline conversion by all of the narrow fractions, but the most significant inhibition was due to relatively light material (bp. 343-393°C), in comparison to higher-boiling fractions. Working with Alan Nelson and John Adjaye, technologist Iva Siu extended this work to examine the role of hydrogen pressure and sulfur content on the inhibition, and to verify that alkyl carbazoles and other pyrrolic nitrogen compounds accounted for the bulk of the inhibition.
One of the major benefits of the chair program to the University of Alberta was the hiring of Dr Alan Nelson to initiate research into the fundamentals of catalytic processing of oilsands components. Under the chair program, research on HDN catalyst fundamentals was began in 2002 in collaboration with Alan Nelson and John Adjaye (Syncrude). The focus of this study is to understand the adsorption and denitrogenation reaction networks of basic and non-basic organonitrogen species on model catalyst surfaces. Using surface science techniques and molecular simulations, Ph.D. student Wa’el Abdallah is studying the adsorption of pyridine and pyrrole on Mo(110) and C/N-modified Mo(110) to develop fundamental insight into HDN reaction networks. This work has given us new insight into the detailed interactions of pyridine with a metal surface. Through additional studies, this approach will serve to; (1) determine the correlation between surface composition and activity, (2) identify the molecular adsorption species and reactive intermediates, and (3) investigate the way in which carbidic (and nitridic) overlayers affect reaction networks and surface species. The development of new catalysts that are selective to C-N bond scission, through and understanding of catalyst fundamentals, is the ultimate purpose of this research.
An additional objective of the chair program was to expand the capability for oilsands upgrading research through the hiring of an Assistant Professor. To this end, Alan Nelson has established an independent and complementary research program in hydrotreating catalyst fundamentals. One key area of research being actively pursued is the modeling of hydrotreating catalysts and reaction networks. The active sites on hydrotreating catalysts (MoS2-based) are located on the edge surfaces of the MoS2 layered structure, but the exact structure of the active sites is still not fully understood. Working from fundamental structural information, post-doctoral fellow Mingyong Sun is studying the local structures of CoMoS and NiMoS hydrotreating catalysts at reaction conditions by considering the effects of temperature and H2S/H2 ratio using density-functional theory (DFT) investigations. This work has confirmed that catalyst sulfidation conditions not only affect the equilibrium sulfur coverage on the edge surfaces of unpromoted MoS2 catalysts, but also affect the atomic structure of the metal sulfide that form when nickel or cobal promoters are added. This detailed understanding of the catalyst at the atomic level has significant implications for designing new hydrotreating catalysis, and will support further modeling studies on organonitrogen compounds to support the development of oilsands-specific hydrotreating catalysts.
A second area being studied by Alan Nelson is selective aromatic saturation, or selective ring opening, of bitumen derived middle distillates to improve the cetane quality. The difficulty with the current technology is that ring opening is non-selective, and consequently results in low cetane number products. One approach to solve this problem is to develop catalysts that selectively open naphthenic rings while limiting the extent of cracking. Master’s student Carolyn Kenney is studying the structural and thermochemical properties of solid-acid supported metal catalysts in order to relate these properties to ring opening activity and selectivity. Her work suggests substantial benefit may be achieved by utilizing a mixed-metal oxide support to adjust the surface acidity and bulk structure. These results also give insight into the reaction networks for catalytic ring opening, which have been classified into two categories; (1) classical bifunctional and bimolecular, and (2) non-classical Haag-Dessau. The prevalence of each reaction scheme in the reactions of mono- and poly-aromatic ring compounds is one of frequent debate in the literature. Master’s student Alex Abraham is extending the experimental work to lend insight into this area by relating these catalyst properties with ring opening networks using molecular simulation and transition state theory. This fundamental and applied research will effectively elucidate the relationship between metallic and acidic functionality, and catalytic reactivity and selectivity to advance the fundamental understanding of catalytic ring opening networks.