A New Refining Process for Efficient Naphtha Utilization: Parallel Operation of a C 7+ Isomerization Unit with a Reformer

A New Refining Process for Efficient Naphtha Utilization: Parallel Operation of a C 7+ Isomerization Unit with a Reformer Authors: Dr. Cemal Ercan, Dr. Yuguo Wang and Dr. Rashid M. Othman ABSTRACT Gasoline is a complex mixture of hydrocarbons containing 5 to 12 carbon atoms and having a boiling point range of 40 C to 190 C. It is a blending of many streams from various refining processes, which fulfills certain specifications dictated by both performance requirements and government regulations. Reformate makes up approximately one-third of the gasoline pool, and with its 60 vol% to 70 vol% aromatic content, it has been the main octane source for gasoline over the years. Gasoline specifications have been gradually changing in past years due to the regulations dictated by safety and environmental concerns. With the decrease of aromatics in gasoline, the role of reformate as the main octane source is expected to shrink. Based on this trend and the restriction on aromatic content, it is believed that future gasoline will be mostly branched paraffins. The reduction of aromatics in gasoline will create a big octane gap, and refineries will have to find economic solutions to close the octane gap and prepare for the future. Isomerization of gasoline range n-paraffins, C 7 -C 12, is one of the more economical and environmentally acceptable ways to address the aromatic issue and ease the octane gap. The novel process proposed here is a step in the direction of helping refineries in this search. The process, which involves the parallel operation of a C 7+ paraffin isomerization unit with a reformer, will also create additional benefits for refineries by substantially increasing the overall liquid yield and easing the conditions of operation. The process is based on the separation of n-paraffins from heavy naphtha by adsorption, then processing n-paraffins in a dedicated C 7+ isomerization unit and non-paraffinic heavy naphtha in a reformer. In a C 7+ isomerization unit, n-paraffins will be isomerized at optimum operating conditions, and the isomerate produced can be sent directly to the gasoline pool. The reformer unit with non-paraffinic feed will be operated at mild operating conditions, and the products can be blended into the gasoline pool as required or can be used for petrochemicals. The absence of n-paraffins in the feed will allow operating the reformer at mild operating conditions, which is expected to increase the liquid yield up to 10 wt%, substantially improve catalyst life and increase hydrogen production. Additionally, the separation of aromatics and the purification of hydrogen from the concentrated reformer effluents will be easier and cheaper. INTRODUCTION Gasoline is a complex mixture of hydrocarbons containing carbon atoms 5 to 12 and having a boiling point range of 40 C to 190 C. Modern reformulated gasoline is a blend of several refining streams, which fulfills certain specifications dictated by both performance requirements and government regulations. Typical gasoline blending streams are presented in Table 1 1-3. Blending Compound Gasoline (vol%) Comments Desired Gasoline Specifications Specifications Range FCC Naphtha 30 50 Has ~30 vol% aromatics and 20 to 30 vol% olefins Octane number 90 95 LSR Gasoline (Naphtha) 2 5 Sulfur (max) 10 15 ppm Reformate 20 40 Has ~60 to 65 vol% aromatics Aromatics (max) 25 30 vol% Alkylate 10 15 Benzene (max) < 1 vol% Isomerate (C 5 /C 6 ) 5 10 Olefins (max) 10 18 vol% Oxygenate (MTBE) 10 15 Oxygen (max) 2 2.7 wt% Butanes < 5 Vapor Pressure (max) 7 8 psia Table 1. Gasoline blending streams of modern refinery with typical compositions and properties FALL 2015 SAUDI ARAMCO JOURNAL OF TECHNOLOGY

Fluid catalytic cracking (FCC) naphtha and reformate make up approximately two-thirds of the volume of gasoline. Since both FCC naphtha and reformate contain high levels of aromatics and olefins, they are also the major octane source for gasoline. Figure 1 summarizes the octane values of various refining streams. Over the years, safety and environmental concerns have caused gasoline specifications to change, and as a result, refineries have changed their operations and gasoline compositions accordingly. As an example, the evolution of European gasoline specifications over the years is presented in Table 2, which shows a gradual change of the gasoline specifications from 1994 to 2010. A more or less similar trend is observed in other parts of the world 1. Table 2 also shows a gradual decrease in aromatic, olefin and benzene levels while the octane value remains high. The U.S. already requires aromatic levels of less than 30 vol%, with benzene levels limited to 0.8%. Furthermore, the aromatic level limit in gasoline will soon fall even lower; particularly as distillation end points (usually characterized as the 90% distillation temperature) are lowered, thereby disallowing the high boiling point portion of gasoline which is largely aromatics. Since the aromatics are the principal source of octane, decreasing the aromatic level will create an octane gap in the gasoline pool. As such, octane maintenance will continue to be a challenge for refineries. The trends in changing gasoline specifications can be summarized as follows: A high octane number will still be required ~95 Research Octane Number (RON) which means that the demand for high performance will stay or even increase. The aromatic content will continue to be reduced, dropping below 35 vol%. The U.S. has already lowered it below 30 vol%. The benzene content will be lowered to < 1.0 vol%. It is expected eventually to be phased out. The sulfur content will be reduced to ~10 ppm and is expected to be reduced further. The olefin content will continue to be reduced, dropping below 10 vol%. Vapor pressure will be low, around 7.0 psi. Reformulated gasoline specifications after January 1, 1998, introduced a major reduction in the distillation range. New specifications will completely eliminate light products, such as C 4. The optimal oxygen content will be maintained max 2.7 wt%. Fig. 1. Octane values of various refinery streams. As the aromatic content of gasoline goes down, the portion of reformate in the gasoline pool has to go down, too, since reformate is the main source of aromatics. Therefore, refineries can no longer heavily rely on reformate or aromatics as an octane source, and alternative octane sources are needed. An ecologically sound way of closing the octane gap due to the European 1994 1995 2000 2005 2010 Sulfur, wt ppm max 1,000 500 150 50/10 < 10 Aromatics, vol% max 42 35 < 35 Olefins, vol% max 18 18 <10 Oxygen, wt% max 2.7 2.7 2.7 Benzene, vol% max 5.0 1.0 1.0 < 1.0 RVP, psi max 5.8 10.2 6.5 8.7 6.5 8.7 6.5 8.7 Distilled at 100 C (min), v/v% 54 65 46 71 46 71 46 71 Distilled at 150 C (min), v/v% 75 75 75 RON/MON (min) 95/85 95/85 95/85 Table 2. European gasoline specifications SAUDI ARAMCO JOURNAL OF TECHNOLOGY FALL 2015

reduction of aromatics is by increasing the concentration of the branched alkanes at the expense of normal paraffins. Consequently, an increase in iso-alkanes with a high octane number is desirable. The novel refining process described here is intended to address the octane gap generated as a result of aromatic reduction in gasoline. The new process will also provide a substantial improvement in the performance of the reformer. PROCESS The novel process for refining naphtha proposed 4 here begins with first separating a naphtha feed into light naphtha C 5 /C 6 paraffins and heavy naphtha. Second, introducing light naphtha to the first isomerization unit C 5 /C 6 isomerization under usual isomerization conditions will produce a light isomerate, separating the heavy naphtha into a heavy n-paraffin C 7+ n-paraffins and a heavy non-paraffinic naphtha. Third, introducing the heavy n-paraffins to the second isomerization unit C 7+ isomerization unit under different isomerization operating conditions will produce a heavy isomerate. Fourth, introducing the heavy non-paraffinic naphtha to a reforming unit under milder reforming conditions will produce reformate. Finally, combining at least a portion of each of the light isomerates, the heavy isomerate and the reformate will form a gasoline blend. Figure 2 is the process flow diagram of the proposed novel process. In this process, the initial steps separating a naphtha feed into light naphtha and heavy naphtha, and processing the light naphtha in the C 5 /C 6 isomerization unit is a standard refining process, and nothing is new about it. Novelty comes after that, i.e., separating the heavy naphtha into heavy n-paraffins, C 7 C 12, and heavy non-paraffinic fractions, then processing these fractions in C 7 C 12 isomerization and reforming units, respectively. To understand the advantages of this novel process, the chemistry going on inside the reformer first needs to be reviewed in detail. As previously mentioned, the reformate with high aromatic content is typically the main octane source for gasoline. The conventional feed to a reformer, e.g., heavy naphtha, contains mostly C 7 -C 12 paraffins, naphthenes and aromatics. The purpose of reforming is to produce aromatics from naphthenes and paraffins for use in various applications. Among the group of chemicals, aromatics go through the reforming reactor as they are; naphthenic dehydrogenation to aromatics is rapid and efficient. So, naphthene conversion dehydrogenation goes almost to completion at the initial part of the reactor or in the first reactor of a multi-reactor reforming unit (the latter requiring an even less severe operation) at a mild temperature. Paraffins are very difficult to convert, however, and require a higher temperature and longer residence time in the reformer. Some conversion of the paraffins occurs toward the end of the reactor system at highly severe operating conditions, where most of the paraffins undergo cracking into light gases. Therefore, to increase the paraffin conversion, highly severe operation conditions are needed, which decreases liquid yield due to excessive cracking. The cracking of paraffins also decreases the hydrogen yield. As shown in Fig. 3, although the octane number increases due to the concentrated aromatic content, a substantial liquid yield loss is observed. Therefore, in a conventional reformer, aromatics are primarily made via the dehydrogenation of naphthenes, and paraffins are considered to be the main source of liquid loss due to cracking. Hydrogen is also produced primarily by naphthene dehydrogenation 4, 5. Heavy naphtha, which is normally fed into a reformer, contains paraffins, naphthenes and aromatics. According to its paraffin content, heavy naphtha is classified as lean or rich naphtha. The naphtha with a high concentration of paraffins is referred to as lean naphtha. Lean naphtha is difficult to process and typically produces too much light cracked products, so has a low liquid yield. In comparison, rich naphtha makes the reforming unit s operation much easier and more efficient, so it is more desirable as a reformer feed than lean naphtha. The rich naphtha is relatively easier to process and has a higher liquid yield. Figure 4 schematically illustrates the typical conversions of lean and rich naphthas at typical reformer operating conditions, and also indicates that for this Fig. 2. Process flow diagram of the proposed novel process for naphtha processing 4. Fig. 3. Reformate yields and aromatic content of a typical reformer. FALL 2015 SAUDI ARAMCO JOURNAL OF TECHNOLOGY

Fig. 4. Schematics of typical conversion of lean and rich naphthas in a reformer: A = aromatics, N = naphthenes, and P = paraffins. typical case, the reformate produced using a rich naphtha has a liquid yield of approximately 10 wt% greater than the reformate produced using a lean naphtha basically due to the presence of low levels of paraffins. Therefore, paraffin conversion in a reformer is considered slow, inefficient and unable to produce the desired products. The most desirable reaction in the reformer is dehydrogenation of naphtha to aromatics, which also produces 3 moles of hydrogen per mole of naphtha dehydrogenated. Because isomerization is an equilibrium reaction where low temperature favors isomerization to normal, the reformer temperature is too high for a quantifiable amount of paraffin isomerization to occur. Similarly, hydro-cyclization of paraffins does not occur in an appreciable amount in the reformer. So, in a conventional reformer, naphthene dehydrogenation to aromatics is the main aromatic octane maker, with hydrogen production an added result of this reaction. To summarize, paraffins in a reformer mostly undergo cracking, thereby reducing the liquid yield, and require otherwise unnecessarily severe operating conditions and a longer residence time (or more reactors). The separation of n-paraffins from heavy naphtha with a known method, such as adsorption, therefore produces a much more desirable feedstock one with negligible n-paraffin content for the reformer. With this kind of feedstock, reformer performance is expected to be improved, as will be discussed next. Typically, heavy naphtha feed contains around 10% to 40% of n-paraffins. Separating the n-paraffins from heavy naphtha with adsorption produces two feedstocks, namely heavy n- paraffins, C 7+, and heavy naphtha without n-paraffins, or nonparaffinic heavy naphtha. Non-paraffinic heavy naphtha is a feedstock ideal for a reformer due to its negligible paraffinic content. With the reduction of paraffins within the heavy nonparaffin naphtha, the naphthene and aromatic content increases, and the feedstock becomes rich naphtha. The processing of this feedstock in a reformer is easier, and the performance of the reforming unit improves substantially; a higher liquid yield is achieved at a lower reactor temperature, which means a longer catalyst life. Figure 5 shows the expected increase in liquid yield and the decrease in operating temperature as a function of naphthene plus aromatics in the feedstock 5. The points in Fig. 5 show experimental data for the operating temperature and liquid yield, C 5+ liquid product, as a function of the total percentage of napthene plus aromatics in the reformer feed. As seen, with the reduction of paraffins within the heavy naphtha, the naphthene plus aromatics content of the feed increases. As the naphthene plus aromatics content increases, the C 5+ liquid yield increases, and the reaction temperature decreases. Therefore, the performance of the reformer improves substantially providing a higher liquid yield due to less cracking and a longer catalyst life due to lower reactor temperature. The expected increase in liquid yield and decrease in operating temperature can be estimated from Fig. 5. The next question is: What to do with the heavy n-paraffins? Although the heavy n-paraffins, as they are, present a desirable feedstock for a steam cracker to produce light olefins, the proposed process uses a dedicated C 7+ isomerization unit to produce branched paraffins, which can be blended into gasoline. As previously mentioned, the operating temperature of a reformer is not suitable for isomerization. On the other hand, a dedicated C 7+ isomerization unit operating at optimum temperature will substantially improve isomerization while also minimizing cracking. The following example clearly illustrates the benefits of the proposed process. Assume that 100 kg of heavy naphtha of which 60 wt% are paraffins, 27.5 wt% are naphthenes, and 12.5 wt% are aromatics is sent to a reformer under typical reforming conditions. The resulting reformates would include 20.4 kg of non-aromatics and 47.6 kg of aromatics, thereby yielding a total liquid yield of 68 kg, or 68 wt% of the original feed. Now, assume that the same amount of naphtha, which has the same composition, is used as a feedstream. But, before sending the heavy naphtha to the reformer, approximately 40 kg of paraffins, about 67%, are extracted from it and sent to Fig. 5. Reactor temperature and reformate yield a function of naphthene plus aromatics (N+A) value; reaction conditions: 100 RON, p = 30 bar, WHSV = 2.0 h-1, and H2/HC = 4.5. SAUDI ARAMCO JOURNAL OF TECHNOLOGY FALL 2015

The first isomerization (C 5 /C 6 Isomerization) is not included because its performance is the same in both cases. Case I Case II-A Case II-B Reformate Feed Reformate Feed C 7+ Isomerization Feedstock (units) 100 60 40 C 5+ Yield 68 90 95 (wt%) Heavy Naphtha wt% wt% Paraffins 60 33.33 100 Naphthenes 27.5 Non-aromatics 29.96 45.83 24.77 100 Aromatics 12.5 Aromatics 70.04 20.83 75.23 Total 100 Total 100 100 100 N+A 40 RON 100 105 +20 Total Liquid Yield (wt%) 68 54 38 Case I: Reformer runs at typical operating conditions to obtain RON of 100. Case II: Reformer runs at milder conditions and C 7 + Isomerization runs at temperature lower than 300 C. Table 3. Comparison of performance of stand-alone reformer (Case I) and parallel operation of a reformer and a C 7 + isomerization unit the C 7+ isomerization unit. The remaining 60 kg of heavy naphtha is sent to a reformer, which can operate at a milder condition due to the lower paraffin content including at a lower temperature of approximately 10 C to 20 C. The resulting reformate includes 13.4 kg of non-aromatics and 40.6 kg of aromatics, for a total liquid yield of about 54 kg, which is about 90 wt% of the reformer feed. Meanwhile, the second isomerization unit has produced a total liquid yield from the extracted paraffins of approximately 95 wt% 38 kg out of 40 kg. Therefore, the overall total liquid yield for both the isomerization unit and the reformer is approximately 92 wt%. So, a net increase in liquid is 24 wt%. The summary of the results for both cases is presented in Table 3. CONCLUSIONS A new process for the efficient processing of naphtha, which is based on separating C 7+ n-paraffins from heavy naphtha, processing non-paraffinic heavy naphtha in a reformer and processing C 7+ n-paraffins in a C 7+ isomerization unit, is proposed. This unique process, using the parallel operation of a reformer and a C 7+ isomerization unit, produces branched C 7+ paraffins to be used in gasoline, concentrated reformate for use in gasoline as needed as well as in petrochemicals, and hydrogen, also for various uses. The absence of n-paraffins in the reformer feed allows the operating of the reformer at milder conditions with a substantial improvement in liquid yield. The benefits of this new process can be summarized as follows: Improved reformer performance: Higher liquid yield minimum 10 wt% higher than yields from a conventional reformer and hydrogen production. Milder operating conditions: ~10 C lower temperature, which has a positive effect on catalyst life. Shorter residence time in the reformer or a reduced number of reformer reactors. Reformate with a high aromatic content, making the aromatic separation for petrochemical use easier. Higher hydrogen concentration in the off-gas due to less cracking, making hydrogen purification also easier. An octane booster: Dedicated C 7+ isomerization at optimum operating conditions will provide the necessary branched paraffins. The ability to revamp existing extra reformers to isomerization. Flexibility in blending streams at the desired levels for gasoline. ACKNOWLEDGMENTS The authors would like to thank the management of Saudi Aramco for their support and permission to publish this article. REFERENCES 1. Ercan, C., Dossary, M. and Wang, Y.: Gasoline Specifications: Historical Trend & the Future, Focus, Fall 2009, pp. 12. 2. Ercan, C.: C 7 C 10 n-paraffin Isomerization, presentation at R&DC, Saudi Aramco, May 28, 2011. 3. Wang, Y., Dossary, M., Sameer, G. and Ercan, C.: Review FALL 2015 SAUDI ARAMCO JOURNAL OF TECHNOLOGY

of C 7 C 10 Paraffin Isomerization, Proceedings of the Future Challenge for Catalysis Workshop, King Fahd University of Petroleum and Minerals, Dhahran, Saudi Arabia, May 2009. 4. Ercan, C., Wang, Y., Dossary, M. and Othman, R.M.: Process Development by Parallel Operation of Paraffin Isomerization Unit with Reformer, U.S. Patent 8,808,534 B2, August 19, 2014. 5. Antos, G.J. and Aitani, A.M. (editors): Catalytic Naphtha Reforming, CRC Press, Boca Raton, Florida, 2004, 624 p. BIOGRAPHIES Dr. Cemal Ercan joined Saudi Aramco s Research & Development Center in 2005 and is now a member of the Oil & Gas Treatment R&D Division of the R&D Center. He has more than 25 years of experience in the oil and gas, and petrochemicals industries. Prior to joining the company, Cemal had worked for ABB Lummus Global, Bloomfield, NJ, as a Senior & Principal Process Development Engineer; for the Syntroleum Corporation in Tulsa, OK, as a Technical Manager; and for ConocoPhillips as a Chief Engineer. He received his B.S. and M.S. degrees from Middle East Technical University, Ankara, Turkey, and his Ph.D. degree from McGill University, Montreal, Quebec, Canada, all in Chemical Engineering. Dr. Yuguo Wang joined Saudi Aramco s Research & Development Center in 2006 and is now a member of the Oil & Gas Treatment R&D Division of the R&D Center. Previously, he had worked for the Center for Applied Energy Research and Fossil Fuel Science at the University of Kentucky, Lexington, KY, and as a Modeling Engineer at Cypress Semiconductors. Yuguo has more than 15 years of experience in the coal, and oil and gas industries. He received his B.S. degree in Physics from Shandong University, Shandong, China, and his Ph.D. degree in Chemistry from the University of Kentucky, Lexington, KY. Dr. Rashid M. Al-Othman is a Science Consultant in the Oil & Gas Treatment R&D Division of Saudi Aramco s Research & Development Center. He has 30 years of experience with Saudi Aramco, during which he has completed internal assignments in both the Ras Tanura Refinery and Shedgum Gas Plant and an external assignment at the French Institute for Petroleum. Rashid received his B.S. degree from Seattle University, Seattle, WA, and M.S. and Ph.D. degrees from the University of Washington, Seattle, WA, all in Chemistry. SAUDI ARAMCO JOURNAL OF TECHNOLOGY FALL 2015