Patent Application
20160046881
This invention relates to a method of using thermally cracked naphtha feedstocks effectively in the refinery gasoline pool by catalytic desulfurization in a naphtha blend.
A large proportion of the gasoline pool in the United States, Europe and elsewhere is filled by naphtha from fluid catalytic cracking units. While this naphtha has good octane as a result of a relatively high olefin content resulting from the catalytic cracking process, it also tends to have an unacceptably high sulfur content under the regulatory standards for motor gasoline, the United States currently sets a standard with a maximum of 30 ppm sulfur but in 2017 the Tier 3 standard of not more than 10 ppm which has been the European standard since 2009. Over the same time period, the progressive decreases in permitted sulfur content have been accompanied by a requirement to reduce mobile source emissions, especially carbon monoxide, by the addition of oxygenates, usually provided in the form of bioethanol since the use of ether type oxygenates such as MTBE became controversial in the US and legislation favoring ethanol was became effective.
Ethanol has a high octane number, commonly reported as about 108 RON (Research Octane Number) although some reports from ethanol manufacturers place it as high as 113 RON although not specifying the MON (Motor Octane Number). However, it is accepted that blending ethanol with gasoline will have a favorable effect on the RON so that there is usually no difficulty for the refiner to meet the PON (Pump Octane Number) requirements of most vehicles with ethanol blends. The octane boost provided by the use of ethanol in gasoline blends, typically the E10 blend with 10 percent ethanol has provided an opportunity to bring feedstocks previously considered of marginal acceptability into the refinery pool for processing into motor gasoline.
Coker naphtha, being derived by the thermal cracking of residual feedstocks is generally high in sulfur as well as olefins although its octane is typically low. Retention of olefins is therefore necessary if octane targets in the gasoline pool are to be met; in addition, sulfur must be significantly reduced to meet current and future sulfur limits in the gasoline pool.
A number of sulfur reduction techniques for gasoline blend components have been developed, summarized briefly in U.S. Pat. No. 7,837,861. The ExxonMobil selective naphtha naphtha hydrofining process, SCANfining™, developed for deep hydrodesulfurization with maximum preservation of the olefins (octane), which is commercially available under license from ExxonMobil Research and Engineering Company, is a very effective naphtha desulfurization process which incorporates aspects of the processes described in the following patents: U.S. Pat. No. 5,985,136; U.S. Pat. Nos. 6,231,753; 6,409,913; U.S. Pat. No. 6,231,754; U.S. Pat. No. 6,013,598; U.S. Pat. No. 6,387,249 and U.S. Pat. No. 6,596,157. The single stage version of the process can be used with a full range catalytic naphtha or with an intermediate catalytic naphtha (ICN), for example a nominal 65-175° C. (150-350° F.) or a heavy catalytic naphtha (HCN), for example, a nominal 175° C.+(350° F.+) naphtha, or both. The two-stage version of the process, as described in U.S. Pat. No. 6,231,753, WO 03/048273 and WO 03/099963, adds a second reactor and inter-stage removal of H2S allowing very deep HDS with very good olefin retention. The operation of this process relies on a combination of a highly selective catalyst with process conditions designed to achieve hydrodesulfurization with minimum olefin saturation.
While the octane retention afforded by the SCANfining process would be useful if the process were applied to the desulfurization of the olefinic coker naphthas, the difficulty encountered in many cases results from the fact that coker naphthas frequently contain silicon which has been found to have a deleterious effect on the SCANfining catalyst. This silicon, in various combined forms, is generated from the silicone oil anti-foaming additives typically based on polydimethylsiloxane, used in the delayed coking process to suppress foaming caused by gas formation; pilot plant testing has shown that anywhere from 25 to 75% of the silicon from the anti-foamant will be found in the naphtha draw. In fact, it has been reported that 70-80% of the silicon at the coker ends up in the coker naphtha fraction. The silicone oils decompose in the coker to form modified silica gels and silicone fragments. The fragments are typically in the naphtha boiling range though some will boil at higher temperatures. The deactivating silica gels which have been found on catalyst surfaces have been primarily SiO2, Si(OSi)4, Si(OSi)3OH and methylated forms such as ═Si(CH3)2.
In hydroprocessing units downstream of the coker fractionator, silicon-containing fragments from the antifoam agent undergo a condensation reaction with alumina surfaces of the catalyst, forming a strong chemical bond. Once the silicon is bound to the alumina surface, it cannot be removed by regeneration or other means. It is a more moderate poison compared to contaminants like sodium or arsenic, but it nonetheless results in activity loss of the order of 5-10° F. (3-6° C.) for each 1.0 wt % Si deposited on the catalyst.
For these reasons, coker naphthas coming from the delayed coking process have not usually been considered suitable feedstocks for the process even though their desulfurization in this way and subsequent incorporation into the gasoline pool would be desirable. While alternative desulfurization processes for coker naphthas are, of course, at hand, they are generally hydrogenative with a high hydrogen consumption resulting from olefin saturation. Thus, the refiner is faced with a true dilemma, he may use SCANfining and retain olefins but inhibit the catalyst or he may use other hydrodesulfurization methods and spend money on hydrogen. Neither option is attractive.
We have now found that it is possible to process coker naphthas as a naphtha blend component in the feedstocks for olefin retentive selective catalytic naphtha hydrodesulfurization processes if a suitably selected silicon trap is used. To minimize the octane loss which occurs in the process a high surface area inert alumina with a controlled low metal content is required. In particular, the levels of metals with hydrogenation capability such as the base metals Ni, Co and Mo should be held at low levels and preferably should be absent.
According to the present invention the delayed coker naphtha is desulfurized in a catalytic naphtha desulfurization process in which the feed comprising the delayed coker naphtha is passed over a silicon trap comprising a high surface area inert alumina of low metals content prior to being hydrodesulfurized in an olefin-retentive, catalytic naphtha hydrodesulfurization process. Unpromoted (no intentional metals content), inert alumina is preferred since it will not affect the olefin-retentive qualities of the hydrodesulfurization catalyst and for maximal silica capture, a high surface area alumina is employed.
The feed to the catalytic naphtha hydrodesulfurization process is a blend of FCC naphtha and coker naphtha. Generally, the coker naphtha is the minority component, present in any amount only up to about 20% v/v and more commonly not more than 10% v/v but at least 5% v/v will provide economic benefit. API gravity of the blend will normally be at least 50, preferably 55 or high, e.g. 58, 60 or higher. Bromine number, representative of olefin content, will typically from 50 to 70, in most cases about 60. RON will be at least 90, e.g. 92 or 93, but the addition of the coker naphtha will not usually result in a high MON which will in most cases be about 80 or slightly higher, e.g. 81, 82 or 83. End point will normally be in the range 200-250 C, usually 210-240 C. T10 (D86) will normally be determined by fractionation settings but T50 will typically be in the range of 75-90 C.
A typical feed for the process is shown in Table 1 below.
Silicon (Si) poisons catalyst by bonding with the aluminum carrier and other acidic sites forming silica (SiO2). The silicon blocks the pore system of the catalyst and therefore, access to the active sites, reducing both the activity of the catalyst and its ability to absorb other poisons such as arsenic. The reaction is quick and diffusion limited so smaller catalysts absorb more silicon because their pore system is more accessible. A corollary to this finding is that the silicon trap should have a large surface area: the silicon species from the silicone additives will form a mono-layer on the catalyst surface; higher surface area catalysts therefore have higher silicon absorption capacities. High surface area of at least 250 m2/g is preferred and at least 300 m2/g is better with optimal surface areas being at least 350 m2/g. While the pore diameter of the alumina may have a role in the silicon retention as a result of limited diffusion, the diffusion limits of larger pores may be relaxed relative to smaller pores, so increasing silicon pickup; pore diameter does not appear to have a strong impact on silicon capacity.
Catalyst particle size can impact silicon removal. Small catalyst particles (less than 1/16 inch, about 1.6 mm, preferably not more than 1/20 inch, about 1.25 mm) are preferred to maximize silicon pickup as silicon removal becomes diffusion limited when using large size catalyst. Table 1 below illustrates the extent of silicon pickup which can be achieved with various alumina trap materials.
While silicon trap materials incorporating nickel and cobalt have been recommended for use in hydrotreating application, their use is not favored in the present process: the feed is passed through the silicon trap bed in the presence of hydrogen and any catalytically active metals present on the alumina base tend to catalyze olefin hydrogenation reactions which are not desired since an objective of the olefin-retentive desulfurization process is to retain olefin octane values to the extent possible consistent with the desired degree of desulfurization. In addition, metals on the trap material catalysts inhibit silicon absorption because the active metals (Co, Ni, Mo) bond with the alumina and reduce the surface area available for bonding with the silicon. The concentration of base metals with hydrogenation activity should be low enough to minimize olefin saturation that can result in an increased octane loss. The metal loadings for the common active base metals for this type of silicon trap catalyst should be Ni or Co content is less than 1 wt % of the finished catalyst and Mo content is less than 5 wt % of the finished catalyst.
The silicon trap will normally be loaded at the top bed in the hydrodesulfurization reactor and operated at the temperature appropriate to the HDS catalyst, as described below. Space velocity can be relatively high since the silicon/alumina reaction is fast, enabling guard bed volume to be limited. Space velocities of 5 or higher, e.g. 6 hr−1 LHSV are suitable.
Silicon contaminated guard bed catalysts are usually not regenerable. The silicon on the catalyst will turn into glass (SiO2) during regeneration coating the catalyst particle and blocking the pore system and active sites. Since the silicon deposits at the top of the guard bed first, catalyst from the bottom beds may be regenerable. If arsenic or other poisons are present in the feed, the dominant poison should be removed first. Arsenic traps may be located in the DIOS reactor or in the desulfurization reactor.
In addition to silicon-containing fragments, coker naphthas typically have a relatively high level of diolefins, including conjugated diolefins which adversely affect hydrodesulfurization catalysts by polymerizing at bed temperatures and plugging the bed. A diolefin saturation (DIOS) reactor should precede the hydrodesulfurization reactor to stabilize the feed. Diolefin saturation of is a fast reaction and can therefore be carried out at high LHSVs and at low temperatures. Saturation of the troublesome conjugated diolefins can be carried out with the DIOS unit operated at a relatively low temperature in the presence of hydrogen. Using a Group VI/VIII catalyst such as Ni/Mo, Co/Mo at a temperature of about 125-175 C, preferably about 140-160 C; the exotherm across the DIOS catalyst bed in usually quite low, no more than 10 C or less. Pressure is typically about 20 Barg with a space velocity of about 3-4, normally about 4 hr.−1 LHSV.
The hydrodesulfurization (HDS) of the naphtha feed is carried out in a process which in which sulfur is hydrogenatively removed while retaining olefins to the extent feasible. The HDS conditions needed to produce a hydrotreated naphtha stream which contains non-mercaptan sulfur at a level below the mogas specification as well as significant amounts of mercaptan sulfur will vary as a function of the concentration of sulfur and types of organic sulfur in the cracked naphtha feed to the HDS unit. Generally, the processing conditions will fall within the following ranges: 250-325° C. (about 475-620° F.), 1000-3500 kPag (about 150-500 psig) total pressure, 600-2500 kPa (about 90-350 psig kPa) hydrogen partial pressure, 200-300 Nm3/m3 hydrogen treat gas rate, and 1-10 hr.−1 LHSV. The olefin-retentive selective catalytic naphtha hydrotreating process available under license from ExxonMobil Research and Engineering Company as the SCANfining process is preferred for this purpose as it optimizes desulfurization and denitrogenation while retaining olefins for gasoline octane. SCANfining is described in National Petroleum Refiners Association Paper AM-99-31 titled “Selective Cat Naphtha Hydrofining with Minimal Octane Loss”. Other olefin retentive selective catalytic naphtha hydrodesulfurization processes such as those described in U.S. Pat. No. 5,853,570; 5,906,730; U.S. Pat. No. 4,243,519; U.S. Pat. No. 4,131,537; U.S. Pat. No. 5,985,136 and U.S. Pat. No. 6,013,598 (to which reference is made for descriptions of such processes).
Typical SCANfining conditions include one and two stage processes for hydrodesulfurizing a naphtha feedstock comprising reacting the feedstock in a first reaction stage under hydrodesulfurization conditions in contact with a catalyst comprised of about 1 to 10 wt. % MoO3; and about 0.1 to 5 wt. % CoO; and a Co/Mo atomic ratio of about 0.1 to 1.0; and a median pore diameter of about 6 to 20 nm; and a Mo03 surface concentration in g Mo03/m2 of about 0.5×104 to 3×104; and an average particle size diameter of less than about 2.0 mm. The reaction product of the first stage may then be optionally passed to a second stage, also operated under hydrodesulfurization conditions, and in contact with a catalyst comprised of at least one Group VIII metal selected from Co and Ni, and at least one Group VI metal selected from Mo and W, preferably Mo, on an inorganic oxide support material such as alumina. The preferred catalyst is the Albemarle Catalyst RT-235.
In a preferred two-stage SCANfining process configuration, typical process conditions will contact the naphtha with hydrogen over the first hydrotreating catalyst in the vapor phase to remove at least 70 wt. % of the sulfur, to produce a first stage effluent which is cooled to condense the naphtha vapor to liquid which contains dissolved H2S which is then separated from the H2S containing gas. The first stage naphtha reduced in H2S is then passed with hydrogen treat gas into the second vapor phase stage in the presence of a hydrodesulfurization catalyst at a temperature at least 10 C (about 20° F.) greater than in the first stage and at a space velocity at least 1.5 times greater than in the first stage, to remove at least 80 wt. % of the remaining sulfur from the naphtha and form a desulfurized naphtha vapor. The second stage vapor effluent is then cooled to condense and separate the naphtha from the H2S to form a desulfurized naphtha product liquid which contains less than 5 wt. % of the amount of the sulfur present in the feed but retaining at least 40 vol. % of the olefin content of the feed. In this configuration, the catalyst in both stages comprising Co and Mo on a support and present in an amount of less than a total of 12 wt. % calculated as the respective metal oxides CoO and MoO3 with a Co to Mo atomic ratio from 0.1 to 1.0. Reaction conditions in each stage normally range from 230-400 C (about 450-750° F.), a pressure of from 400-34000 kPag (about 60-600 psig), a treat gas ratio of from 1000-4000 scf/b and a space velocity of from 1-10 v/v/hr; under these conditions, the percent desulfurization in the second stage is typically at least 90%. Space velocity in the second will normally be greater than that in the first stage and can range up to 6 hr.−1 LHSV.
During hydrogenative sulfur removal processes such as SCANfining, the organic sulfur is converted to inorganic form as hydrogen sulfide (H2S) but some mercaptans may remain in the product stream. A subsequent mercaptan decomposition stage may be included to reduce the content of mercaptan sulfur in the final product stream. Suitable mercaptan removal processes are described in US 2007/114156 and US 2014/174982.
The sulfur content of the desulfurized product will be substantially reduced but it may not comply itself as finished gasoline; it is however, satisfactory as a component in the gasoline pool when blended with other gasoline range fractions such as straight run naphtha, alkylate, reformate etc to make a blend for ethanol blending at the product terminal. Typical product quality is shown in Table 4 below.
