~LZ~293 .
BACKGROUND OF THE INVENTION
This invention relates to catalytic hydrotreating or catalytic hydrocracking of hydrocarbon materials and, more particularly, to catalytic hydrocracking of heavy hydrocarbon oils, such as bitumen from tar sands.
Proven reserves of conventional crude oil supplies are expected to diminish significantly within the next two decades. As the available amount of relatively light crude oil decreases, other sources which can be used as raw materials to produce hydrocarbon fuels must be utilized. Initially there will be a tendency to use progressively heavier and heavier crude oils. These substances must be refined to a greater extent in order to make products which are equivalent to those which in the past have come from conventional light crude oils. For example, heavier crude oils tend to have higher concentrations of asphaltenes and also larger values of Conradson carbon residue, which is an indication of the coke forming tendency of a particular hydrocarbon material. Several changes must be made to these heavy crude oils in order to transform them into usable fuel products. The major change is the conversion of the higher boiling larger molecules into lower boiling smaller molecules, . . ~ . . -. . . . ` . . . .
. ~ . . ~: , : :
~ ' ''', '`' , :
, ` ` . ~ .
- 2 -..
There are several potential processes for this molecular weight reduction. For example, coking processes have been used to a considerable extent in the past. The :~ most undesirable feature of the coking processes is that large quantities of coke are produced as an unusable by-product. The coke originates in the asphaltenes and - coke-precursors contained in the original crude oil.
The heavier crude oils contain larger quantities of asphaltenes and coke precursors and, therefore, produce larger quantities of by-product coke. This coke is high in sulphur and also contains large concentrations of metals, such as nickel, vanadium and frequently iron.
Disposal of this coke is a major problem. ThUs, it cannot readily be used as a conventional fuel because the large quantities of sulphur in the coke necessitates the use of expensive stack gas scrubbing facilities to prevent excessive dioxide emissions to the atmosphere and consequent acid rains. This coke is also not well suited to the production of electrodes such as those used in the aluminum or related industries because the metal concentrations within the coke are unacceptably high~ The end result is that the coke by-product obtained from the coking processes is at best very difficult to dispose of.
It is, of course, highly desirable to be able to replace the coking processes with a process which is capable of converting all of the high molecular weight species in the heavy crude oils into usable liquid `
products. Hydrocracking is such a process and various configurations of this process have been used commercially. ~or example, there are fixed bed reactors in which a heavy oil feedstock and hydrogen are combined and passed through a vessel containing a fixed bed of catalyst. Alternatively, fluidized beds or ebullating beds of catalyst have been used in which the combination r of hydrogen and liquid hydrocarbon feedstock enters the bottom of the vessel where small sized catalyst particles are suspended in a fluidized state. Using the hydrocracking process, virtually all of the feedstock is converted into usable liquid fuels or into liquid products for which conventional refining technologies are available. However, the hydrocracking process also has one operational deficiency. When these heavy crude oils are processed in the presence of a catalyst, there is a tendency for large quantities of coke and also large concentrations of metals (which originate in the organo-metallic compounds in the heavy crude oil) to deposit on the catalyst surface. This catalyst deactivation has a profound influence on the operation of the process, as the conversion rate decreases substantially with time.
Initially, coke deposition is the predominant deactivation mechanism. After longer periods of time, both coke deposition and metals deposition are responsible for the deactivation of the catalyst.
Studies have shown that catalysts contain at least two kinds of coke. One type of coke acts as a reaction intermediate which is subsequently converted into reaction products. The other type of coke is an unreactive material which blocks catalytic sites and decreases activity of the catalyst. Ideally, the formation of coke which blocks catalytic sites should be minimized.
The commonly used hydrocracking catalysts consist of combinations of compounds containing group VI elements (chromium, molybdenum and tungsten), group VIII elements i`
(cobalt and nickel) and a catalyst support such as alumina, silica or chemical or physical combinations of silica-alumina. Such catalysts are expensive to produce and deactivate rapidly at hydrocracking conditions, making ;` them poorly suited to catalytic hydrotreating or Z~3 .- -- 4 hydrocracking of large volumes of many types of heavy hydrocarbon oils.
Other elements that have been used in catalysts for the conversion of hydrocarbons are the alkali metals and alkaline earth metals. For instance, S.P.S. Andrew I. &
E.C. Prod. Res. Dev. 8, 321 (1969) describes a steam reforming catalyst which is a nickel catalyst containing alkaline compounds as activators. In discussing the function of such catalyst, the article states ""the success of alkalized nickel reforming catalysts.O.is primarily due to its ability to increase the rate of removal of carbon residues from the nickel surface by steam gasification." This represents but one example and it has been known for many years that alkaline eompounds have the ability to eatalyze the steam-carbon reaction.
Another example of the use of an alkaline earth metal in a steam reforming catalyst is Canadian Patent 811,139 in which a large amount of alkaline earth metal is used in a nickel, alumina catalyst.
~UMMARY OF THE INVENTION
Aecording to the present invention, it has been quite surprisingly discovered that the alkali metals and alkaline earth metals when present in small quantities in the usual hydrocraeking eatalyst are highly effective in preventing the deposition of earbonaceous materials on the surface of the catalyst. Thus, the present invention in its broadest aspect relates to an improvement in the proeess for catalytically hydrotreating or catalytically hydroeracking a hydrocarbon material in which a hydrocarbon material and hydrogen are reacted under heat and pressure in the presence of a catalyst comprising at least one group VIb element and at least one group VIII
element on a silica, alumina or silica-alumina support, the improvement being the presenee of a small amount of at ~lZ~; 293 ;
least one group Ia element, whereby deposition of carbonaceous materials on the surface of the catalyst is substantially decreased.
As the group Ia elements there can be specifically mentioned lithium, sodium, potassium, rubidium, cesium and francium. The added materials interact with compounds in the hydrocracking catalyst to produce a decreased rate of formation of unreactive carbonaceous deposits which block catalytic sites. By having fewer unreactive carbonaceous deposits on the catalyst, it is possible for a higher prop-ortion of the reactive sites to participate in reactions.
One function of the group Ia elements may be that they affect the acid-base properties of the solid catalyst support. Frequently the materials which are used as ; 15 catalyst supports, such as silica, alumina or their mixtures, have acidic sites. Two types of acidic sites have been described theoretically. Bronsted acids are those which are hydrogen donors. Lewis acids are those which are electron pair acceptors. Hydrocracking and ; 20 hydrotreating catalysts of the type used in the present invention sometimes contain both types of acid sites.
. Adding the group Ia elements tends to make such catalyst ~` more basic. In principal this should involve the elimination of some of the acidic sites, particularly of the Lewis type. Any undesirable reactions which are promoted by acid sites, such as coke deposition, are believed to be hindered when the number of acid sites are decreased and the basicity of the catalyst is increased.
The process of this invention can be carried out with advantage ~or a variety of different purposes. For instance, a heavy crude oil may be hydrocracked with the object of reducing its molecular weight. .Other reactions, such as sulphur removal, nitrogen removal, oxygen removal, hydrogenation of aromatics, olefins and other unsaturated . ; ~ . ' !: ' ` , . ' ~ ' `
` ~``~ ' . ` ' ` , ' `. ' ' " ' ~' '` " `` `" , " . ` . ' compounds will occur simultaneously. The process may also be used to hydrotreat hydrocarbon feedstocks for sulphur removal, nitrogen removal or hydrogen addition without extensive molecular weight reduction.
Heavy hydrocarbon feedstocks which can be p~ocessed can include such materials as petroleum crude oil, atmosphere tar bottoms products, vacuum tar bottoms products, heavy cycle oils, shale oils, coal derived fluids, crude oil residuum, topped crude oils and the heavy bituminous oils extracted from tar sands. Of particular interest are the oils extracted from tar sands which contain wide boiling range materials from naphtha through kerosene, gas oil, pitch, etc., and which contain a large proportion, usually more than about 40 to 50~ by weight of materials boiling above 524C, equivalent atmospheric boiling point.
~n a typical process in accordance with the invention, the feedstock is heated to about 50-400C, preferably 75-250C, and mixed with hydrogen. After the hydrogen addition, the mixture is heated further to about 250-550C, preferably 375-~75C. This oil-hydrogen mixture enters the reactor and flows through a catalyst bed. This catalyst bed can be either a fixed bed or a fluidized or ebullated bed. The hydrogen gas and oil feedstock contact the catalyst in such a way that both reactants are present and available and the reaction can occur. The reaction is preferably carried out at a pressure of up to 3000 psig and a hydrogen partial pressure in the range of 1 atm to 1000 psig. The liquid space velocity is conveniently in the range 0.1 hr 1 to 35 hr 1, preferably 0.5 hr 1 to 20 hr 1.
Reaction pressures used are a function of the properties of the feedstock, as characterized by molecular type and boiling point range. Thus, for hydrocarbon ` ` . ' ~ ' '.
. .
feedstocks containing components boiling above 525C, a pressure of 500-3000 psig is preferred. For gas oil feedstocks containing components boiling between 200-525C, a pressure of 50-1500 psig is preferred, while for naphtha feedstocks having components boiling between room temperature and 250C a pressure of 20-1000 psig is preferred. The reactor can be a downflow trickle bed reactor or an upflow fixed bed reactor. Alternatively, it may be a fluidized bed or ebullated bed reactor. After the products leave the reaction vessel they are separated.
Either a single separation vessel or a series of separation vessels may be used. When a series of separation vessels are used, subsequent vessels are maintained at decreasing pressures. The vapours leave through the top of the separation vessel and the liquid ~ through the bottom. A hydrogen rich vapour stream can be - recycled to the reactor after removing some of the H2S~
acid gases and light hydrocarbon gases.
The liquid product obtained can be used in a number of different ways. Thus, the total liquid product may be fed directly into a pipeline. Alternatively, the liquid ; product can be separated by fractionation, distillation or other means and the resulting streams can be hydrotreated separately. The highest boiling material, e.g. that ~oiling above about 525C, may be used on the site to provide energy for the process. Alternatively, it may be gasified and the gasification product may be used as a source of process energy.
The alkali metal or alkaline earth metal compounds used in the catalyst may conveniently be in the form of an oxide, hydroxide or carbonate. ~ typical catalyst may contain 0.5-35% by weight of a group VI element, - preferably molybdenum or tungsten. It may also contain up to 10% by weight, preferably 0.5-6~ by weight of a group - ~:
. . . ~,~ ,, - , : , 1 : ' ; .. ~ .
VIII elementr preferably cobalt or nickel oxide. The amount of group Ia element in the catalyst is usually based on the amount of group VI element present and is usually in an atomic ratio of 0.1:10, preferably 0.3:4~ to the group VI elements.
The catalyst used in this invention may have a variety of geometries. The catalyst typically have pore sizes, from 2 to 1000 nm, preferably 8 to 60 nm and a surface area greater than 50 m2/g. Small pore zeolites, such as those having pore sizes smaller than 2 nm,are not normally utilized in the present invention. The reason for this is that the pores in the zeolites are so small that they do not permit the entrance of the high molecular weight hydrocarbon species involved in the hydrocracking or hydrotreating reactions of this invention. The catalyst supports are of the amorphous type and are not crystalline when they contain pores of the sizes utilized in the present invention. These pore sizes have several effects on the reaction rate. In general, the larger the pore size the laeger the size of the molecules which can enter the pore and therefore react within the interior of the catalyst. If the pores are relatively small, for example, 5-6 nm, some of the larger molecules are unable to enter the pores. In other cases where the hydrocarbon molecules a~e small enough to enter the pores, the rate of diffusion of the molecule within the pore is quite slow. Larger molecules quite often react predominently within the surface region of pellets or extrudates of the catalyst.
On the other hand, if the catalyst contains pores of larger sizes, for example 40 to 60 nm, the large organo-metallic species are able to re-enter the pores and penetrate throughout the catalyst form. When heavy crude oils are processed, the metals are neither deposited at the pore mouth exclusively nor predominantly within the , ` "; ~ :
: , .
LZ~ 3 exterior shell of the catalyst shape. Instead, they are distributed more evenly throughout the entire catalyst.
This means that the reaction rate tends to be enhanced as the metal deposits are uniform and therefore the whole catalyst can be used. The catalyst life is extended and the catalyst can be used for longer periods of time. When distillate fractions are hydrotreated in catalysts having large pores, the rates of molecular diffusion are higher than in small pore catalysts. As a result, the reactants diffuse further toward the center of the catalyst particle so that reaction is not restricted to a thin shell on the exterior of the catalyst particle.
Unfortunately, the effect of catalyst size is not single valued. By making the catalyst pores larger it is possible for large molecules to go to the interior of the catalyst. Some of these large molecules are coke precursors and as the pores increase in size, the amount of coke per unit surface area of catalyst increases. This tends to increase the rate at which the catalyst deactivates.
The catalysts used in the process of this invention can be prepared in a variety of different ways. For example, a calcined catalyst support may be impregnated by solutions contalning the desired components. As an example, gamma alumina can be impregnated with solutions containing cobalt ions, molybdate ions and group I metal ions. The resulting material is then calcined again to ` produce the catalyst product. According to another procedure, the catalyst may be prepared by adding compounds containing the desired components into a gel composed of water and the material which would become the catalyst support. When the gel is dried and calcined, the excess water is driven off and the metals converted to the corresponding oxides. They may also be prepared by ~;
. ` .
:.
;'` ',, `'. '' :' . : i :, :' ~ ~ Z~ 3 coprecipitation or mechanical mixing of components. These all represent well known techniques in the industry.
Ceetain preferred embodiments of this invention will now be further illustrated by the following non-limitative examples. The results are shown on the attached drawings in which:
Figure 1 is a series of plots showing +525C pitch, oxygen, sulphur and nitrogen conversions v. lithium to molybdenum atomic ratio;
Figure 2 is a series of plots showing surface area, hydrogen to carbon ratio and coke content v. lithium to molybdenum atomic ratio;
Figure 3 is a series of plots showing ~525C pitch, oxygen, sulphur and nitrogen conversions v. sodium to molybdenum atomic ratio;
Eigure 4 is a series of plots showing surface area, hydrogen to carbon ratio and coke content v. sodium to molybdenum atomic ratio;
Figure 5 is a series of plots showing +525C pitch, oxygen, sulphur and nitrogen conversions v. potassium to molybdenum atom ratios;
~igure 6 is a series of plots showing surface area, hydrogen to carbon ratio and coke content v. potassium to molybdenum atom ratio; and Figure 7 is a series of plots showing specific gravity, sulphur content and nitrogen content v. time on stream.
Example 1 A series of catalyst samples were prepared, each ` sample having a base consisting of 5 kg o~ alpha~alumina monohydrate powder (a mixture of 20% by weight of Conoco Catapal~ SB and 80% by weight of Catapal~N). This powder was placed in a mix muller and solutions of cobalt nitrate and ammonium paramolybdate were mixed into the powder such that when calcined it contained 2.2% by weight . ` .
: :' of the molybenum trioxide and 1.1% by weight of the cobalt oxide. The group la element was added to the catalyst in the form of a solution of lithium carbonate. Four catalyst samples were prepared in the above manner having the same amounts of cobaltand molybdenum but with increasing amounts of lithium carbonate at atomic ratios of lithium to molybdenum of 1, 1.8, 4.2 and 8.5. These were compaeed with one catalyst which did not contain a group Ia element.
These materials were mixed in the muller and a small amount of stearic acid was added before the mulled powder was extruded into 3.18 mm diameter extrudates. These extrudates were dried at 110C overnight and then calcined at 500C for approximately 8 hours.
The samples of the catalysts were analyzed for cobalt and molybdenum via a standard atomic absorption technique. The lithium concentrations were determined by a standard flame photometer approach after dissolving the extrudates in hydrochloric acid.
As a feedstock for hydrocracking there was used an " Athabasca bitumen obtained the from ~reat Canadian Oil Sands Ltd. at ~ort McMurray, Alberta. This bitumen had the general properties shown in Table 1 below _ General ProPerties of At~ !abaSca Bi tumen Specific Gravity, 16/16C 1.009 As`h (wt.~) 700C 0.59 Iron (ppm) 358 Nickel (ppm) 67 Vanadium (ppm) 213 Conradson Carbon Residue (wt.%) 13.3 Pentane Insolubles (wt.%) 15 5 Benzene Insolubles (wt.%) 0 72 Sulphur (wt.%) 4.48 Nitrogen (wt.%) 0.43 Oxygen (wt.~) 0.95 Carbon (wt.%) 83.36 Hydrogen (wt.%) 10 52 +525C Residuum twt.%) 48 03 : . _ . .. ___ .__ I
- . . . . .
.
25~3 Catalyst evaluation experiments were performed in a fixed-bed reactor system using the above feedstock. The reactor was filled completely with extrudates and the bitumen, mixed with hydrogen (purity 99.9 wt.%), flowed into the bottom of the reactor as a bottom feed. Each ; catalyst sample was evaluated at a pressure of 13.9 MPa and a temperature of 450C with a liquid volumetric space velocity of 0.29 ks based on the total empty reactor volume. The hydrogen flow rate was set at 43.1 ml/s at S.T.P. The reactor was kept at steady state for one hour preceding and for two hours during the liquid product collection period. Prior to the evaluation experiments, each catalyst was presulphided in the presence of bitumen and hydrogen for approximately 14.4 ks. The feedstock and liquid products were analyzed for carbon, hydrogen, sulphur, nitrogen, oxygen and the fraction boiling above 525C. Hydrogen and carbon elemental analysis were performed using a Perkin Elmer model 240 analyzer. Oxygen was determined via neutron activation analysis using a neutron generator.
The coke concentration was determined on both a used catalyst and on a ~resh catalyst which acted as a ` reference. The fresh reference catalyst was saturated with the appropriately matched liquid product obtained from the corresponding catalyst evaluation experiment.
During saturation, the fresh reference catalyst was evacuated and then impregnated with liquid reaction product for approximately 0.9 ks at 55.3 MPa and at room temperature. Paired catalyst samples (i.e. reference and used) were then deoiled in a flowing hydrogen stream (7.29 x 105 m3/s at S.T.P.) while the temperature was raised over a one hour period to 600C. The samples were kept at 600C for 0.9 ks then allowed to cool to 200C in the flowing hydrogen stream. The paired catalysts were then . . .
. .
~L~2~293 weighed and left overnight in a mufEle furnace at 600C.
The change in weight after oxidation in the muffle furnace was ascribed to the coke being removed from the catalysts. Each of the weight changes was expressed as a percent of the final catalyst weight (i.e. coke free catalyst basis). The amount of coke on the fresh reference catalyst was ascribed to coke formed during deoiling from the hydrocarbon product adhering to the catalyst. The amount of coke on the reference catalyst was subtracted from the amount of coke on the used catalyst to obtain the coke values reported. This empirical definition of coke is used to represent the amount of coke on the used catalyst after removal from the reactor but prior to the deoiling procedure.
The results of these tests are shown in Figures 1 and 2 of the drawings. In ~igure 1 the conversion results are shown in solid lines and solid circles, while the amounts of reaction per unit surface area are shown by the dotted lines and open squares for catalysts containing lithium.
The conversion results generally go through a maximum as additional lithium is added to the catalyst. The amount - of reaction per unit area went through a maximum as lithium was added to the catalyst.
Figure 2 shows the surface area, H/C ratio in the liquid, and coke content of catalysts containing lithium.
The surface area of the fresh catalyst and hydrogen to carbon ratio in the liquid product go through a maximum and then decrease with increasing lithium content of the catalyst. The coke content goes through a minimum and then increases with increasing lithium content of the catalyst. The decreased quantities of coke on catalysts containing small quantities of lithium illustrate the object of the present invention.
.
: -:, . . ,:
~; ,~ ` ' , ,, . . , ~L~Z1293 Example 2 A new series of catalyst samples were prepared using the same procedures and materials as in Example 1 except that the lithium carbonate was replaced by sodium carbonate.
The same feedstock was used as in Example l and the same series of catalyst evaluation experiments were carfied out in the fixed bed reactor.
The results of these tests were shown in Figures 3 and 4, with Eigure 3 showing the conversion results as solid lines and solid circles and the amounts of reaction per unit surface area as dotted lines and open squares for the catalysts containing sodium. In this case, all the conversion results decreased as the amount of sodium in the catalyst increased. On a unit surface area basis, the amounts of pitch removed and oxygen removed increased slightly with increasing concentration of sodium in the catalyst.
Eigure 4 shows the surface area, H/C ratio in the ` 20 liquid product and coke content of catalysts containing sodium. All the quantities decreased with the addition of sodium and the decreased coke content showed that there -are smaller quantities of carbonaceous deposits on the catalyst.
Example 3 The procedures of Example 1 were repeated once again ` and catalyst samples were once more prepared with the only change being the replacement of lithium carbonate with ~- potassium carbonate. The catalyst evaluation experiments were repeated again in the fixed bed reactor using the ~ same feedstock and conditions as in Example l. The :~ results obtained are shown in Eigures 5 and 6 of the ~` drawings.
~ .
, , : ' : `' ' ;:
Figure 5 shows the conversion results as solid lines and solid circles and the amounts of reactions per unit surface area as dotted lines and open squares for the catalysts containing potassium. All of the conversion results decrease as the amount of potassium in the catalyst increases and on a unit surface area basis, the amounts of pitch removed and oxygen removed increased slightly with increasing concentration of potassium in the catalyst. The amounts of sulphur removed and nitrogen ` 10 removed per unit area generally decreased with increasing potassium content in the catalyst.
Figure 6 shows the surface area, H/C ratio in the ` liquid product, and coke content of the catalyst containing potassium. All three quantities decreased with the addition of potassium. The decreased coke content shows that there are smaller quantities of carbonaceous deposits in the catalyst of this invention.
, , " , - . . : , '