EP1171486A1

EP1171486A1 - Process for changing between incompatible polymerization catalysts

Info

Publication number: EP1171486A1
Application number: EP00919524A
Authority: EP
Inventors: William D. Bybee; Mitchell A. Krutilek; John F. Kominek
Original assignee: Chevron Phillips Chemical Co LLC; Chevron Phillips Chemical Co LP
Current assignee: Chevron Phillips Chemical Co LLC; Chevron Phillips Chemical Co LP
Priority date: 1999-03-31
Filing date: 2000-03-22
Publication date: 2002-01-16
Also published as: WO2000058377A1; JP2002540266A; CN1352657A; AU4019500A

Abstract

An improved process for changing between two incompatible polymerization catalysts in a commercial scale olefins polymerization process. More particularly, the first polymerization is stopped and substantially all the polymer is removed, followed by rapid purging with nitrogen and start-up with a seedbed of high melting polymer.

Description

PROCESS FOR CHANGING BETWEEN INCOMPATIBLE POLYMERIZATION CATALYSTS

FIELD OF INVENTION The process of the invention is related to olefins polymerization and the changeover between two incompatible polymerization catalysts. The process involves emptying the reactor as part of this changeover. For example, the first catalyst may be a Ziegler-Natta catalyst and the second catalyst may be a metallocene catalyst.

BACKGROUND AND RELEVANT REFERENCES From time to time during the production of olefin polymers in a commercial reactor, it is necessary to convert from one type of catalyst system producing polymers having certain properties and characteristics to another catalyst system capable of producing polymers with somewhat different properties and characteristics. However, this catalyst changeover is often problematic and can be quite time-consuming. For example Williams et al. in USP 4,460,755 teach that "when converting from a Ziegler-type catalyst to a chromium-based catalyst, the components of the Ziegler-Natta catalyst act as poisons for the chromium- based catalyst, and that the chromium-based catalyst is ineffective in promoting polymerization if it is introduced directly into the polymerization reactor in the presence of the Ziegler-Natta catalyst. This is true even if a large excess of the amount of chromium-based catalyst required to promote polymerization is employed." The Williams et al. patent teaches that "since direct conversion from the use of a Ziegler-Natta catalyst to a chromium-based catalyst had proven unsuccessful, such change of catalyst could heretofore be effected only "by halting the polymerization reaction completely, emptying the reactor to rid it of the original catalyst system, and then re-initiating polymerization with the new catalyst system." They also teach that "in order to reinitiate polymerization with the new catalyst, the recharged reactor first had to be purged to low levels of oxygen and water, scavengers had to be employed to remove other catalyst poisons, and prolonged induction periods had to be undergone after such catalyst was introduced into the reactor before polymerization would begin again. Consequently, such catalyst conversions were both time consuming and costly, ordinarily requiring about 4 days or more of reactor shutdown in a commercial operation before polymerization could once again be re-initiated with the new catalyst." Williams et al., solve this transition problem by using a hydroxyl-containing compound to interact with the Ziegler- Natta catalyst before the Cr catalyst is introduced. Similar problems exist when transitioning from Ziegler-Natta catalysts to metallocene catalysts. In part this is because trying to operate incompatible catalysts together leads to polymers with unacceptable performance properties. For example, in making a polyethylene with a Ml of 1.0, Ziegler-Natta catalysts require operating at high hydrogen concentrations (about 6 mole% in the reactor) while metallocene catalysts must be operated at low hydrogen concentrations (about 500 ppm). If the Ziegler-Natta catalysts are operated at low hydrogen, they produce very high molecular weight (MW) polymers. If the metallocene catalysts are operated at high hydrogen concentrations they produce low molecular weight polymers. Combining the two catalysts and operating at 500 ppm leads to a polymer that contains ultra-high MW chains and hard particles. Processing this polymer mix results in polyethylene films containing gels. The operating conditions for the two catalyst systems are incompatible. Similarly, the comonomer incorporation or uptake rates for the two catalyst systems are very different. The Ziegler-Natta catalyst requires a much higher concentration of comonomer in the feed to get the same degree of incorporation in the polymer chains. At conditions useful for the Ziegler-Natta catalyst, the metallocene catalyst produces very low-density polymer with lower sintering temperature, which may cause sheeting in the bed. Thus it is necessary to avoid the presence of both an active Ziegler-Natta catalyst and an active metallocene catalyst during polymerization. Agapiou et al. have patented a process for transitioning from one incompatible catalyst to another. (See USP 5,442,019; 5,672,665; 5,753,786 and 5,747,612 which are incorporated herein by reference in their entirety.) The process comprises the steps of: a) discontinuing the introduction of one of the incompatible catalysts or catalyst systems into a reactor; b) introducing an irreversible catalyst killer; and c) introducing into the reactor a second catalyst or catalyst system incompatible with the first catalyst system. In one embodiment the transition is between a traditional Ziegler-Natta type catalyst system and a metallocene-type catalyst. Although the Agapiou et al. method has some advantages, it also has some disadvantages. For example, having the polymer product from the first catalyst present during the transition results in product made from both catalysts; this mixture may have less than optimum polymer properties (e.g., MWD, density, etc.). Also, the first catalyst must be killed by titration. Too much kill agent will inactivate the second catalyst; too little will not meet the process objective. A serious disadvantage for the Agapiou et al. process lies in the fact that an overly active metallocene catalyst can generate hot spots and polymer sheets on startup, if the polymer bed is composed of LLDPE. Another disadvantage is that any reactor chunks or sheets that have formed and adhered to the reactor wall during operation of the first catalyst cannot be removed during the catalyst change, since the reactor is not opened. With the typically more active metallocene catalysts, failure to clean the grid and reactor walls 1-2 meters above the grid often leads to problems. These include faster grid plate and polymer withdrawal line plugging and wall fouling. Borealis in WO 9732905 also teaches a method for transitioning between two different catalysts in olefins polymerizations. The method comprises: (a) discontinuing the feed of the first catalyst into the polymerization reactor; and (b) introducing the second catalyst into the reactor. The transitioning is performed between a chromium-based catalyst and a metallocene catalyst. It is taught that the transition may be performed rapidly without emptying the reactor. The reaction may be stopped and the first catalyst deactivated, or a "flying" transition may be performed by stopping the feed of the first catalyst, adjusting the reactor conditions and starting the feed of the second catalyst. None of the above patents are concerned with ways to shorten the turnaround time for catalyst transitions where the reaction is stopped and polymer, along with catalyst, are removed from the reactor.

SUMMARY OF THE INVENTION In one embodiment the invention is an improved process for changing between two incompatible polymerization catalysts in a commercial-scale olefins polymerization process. The invention comprises: polymerizing olefins with a first polymerization catalyst; stopping the polymerization; removing substantially all of the polymer from the reactor, preferably including non-powder polymer chunks; rapidly purging with nitrogen; adding a seedbed of polymer particulates to the reactor, and polymerizing olefins with a second polymerization catalyst. We have now found that, for reactors having easy dump means, it is just as simple and time-efficient to dump the reactor contents as to use the poisoning techniques of the prior art. By the term "easy dump means" is meant a mechanism whereby the reactor contents can readily and quickly be removed. One such system with an easy dump means is shown in Figure 1. Here the reactor incorporates a fluidization grid plate with a central opening for polymer removal. It is preferred that the reactor also include a man-way that allows for inspection and cleanup of the grid plate. Preferably there is a vessel, e.g., a silo, external to the reactor to store a high melting polymer seedbed. Preferably, the vessel has a high enough pressure rating to allow using a high pressure N₂ line to quickly transfer the seedbed polymer into the reactor. The reactor system also includes a high-pressure nitrogen line to allow for rapidly purging the reactor. With this sort of reactor system, one can readily stop polymerization and without the use of catalyst kill agents, rapidly remove the olefin and polymer particulates containing catalyst. The gas circulation is stopped (i.e., the compressor is shut down) and the reactor is opened and inspected. After the polymer bed is removed, the reactor is opened. The reactor is then inspected for any strips or chunks on the ground and walls, which are removed if present. No special care is needed while inspecting the reactor since no poisons are present. The reactor is closed and a N₂ purge is then performed to remove any oxygen. Having high pressure N₂ and large diameter purge lines significantly speeds up this process. A drying agent, such as triethyl aluminum (TEA), is generally added at this point to dry the reactor. Then a seedbed of high melting polymer particulates, also called a startup bed, is rapidly added to the reactor. This high melting point polymer is selected to eliminate sheeting and chunking of the polymer during start-up. It minimizes the potential for overheating and melting when the catalyst lights off. Compared to using a low melting polymer, the use of a high melting polymer particulates allows the catalyst to be added more rapidly. In a preferred embodiment the invention is utilized with a continuous polymerization process to make polyethylene, more preferably linear low density polyethylene (LLDPE). Although the process is generally described and exemplified herein as pertaining to ethylene polymerizations, it should be understood that this is merely illustrative, and the invention encompasses polymerization of other olefin or combinations of olefins. When making LLDPE, the high melting polymer is preferably high density polyethylene (HDPE). HDPE has a higher melting point (i.e., higher sintering temperature) than the LLDPE. After the reactor is closed and a high melting polymer seedbed is added, reaction gases such as monomer, comonomer, and hydrogen, are then added and circulated. Polymerization is initiated by feeding the second catalyst and adjusting polymerization temperatures and pressures. An important aspect of the present invention is to do the whole process rapidly without fouling the reactor during startup with an overactive catalyst. Preferably the down time is less than 3 days, more preferably less than 2 days. In one preferred embodiment, the invention further comprises a transition between comonomers. For example, a transition between a non- polar and a polar comonomer coincides with the catalyst changeover. In a preferred embodiment the invention is a process for changing between two incompatible catalysts, a first catalyst and a second catalyst, during olefins polymerization, said process comprising: (a) operating an olefins polymerization plant which comprises a reactor system having a reactor, means for easy polymer removal and a means for polymer introduction, said plant utilizing a first catalyst at a first set of polymerization conditions; (b) stopping the polymerization; (c) removing substantially all of the polymer from the reactor system; (d) rapidly purging the reactor with nitrogen, preferably using a high pressure purge line; (e) drying the reactor; (f) adding a seedbed of high melting polymer particulates to the reactor; (g) adding reaction gases comprising monomer, comonomer (if desired), hydrogen, nitrogen, and an inert hydrocarbon (preferably n-pentane, isopentane, n-hexane); (h) adding said second polymerization catalyst to the reactor; (i) polymerizing one or more olefins at a second set of polymerization conditions. Preferably, the polymerization is stopped without poisoning the catalyst in the reactor. It is also preferred that any non-powder polymer chunks be removed as part of step (c), above. Moreover, it is also preferred that the process is completed in less than 3 days, more preferably in less than 2 days. As will be discussed below, in one embodiment, the invention further comprises a transition between comonomers, which coincides with the catalyst change. In step (g), preferably inert hydrocarbon and nitrogen are also included in the gases. Also, comonomer is preferably included if co-polymerization operation is desired. Among other factors, the present invention is based on our finding that catalyst changeovers can be accomplished without using catalyst kill agents using a rapid changeover procedure in accordance with the procedure described herein. Our time-saving process preferably uses high pressure nitrogen introduced via one or more supply lines, and uses rapid purging, preferably also in addition to using a seedbed of high melting polymer particulates for controlling polymerization light-off. The process of the present invention allows for rapid catalyst changeovers with short turnaround times, even though polymer is removed from the reactor. The present process also allows for reactor inspection and clean-up before polymerization with the second catalyst begins. BRIEF DESCRIPTION OF THE FIGURE The Figure is a schematic of a reactor system useful in the present invention.

DETAILED DESCRIPTION OF THE INVENTION In one broad aspect, the invention relates to changing between two incompatible catalysts in a commercial scale process operating in olefins polymerization plant. As used herein, "a commercial scale olefins polymerization process" is one that is primarily operated to produce polymer product for sale to customers, either internal or external customers. The process of the invention includes removing polymer from the reactor during the catalyst changeover. The term "olefins" is used herein to preferably include ethylene; propylene; butene-1 ; hexenes such as 4-methylpentene-1 and hexene-1 ; octene-1 ; and decene-1. The olefins polymerization may be a homopolymerization, such as ethylene polymerization, or a co-polymerization, such as ethylene plus butene, propylene, or other olefins such as mentioned above. In one especially preferred embodiment, the invention is a process for changing between two incompatible catalysts, a first catalyst and a second catalyst, during olefins polymerization, the process comprising: (a) operating an olefins polymerization plant which comprises a reactor system having a reactor, a grid plate, a means for easy polymer removal, and a means for polymer introduction, said plant utilizing said first catalyst at a first set of polymerization conditions; (b) stopping the polymerization without using a catalyst kill agent; (c) removing substantially all of the polymer from the reactor; (d) shutting down the compressor; (e) opening the reactor, and inspecting and removing non-powder chunks; (f) rapidly purging the reactor with nitrogen; (g) drying the reactor; (h) adding a HDPE seedbed that is at least 40% of the weight of the operational polymer bed to the reactor; (i) adding reaction gases comprisinging monomer, comonomer, hydrogen, nitrogen and an inert hydrocarbon; (j) adding said second polymerization catalyst to the reactor; (k) polymerizing one or more olefins at a second set of polymerization conditions to produce a polyethylene product. Preferably, the process is completed in less than 3 days, more preferably in less than 2 days. The process may further comprise a transition between comonomers. As used herein, the term "incompatible catalyst" is used in the same way as used in Agapiou et al., USP 5,753,786, bottom of Column 3. "Incompatible catalysts" are those that satisfy one or more of the following: (1) those catalysts that in each others presence reduce the activity of at least one of the catalysts by greater than 50%; (2) those catalysts such that under the same reactive conditions one of the catalysts produces polymers having a molecular weight greater than two times higher than any other catalyst in the system; and (3) those catalysts that differ in comonomer incorporation or reactivity ratio under the same conditions by more than about 30%. The process of this invention can be used in a gas phase, solution phase, slurry or bulk phase polymerization process. A gas phase polymerization process in a fluidized bed reactor is preferred, and may be operated in condensed mode or in 100% gas phase. In a typical continuous fluidized bed polymerization process for the production of polymer from monomer, a gaseous stream comprising monomer, with or without entrained liquids, is passed through a fluidized bed reactor in the presence of a catalyst under reactive conditions. A polymer product is withdrawn. Also withdrawn is a gas stream, which is recycled and usually cooled, and together with additional monomer sufficient to replace the monomer polymerized is returned to the reactor. In one embodiment, the recycle gas stream is cooled to form a gas and a liquid phase mixture that is then introduced into the reactor. For a detailed description of a gas phase process, see U.S. Pat. Nos. 4,543,399; 4,588,790 and 5,668,228 which are incorporated herein by reference. While the preferred embodiment of the process of the invention is directed to changing between a traditional Ziegler-Natta catalyst and a metallocene catalyst, it is within the scope of this invention that the process of the invention would apply to any change between incompatible catalysts. For example, changing between a traditional Ziegler-Natta catalyst and a chromium catalyst or changing between a chromium catalyst and a metallocene catalyst or even change between a traditional Ziegler-Natta titanium catalyst to a Ziegler-Natta vanadium catalyst. This invention contemplates that the direction of changing between incompatible catalysts is not limiting. However, it is preferred that the process of the invention involve a change from or to a metallocene catalyst. Traditional Ziegler-Natta catalysts typical in the art comprise a transition metal halide, such as titanium or vanadium halide, and an organometallic compound of a metal of Group 1 , 2 or 3, typically trialkylaluminum compounds. The latter serve as an activator for the transition metal halide. Some Ziegler-Natta catalyst systems incorporate an internal electron donor that is complexed to the alkyl aluminum or the transition metal. The transition metal halide may be supported on a magnesium halide or complexed thereto. This active Ziegler-Natta catalyst may also be impregnated onto an inorganic support such as silica or alumina. For more details on traditional Ziegler-Natta catalysts, see for example, U.S. Pat. Nos. 3,687,920, 4,086,408, 4,376,191 , 5,019,633, 4,482,687. 4,101 ,445, 4,560,671 , 4,719,193, 4,755,495, 5,070,055, all of which are herein incorporated by reference. Metallocene catalysts, for example, are typically those bulky ligand transition metal compounds derivable from the formula: [ ]_mM[A]_n[B]_p where L is a bulky ligand, A is an anionic ligand, B is a neutral donor ligand, p is zero, one, two or three, M is a transition metal, and m and n are such that the total ligand valency corresponds to the transition metal valency; m can be one, two, three or four, and n can be zero or the valency state of the transition metal minus m. The ligands L and A may be bridged to each other, and the ligands L and B may be bridged to each other, and if two ligands L and/or A are present, they may be bridged. A may be a halide, an alkyl, an amide, or phosphide, etc. B may be an oxygen donor such as tetrahydrofuran , pyridine, thiophene, an alkene, or a diolefin. The metallocene compound may be a full-sandwich compound having two or more ligands L, which may be cyclopentadienyl ligands or cyclopentadiene derived ligands, or half-sandwich compounds having one ligand L, which is a cyclopentadienyl ligand or a cyclopentadiene derived ligand. The metallocene compounds contain a multiplicity of bonded atoms, preferably carbon atoms, forming a group that can be cyclic. The bulky ligand can be a cyclopentadienyl ligand or cyclopentadienyl derived ligand, which can be mono- or poly-nuclear or any other ligand capable of .eta.-5 bonding to the transition metal. One or more bulky ligands may be .pi.-bonded to the transition metal atom. The transition metal atom may be a Group 4, 5 or 6 transition metal and/or a transition metal from the lanthanide and actinide series. Other ligands may be bonded to the transition metal, such as those indicated above. Non-limiting examples of metallocene catalysts and catalyst systems are discussed in for example, U.S. Pat. Nos. 4,871 ,705, 4,937,299, 5,017,714, 5,120,867, 5,057,475, 5,096,867, 5,055,438, 5,227,440, 5,153,157, 5,198,401 , 5,241 ,025, 4,530,914, 4,952,716, 5,064,802, 5,124,418, 4,808,561 , 4,897,455, 5,470,993 and 5,616,664 all of which are incorporated herein by reference. In the preferred embodiment, the metallocene catalyst of this invention is supported on support materials known in the art, for example, inorganic oxides, like silica, alumina or magnesia or polymeric, such as polyethylene. It is also contemplated by this invention that the metallocene catalyst of the invention can be supported on a single support or separately supporting the metallocene on one support and an activator on another support. In the present invention, the first polymerization catalyst may be any olefins polymerization catalyst including those described hereinabove. The second polymerization catalyst may also be any olefins polymerization catalyst including those described above, with the proviso that the second catalyst be incompatible with the first catalyst. The first and second catalysts may be added to the reactor system in various ways. U.S. Patent No. 5,028,669 to Rowley et al., which is incorporated herein by reference in its entirety, shows a variety of ways to add catalyst that can be advantageously employed in this invention. Many other ways known in the art for feeding particulates to a process stream will also work with the present invention. In a preferred embodiment, the catalyst is added to the reactor by a method described in U.S. Pat. No. 5,851 ,493 to Lawson et al., which is incorporated herein by reference in its entirety. This patent describes an injection system that includes a valve with at least one cavity for receipt of particulate, and a sweep stream source for providing a sweep stream to remove substantially all the particulate from the cavity of the valve. The invention utilizes a reactor system that includes a reactor, at least one heat exchanger, at least one compressor and associated piping. In a preferred embodiment, the reactor system further comprises a gas distribution grid plate. By way of example, a reactor system useful in this invention is shown schematically in FIG. 1. It includes a fluidized-bed reactor (1 ) consisting of a vertical cylinder (2) surmounted by an expanded chamber, also known as the disengaging section (3) and provided at its bottom part with a fluidization grid (4). The grid has small openings for distributing the entering reactant mixture and a larger opening (5) for polymer withdrawal. The reactor system also comprises a line (6) for recycling the reaction mixture, which joins the top of the fluidized-bed reactor to its bottom part. The recycling line (6) includes, in succession in the direction of flow of the gaseous reaction mixture, a cyclone (7), a first tube heat exchanger (8), a compressor (9) and a second tube heat exchanger (10). Ethylene and hydrogen are added through lines (11 ) and (12), respectively. Comonomer is added through line (13). Line (14) makes it possible to feed the reactor (1 ) with solid catalyst. Lines (24) and (25) are the feed lines for the different catalysts. Line (14) needs to be clear of the first catalyst before it is used to feed the second catalyst to the reactor. While on-stream the produced polyolefin polymer particulates (27) are discharged from the reactor (1 ) into line (15). During operation, fines are reinjected into the reactor via line (16). Line (23) is a purge line for removing gases from the reactor loop. As part of the catalyst changeover, residual polymer particulates are removed through grid withdrawal line (17). During operation, line (17) is nitrogen purged. After the compressor (9) is shut off the blinds are rolled for safety. Portal (18) is a man-way that may be opened to inspect the grid and reactor walls. Non-powder polymer chunks can be manually removed during inspection of the grid plate (4) after entry through the man-way (18). Residual particulates can also be vacuumed out during this inspection. The high- pressure nitrogen inlet line (19) is used to purge the reactor. High melting polymer particulates (20), which are blanketed with nitrogen, from the seedbed silo (21 ) are transferred into the reactor (1 ) through the seedbed inlet line (22). Line (26) is used to fill the seedbed silo with polymer. Line (27) represents the polymer in the bed. The polymerization may be stopped in a variety of ways, such as stopping catalyst feed, introducing a catalyst poison, or dropping the temperature or pressure or monomer concentration below the minimums necessary to sustain a polymerization reaction. It is preferred that the polymerization be stopped by stopping the addition of the first catalyst. After the polymerization is stopped, substantially all of the polymer is removed from the reactor. First powder is removed via line (15) and then via the grid opening (5). More preferably, before the compressor is turned off, the bed is maintained above 150°F and olefins are purged through purge line (23). It is preferred that the grid opening be located at the center of the grid plate. It is small relative to the size of the grid plate, for example, about 4 to 6 inches compared to about 16 feet. By the term "substantially all the polymer" is meant at least 90%, preferably at least 95%, more preferably at least 98%, and most preferably at least 99.5 wt. % of the polymer. One objective of the invention is to achieve a rapid turnaround time for the catalyst changeover. The turnaround time is the time between when the first catalyst is no longer added and the second catalyst lights off, i.e., begins sustained olefins polymerization. It is desirable that this is done in less than 3.5 days, preferably in less than 3 days, and more preferably in less than 2 days. As used herein, the term "rapid purge" or "rapidly purging" or the like are intended to mean the quick dilution of catalyst poisons in a minimum time. It is preferable to fill and empty the reactor several times to purge out any residual poisons such as oxygen or water to very low levels preferably below a few ppm. This rapid purging can be accomplished, for example, by using large diameter nitrogen lines and nozzles, or more preferably with high-pressure nitrogen lines supplied at pressures greater than 50 psig, for example at pressures of 100 psig and higher. Preferred pressures are between 125 and 400 psig, more preferably between about 150 and 350 psig, and most preferably at a pressure of between about 300 and 350 psig. A rapid purge which includes filling and emptying the reactor with nitrogen takes less than an hour, preferably less than about 40 minutes, more preferably less than about 30 minutes, and still more preferably less than about 20 minutes. Rapidly purging the reactor with nitrogen significantly reduces the turnaround time for the catalyst changeover. For the purposes of this application, rapid purge means a nitrogen addition rate of at least 340 moles per hour and the subsequent purging of at least that rate, more preferably at least 400 moles per hour, still more preferably at least 500 moles per hour, and most preferably at least 600 moles per hour. Temperature for purging operation is preferably between 0°F and 150°F, more preferably between 40°F and 85°F. For polymerization with Ziegler-Natta, metallocene or other water- sensitive catalysts, the water level in the nitrogen purge gas should be below 1 ppm. Thus, depending on the water content in the reactor, a small amount of a drying agent such as a trialkylaluminum compound may be added to reduce the water level to less than 1 ppm. This drying or "pickling" of the reactors and piping is common industry practice. Any of the well-known drying agents may be used in this step. Preferred drying agents include trialkyl aluminum compounds, such as trimethyl aluminum or more preferably triethyl aluminum (TEA). This portion of the start-up procedure may be accomplished by circulating the gases containing drying agent through the empty bed for about one hour, followed by about three low-pressure N₂ purges to facilitate removal of TEA and water reaction products. The reactor may then be repressured to 150 psig with N₂. Preferably, a high melting seedbed is introduced to reduce bed sheeting on startup after the LLDPE bed is withdrawn. Complete elimination of bed sheeting and chunking on startup has been observed by using, for example, a high melting polyethylene seedbed. It is important to use a high melting polymer as the seedbed, since low-melting polymers can cause problems on start-up. When polymerization is reinitiated, the temperature may increase above the target temperature, due to the rapidly accelerating reaction, and localized hot spots may be generated. This can soften or melt the seedbed polymer and block the grid plate openings or polymer withdrawal lines. For example, it has been found that using LLDPE as a seedbed polymer often results in operational problems. We have found that using a high melting polymer reduces bed sintering and prevents premature shutdown. We have also found that using HDPE of a high Ml, such as greater than 6 Ml, yields an initial product from the reactor that may be sold as a molding resin. It is not critical how the high melting polymer powder (also known as particulates or fluff) is transferred into the reactor, although additional contact with ambient air should be minimized. The powder can be transferred into the reactor by suction or aspiration from a storage silo. Preferably, the silo is appropriately pressure-rated and the polymer is transferred by blowing it into the reactor using a high pressure nitrogen line. The high melting polymer should not deactivate the polymerization catalyst. Any polymer melting above the melt point of LLDPE (above about 220°F) meeting these criteria can be used. HDPE is an especially preferred high melting polymer. For our purposes, HDPE having a density of >0.940 g/cc may be used. HDPE with a density of >0.950 g/cc is preferred, and Ml above 6 is preferred for the HDPE selected for this use. The amount of polymer seedbed should be sufficient to cover the fines re-injection inlet line. This amount is at least 40% of the weight of the operational polymer bed, more preferably at least 50%, and most preferably about 60%. As used herein, the term "operational polymer bed" means the amount of polymer particulates in the reactor when it is operated commercially. According to a preferred embodiment of the present invention, an antistatic agent is introduced to the reactor after the seedbed, such as HDPE seedbed is introduced, and prior to introducing the second catalyst, such as the metallocene catalyst. We have found that such use of antistatic agent is another factor that aids in successful transition and "light-off, including aiding in avoiding of sheeting and/or chunking. Preferred amounts of antistat agent are between about 1 ppm and 1000 ppm by weight based on polyethylene production, more preferably between 5 ppm and 500 ppm, and most preferably between 10 ppm and 80 ppm. Preferred antistat agents include those types as described in USP 5,283,278, which is incorporated herein by reference. The feed or gas in the reactors is adjusted so that the catalyst will produce a given product of a certain density and melt index. This generally depends on how well a catalyst incorporates comonomer and its rate of reaction with hydrogen. Typically, the gas composition also contains an amount of hydrogen to control the melt index of the polymer to be produced. In typical circumstances, the gas also contains an amount of dew point increasing component with the balance of the gas composition made up of a non-condensable inerts, for example, nitrogen. Generally, the gas contains at least one alpha-olefin having from 2 to 20 carbon atoms, preferably 2-15 carbon atoms. Examples of alpha-olefins include alpha-olefins of ethylene, propylene, butene-1 , pentene-1 , 4-methylpentene-1 , hexene-1 , octene-1 , decene-1 , and aromatic containing olefins such as styrene. Other monomers can include dienes, norbomene, acetylene, and polar vinyl compounds such as vinyl esters, vinyl ethers, vinyl halides, vinyl amines and vinyl sulfides. Vinyl acetate is a preferred polar comonomer. In the preferred embodiment, the gas composition contains ethylene and at least one alpha-olefin having 3 to 15 carbon atoms. It is also envisioned that a second, third or fourth comonomer may be added to make terpolymers, tetramers and quintamers. Although discussed herein in terms of catalyst changeovers, the advantages of rapid polymerization transitions achieved through polymerization shutdown and start-up can be applied to other polymerization problems. Thus the advantages associated with seedbeds made of high melting polymer particulates and the use of high-pressure nitrogen lines can also be used for transitioning between comonomers. These comonomers include non-polar comonomers such as alpha-olefins, i.e., transitions between butene and hexene, and especially between polar comonomers containing elements other than carbon and hydrogen, such oxygen, nitrogen, halogen or sulfur atoms, such as vinyl acetate and vinyl chloride, or between polar and non-polar comonomers. It is envisioned that the catalyst changeover process of the invention will, in some instances, also include a transition between comonomers. Depending on the second catalyst to be introduced into the reactor, the concentration of reactants, such as monomer, comonomer and hydrogen gas, can be increased or decreased. In the preferred embodiment, the concentrations of comonomer and hydrogen are decreased, particularly when a metallocene catalyst is utilized as the second catalyst in the process of the invention. Typically, the reactant gas composition is diluted as above, for example, by either pressure purging or flow purging, procedures well known in the art. Once the reactant concentrations are sufficiently diluted to accommodate the second catalyst and substantially all poisons are removed, the next step in the invention is to introduce the second catalyst. Once the bed is fluidized and the new gas composition is introduced into the reactor, the second catalyst is introduced into the reactor under reactive conditions. To obtain a more complete understanding of the present invention, the following examples illustrating certain aspects of the invention are set forth. It should be understood, however, that the invention is not intended to be limited in any way to the specific details of the following examples. EXAMPLES The following example and Table 1 illustrate the invention when making LLDPE in a gas phase reactor. Here the change is from a Ziegler- Natta catalyst to a metallocene catalyst. Example 1 uses an LLDPE seedbed (added by aspiration from a storage silo). It uses a conventional nitrogen supply line at 50 psig. In Example 2 of the invention, a high-pressure nitrogen purge system (at about 325 psig) is used with an HDPE seedbed, which is blown into the reactor using nitrogen pressure. The HDPE seedbed allows more rapid catalyst start-up since concerns about polymer sintering are reduced. The specifics for Example 2 are described in detail below. These two examples both include the same 21 steps. 1. Stop feeding the Ziegler-Natta catalyst. 2. Reduce the reactor pressure from about 325 to about 150 psig. 3. Draw down the polymer bed, thus also removing the bulk of the catalyst. Add nitrogen to displace the hydrocarbons. 4. Empty remaining polymer and catalyst from reactor through grid plate port. 5. Stop gas circulation (i.e., stop the compressor). 6. Close valves to block off feed lines and put blinds in place (for safety). 7. Open reactor and preferably inspect reactor, especially grid. Remove any sheets or chunks on grid or along walls, and ensure grid is good and clean. 8. Close reactor and remove blinds on feed lines. 9. Rapidly purge the reactor of oxygen using a high-pressure (about 325 psig) nitrogen (N₂) purge line, purge 3 times (3X) filling the reactor with N₂ to between 5-20 psig each time. 10. Start compressor. 11. Rapidly purge the reactor using a high pressure N₂ line several times. The purge operations are continued (usually about seven times) until the moisture level is constant. It typically stabilizes at a moisture reading of about 5 ppm or less. With the use of a high-pressure nitrogen line, steps 9, 11 14 and 15 can be done rapidly. 12. Depending on the water content in the reactor, add a small amount of triethyl aluminum (TEA) to reduce water level to <1 ppm. 13. Circulate gases through the empty bed. 14. Rapidly purge 3 times using high-pressure N₂ lines to facilitate removal of TEA and water reaction products. 15. Repressure to 150 psi with N₂. 16. Test for leaks. 17. Drop pressure. In Example 1 , the reactor pressure is dropped to 130 psig. In Example 2, the reactor pressure is dropped to 5-10 psig in order to facilitate transferring the polymer powder with high-pressure nitrogen from the seedbed silo. 18. Transfer the seedbed polymer into the reactor. The amount transferred is equal to about 1/2 the weight of the operational bed. The transfer can be done by various means such as by aspiration of polymer powder from a nitrogen-blanketed storage silo, or preferably by blowing the powder from a storage silo using nitrogen. 19. Make polymerization composition by adding feed and other gases, e.g., ethylene monomer, hexene or butene comonomer, and H₂ and then heat reactor to 180°F. Optionally, an antistatic agent can be added in an amount of 5 to 500 ppm by weight based on the polymer production rate. 20. Start feeding the metallocene catalyst; the reactor will light, i.e., polymerization will begin, within four to ten hours. 21. Adjust polymerization conditions to make product meeting LLDPE polymer specifications.

Table 1 shows the approximate times for each step.

Table 1

^" Example 1 reactor pressure 130 psig, Example 2 reactor pressure 5-10 psig. ^* Example 1 uses LLDPE via aspiration, Example 2 uses HDPE via nitrogen blow-in. As can be seen, the use of HDPE as the seedbed, transferring the seedbed powder using pressurized nitrogen to blow it into the reactor and purging using high-pressure nitrogen supply lines reduces the total turnaround time for the changeover significantly. While the invention has been described above in terms of preferred embodiments, it is to be understood that variations and modifications may be used as will be appreciated by those skilled in the art. Indeed, there are many variations and modifications to the above embodiments which will be readily evident to those skilled in the art, and which are to be considered within the scope of the invention as defined by the following claims.

Claims

WHAT IS CLAIMED IS:

1. An improved process for changing between two incompatible polymerization catalysts in a commercial scale olefins polymerization process, comprising: polymerizing olefins with a first polymerization catalyst; stopping the polymerization; removing substantially all of the polymer from the reactor; rapidly purging with nitrogen; adding a seedbed of polymer particulates to the reactor, and polymerizing olefins with a second polymerization catalyst.

2. A process in accordance with claim 1 wherein the process is completed in less than 3 days.

3. A process in accordance with claim 1 wherein the process is completed in less than 2 days.

4. A process in accordance with claim 1 wherein the polymerization is stopped without poisoning the catalyst in the reactor.

5. A process in accordance with claim 1 wherein the polymer seedbed comprises HDPE.

6. A process in accordance with claim 1 further comprising removing non-powder polymer chunks.

7. A process in accordance with claim 1 wherein the olefins comprise ethylene.

8. A process in accordance with claim 1 wherein the olefins comprise ethylene and a comonomer selected from propylene, butene-1 , 4-methylpentene-1 , octene-1 or decene-1.

9. A process in accordance with claim 1 wherein the polymer produced is polyethylene.

10. A process in accordance with claim 1 wherein the change in polymerization catalysts is between a Ziegler-Natta catalyst and a metallocene catalyst.

11. A process in accordance with claim 1 wherein the change in catalysts is between a chromium-based catalyst and a metallocene catalyst.

12. A process in accordance with claim 1 wherein after the polymerization is stopped, at least 95% of the polymer is removed from the reactor.

13. A process for changing between two incompatible catalysts, a first catalyst and a second catalyst, during olefins polymerization, said process comprising: (a) operating an olefins polymerization plant which comprises a reactor system having a reactor, means for easy polymer removal and a means for polymer introduction, said plant utilizing a first catalyst at a first set of polymerization conditions; (b) stopping the polymerization; (c) removing substantially all of the polymer from the reactor system; (d) rapidly purging the reactor with nitrogen; (e) drying the reactor; (f) adding a seedbed of high melting polymer particulates to the reactor; (g) adding reaction gases comprising monomer, comonomer, hydrogen, nitrogen, and an inert hydrocarbon; (h) adding said second polymerization catalyst to the reactor; (i) polymerizing one or more olefins at a second set of polymerization conditions.

14. A process in accordance with claim 13 wherein the process is completed in less than 3 days.

15. A process in accordance with claim 13 wherein the process is completed in less than 2 days.

16. A process in accordance with claim 13 wherein the polymerization is stopped without poisoning the catalyst in the reactor.

17. A process in accordance with claim 13 wherein the polymer seedbed comprises HDPE.

18. A process in accordance with claim 13 further comprising removing non-powder polymer chunks.

19. A process for changing between two incompatible catalysts, a first catalyst and a second catalyst, during olefins polymerization, the process comprising: (a) operating an olefins polymerization plant which comprises a reactor system having a reactor, a grid plate, a means for easy polymer removal, and a means for polymer introduction, said plant utilizing said first catalyst at a first set of polymerization conditions; (b) stopping the polymerization without using a catalyst kill agent; (c) removing substantially all of the polymer from the reactor; (d) shutting down the compressor; (e) opening the reactor, and inspecting and removing non-powder chunks; (f) rapidly purging the reactor with nitrogen; (g) drying the reactor; (h) adding a HDPE seedbed that is at least 40% of the weight of the operational polymer bed to the reactor; (i) adding reaction gases comprising monomer, comonomer, hydrogen, nitrogen, and an inert hydrocarbon; (j) adding said second polymerization catalyst to the reactor; (k) polymerizing one or more olefins at a second set of polymerization conditions to produce a polyethylene product.

20. A process in accordance with claim 19 wherein the process is completed in less than 3 days.

21. A process in accordance with claim 19 wherein the process is completed in less than 2 days.

22. The process of claims 1 or 13 or 19 wherein the process further comprises a transition between comonomers.