Allemeersch - Polymerisation of ethylene : in slurry loop reactors

The polymerisation products of ethylene are among the largest commodity plastics. Low density polyethylene (LDPE) is a highly branched polymer, generated from the monomer at conditions of extremely great pressure and high temperature. High density polyethylene (HDPE) is a linear polymer, and is obtained by catalysis, at moderate pressures and temperatures. The two types of polyethylene are so different in properties that they can hardly be said to be competitors.

Two main processes are known to produce HDPE. In the gas phase process, the polymer particles are formed on catalyst particles that are fluidized in a fluidized bed reactor by a gas stream carrying the monomer. In the slurry loop process the polymer grows on catalyst particles that are suspended in a liquid, wherein the monomer is dissolved, thus forming a slurry that is circulated in a recycle tubular reactor, commonly called loop reactor.

The slurry loop reactor is a long tube, about 600 mm in diameter and a few hundred metres long, of which the ends are connected to make it one closed loop. Its volume is of the order of 100 m . It is totally filled with the slurry, which is circulated through the loop by one single circulation pump. The larger part of the reactor consists of an even number of vertical tube sections, or legs, connected by short horizontal elbows at the top and the bottom.

The outer aspect of the reactor is that of a tubular reactor, one of the two main model reactors known in chemical reactor engineering, where the reagents are fed at one side of the tube and leave it at the other end. However, since the loop is closed, and the reactor contents are circulated through the loop at a high speed, the behaviour of the reactor is very close to that of the continuous stirred tank reactor, that other model known from reactor engineering textbooks.

Actually, the loop reactor can be regarded as a stirred tank, where the circulation pump has taken over the role of the agitator, and where the vessel has been lengthened to obtain more external cooling surface, and a higher production capacity. Indeed, the heat of reaction of the polymerisation of ethylene is quite high, and the cooling surface of the reactor is one of the limiting factors to its production capacity. Since a conventional vessel is rather spherical, the ratio of its external surface to its volume is inversely proportional to its diameter. In a loop reactor, the cooling of the reactor contents is effected by cooling water flowing through a cooling jacket, concentric with the tube, and so its external surface is simply proportional to its volume. When the reactor tube is made longer, both volume and cooling surface are increased to the same degree. Due to the fact that the slurry and the cooling water are flowing at very high speeds, high heat transfer coefficients are obtained at both sides of the reactor wall.

Three families of catalysts are used in the slurry loop process. They yield polymers of different molecular weight distributions (MWD), and thus allow to produce a very broad range of high density polyethylenes, of which many specific grades are used for specific applications. The reaction mechanisms of the three catalyst families are different, because their active species, that provoke the polymerisation reaction, are different. The active chemical of the chromium oxide catalyst is chromium(VI) oxide stabilised on a silica particle. The resulting MWD is very broad. The ZieglerNatta catalyst is using a complex of TiCl4 in an environment of MgCl2. The resulting MWD is more narrow. The metallocene catalyst, where a Zr-metallocene group is the active chemical, produces the most narrow MWD. A chromium(VI) oxide catalyst cannot be realised without a silica support, but also the two other catalysts are most often supported on silica particles to facilitate their handling and feeding to the reactor. The size of those silica particles is of the order of 100 m.

The catalysts are very sensitive to poisoning, especially by moisture. That is why every feed stream to the reactor is purified in an extensive purification section, where, on a series of adsorber beds, the impurity levels are reduced to the order of 0.1 ppm.

The best way to explain how the reactor functions, is by telling how it is started up. In a first step, the loop reactor is filled with the diluent. That is a liquid wherein the reaction ingredients are dissolved, but wherein the solubility of the polymer should be minimal. The most common diluent is isobutane, and also propane is used. The diluent is in the liquid state at the prevailing reactor conditions, viz. temperatures around 100 C, and pressures around 40 barg.

The reactor is one stage in a greater loop described by this diluent in its course through the process, , where it is passing a purification step before being fed to the reactor, is flashed off upon leaving the reactor, compressed and condensed, purified in the recovery and purification sections, and finally recycled to the reactor. By the filling of the reactor is meant that this circulation of diluent through the process is set up, and that the diluent in the reactor is at the operating temperature and pressure.

Then the required concentrations of monomer, and, if necessary, comonomer and hydrogen are established in the loop reactor. The condensed liquid, entering the recovery section, will then contain, besides the diluent, also concentrations of ethylene, comonomer and hydrogen. This stream is split up in its components in the recovery section by distillation, as far as is necessary to purify them and control their quantities that are all recycled to the reactor. Only two small purge streams, of lights and heavies, are taken off in the recovery section, which means that most of the liquid is recycled.

The reaction is started by the feeding of the catalyst. Polymer will be formed on the catalyst particles, and so a polymer-diluent slurry will be established. The thickening slurry is allowed to reach its steady state concentration, at some 20 vol %. This concentration is controlled by the feed flow of the recycled diluent, and by the number of settling legs in service.

The settling legs are vertical tubes suspended to the bottom of the reactor, wherein the slurry is allowed to settle and reach a higher concentration of solids. Small quantities of concentrated slurry are released from each settling leg to the flash tank at regular intervals. The thus effected thickening of the slurry reaching the flash tank is saving capacity and energy in the recovery section.