GB1578705A - Process for the manufacture of low-density polyethylene in stirred autoclaves with practically complete back-mixing - Google Patents

Description

(54) PROCESS FOR THE MANUFACTURE OF LOW-DENSITY POLYETHYLENE IN STIRRED AUTOCLAVES WITH PRACTICALLY COMPLETE BACK-MIXING (71) We, EC ERDOELCHEMIE GMBH, a body corporate organised under the laws of the Federal Republic of Germany, of Postfach 75 20 02, D-5000 Koeln 71, Federal Republic of Germany, do hereby declare the invention, for which we pray that a patent may be granted to us, and the method by which it is to be performed, to be particularly described in and by the following statement:- The industrial high pressure processes for the manufacture of low density polyethylene differ substantially only in the design of the reactors. According to their constructional principle, the reactors are commonly classified as tubular reactors and (stirred) autoclaves.

Tubular reactors are characterised by a very high L/D ratio (L = length, D = diameter), values between 10,000 and 50,000 being the rule. As a result of this geometry, back-mixing of the material being reacted in such reactors is extremely slight, that is to say, a very close approximation to plug flow is achieved.

The stirred autoclaves in high-pressure polyethylene processes in general have a L/D ratio of 20. Back-mixing in such autoclaves is significantly affected by the L/D ratio. At L/D ratios of < 34 and with a corresponding design of the stirrer, practically complete back-mixing is obtained. On the other hand, in the case of higher L/D ratios and when suitable inserts are used, back-mixing can be restricted in a controlled manner.

The degree of back-mixing is very important under two headings: Process stability and product properties.

Process stability is of outstanding importance in high-pressure polyethylene processes because of the extreme reaction conditions during the manufacture of low-density polyethylene (pressure range of normally 1,500---- 3,000 bars and temperature range of 150C 300"C), the high heat of polymerisation of ethylene (approx. 800 kcal/kg or 3.3 MJ/ kg) and the tendency of ethylene to decompose explosively at high pressures and temperatures above 300"C (see Chemical Engineering Science, 1973, Vol. 28, 1,505--1,514).

Good control of such extreme reaction conditions thus makes a substantial contribution to process stability. For example, a high degree of process stability also means that the number of emergency blow-downs to atmosphere and the associated dangers and consequences, such as ignition of the ethylene/air mixture and environmental pollution by noise and harmful substances, are substantially reduced. In addition to effective control instruments, such as sophisticated control loops for the temperature and pressure, it is above all intense mixing of the reaction space which assists control of the reaction, because it effects a rapid and uniform distribution of all the reactants. In this way, possible inhomogeneities in . the reaction space, which can be caused in case of incomplete mixing by fluctuations in metering and/or flow conditions, are a priori excluded and the formation of so-called "hot zones" which can be the starting point for the explosive decomposition of ethylene is avoided.

According to general opinion a reaction system exhibits practically complete backmixing, when the mixing time, i.e. the average duration of a mixing circuit is equal to a tenth or less of the average residence time of the reaction mixture.

The patent of Christal et al., U.S.

2,897,183 describes a constant environment reaction system in terms of end-to-end mixing of between 30 and 200 average number of circuits of the reacting mixture in accord with the equation NC = Gr/Gf, in which NC is the number of circuits, G, is the end-toend circulation in pounds per hour, and Gr is the monomer(s) feed rate in pounds per hour.

It is known that the end-use properties of low-density polyethylene are substantially influenced not only by the melt index and the density, which normally are set by the reaction parameters, that is to say the pressure, the temperature and the concentration of molecular weight regulator, but also by molecular data, such as long-chain branching A and nonuniformity Mw U= 1 Mn (Mw = weight average and Mn = number average molecular weight, and, in addition to the reaction parameters mentioned, reaction conditions attributable to the homogeneity or inhomogeneity of the reaction environment are responsible for these molecular data (compare Die Angewandte Makromolekulare Chemie 40/41 (1974) 361-389 (No. 586)). Such reaction environments are characterised by the presence or absence of temperature gradients, concentration gradients and pressure gradients during the course of the reaction.

Temperature gradients and concentration gradients are under adiabatic reaction con ditions mutually interdependent and are the consequence of incomplete back-mixing. The pressure gradient is normally given by the lay-out of the reaction system and is inevitable when using tubular reactors. In the case of complete back-mixing (constant environment) none of the three gradients is possible.

A relationship between, on the one hand the temperature gradient, concentration gradient and pressure gradient and, on the other hand, the properties of the product arises from the fact that all the reaction steps concerned in the polymerisation, such as initiation reactions, growth reaction, chain-transfer reactions and chain-stopping reactions, depend on the tem perature, concentration and pressure in various ways. For example, a high concentration of polymer favours the long-chain branching A which, in order to take place, requires the interaction of a free radical with a polymer chain. Under comparable reaction conditions (pressure, temperature and identical final conversion), more long-chain branches are consequently produced in a (homogeneous) reaction system with a constant polymer con centration (final concentration) than in an (inhomogeneous) reaction system in which the polymer concentration rises from a (low) initial value to the final value. The rheology, crystallisability, melt relaxation and other properties of the two products differ in spite of identical melt index.

For certain fields of application, for example the melt-coating of paper, board, metal foils and the like, products having high values of A and U are advantageous because their relaxation behaviour leads to low "neck in" values (narrowing of the melt curtain after leaving the extrusion equipment). In other fields of application, for example the blowing of thin film, products having low A and U values are advantageous since they possess good rheological properties, extensibilities and optical properties.

This comparison demonstrates the influence of back-mixing in the reaction system on the product properties. In tubular reactors backmixing is virtually prevented as a result of the design. In the case of autoclaves, the degree of back-mixing essentially depends on their geometry and their inserts.

Although a practically completely backmixed reaction system should be given preference from the point of view of process stability, as far as certain product properties are concerned the arguments for a non-backmixed reaction system prevail.

This knowledge has led to a number of proposals (U.S.A. Patent Specificarions 3,178,404, 3,536,693, 3,575,950, 3,692,763 and 3,875,128, British Patent Specification 1,071,305 and Belgian Patent Specification 710,392) with the aim of improving certain properties of the product by carrying out the polymerisation in two or more reaction zones operating at differing (rising) temperatures (temperature gradient). This is achieved largely by modifying the reactor to provide separate reaction zones, with no back-mixing in the reactor between zones, for example by means of inserts and special types of stirrer.

In order to obtain the temperature zones, U.S.A. Patent Specification 3,178,404 proposes a two-zone reactor with substantially no back-mixing between the zones as the preferred solution and, moreover, indicates the possibility of using an elongated tubular reactor or separate reactors. It is advantageous that the initiators employed for the different reaction zones have differing activities and are adapted to the particular temperatures. U.S.A.

Patent Specification 3,178,404 claims the use of caprylyl peroxide for the reaction zone having the lower temperature (1300-1900C).

In the British Patent Specification 1,071,305 the combination of two slender autoclaves (L/D 11:1 to 20:1) is described which are operated under very differing reaction conditions in such a way that the (high-molecular) reaction product of the first autoclave is led into the lower zone (mixing zone) of the second autoclave ("wax reactor"). In U.S.

Patent 3,875,128 a combination of two or more slender autoclaves (L/D 5-20) is daimed with cooling units poositioned between the autoclaves, in order to increase the conversion of the polymerisation. The resulting considerable losses of pressure and formation of deposits in the coolers are taken into account.

The advantages of these proposals are, however, contrasted by two substantial disadvantages which are caused by the loss of practically complete back-mixing: 1. a reduc tion of process stability; 2. the loss of the possibility of producing products the optimum properties of which are achieved when the reactor is practically completely back-mixed.

It has now been found that the desirable product properties as obtainable with tubular reactors can be achieved without loss of the process and product advantages of a process with practically complete back-mixing, by operating two specially designed autoclaves in series and with practically complete backmixing. The subject of the present invention is a continuous process for the manufacture of polymers of ethylene in autoclave reactors with practically complete back-mixing under pressures of 800 to 3000 bars in the presence of free-radical-generating initiators, the process being characterised in that a) two autoclaves are connected in series in such a way that all the reaction product from the first reactor fiows to the second reactor, b) each autoclave has a length/diameter (L/D) ratio of 1:1 to 3:1, c) the monomer(s) feed stream is divided between the two reactors in such a way that 1) in the case of operating under tempera ture/concentration and/or pressure con ditions in the first reactor different from that in the second reactor, the feed stream to the first reactor is equal to or greater than the feed stream to the second reactor, or 2) in the case of operating both reactors at the same conditions the monomer(s) feed stream can be divided as desired, d) the reaction temperature in the two autoclaves are set in such a way that 1) in the case of operating with a tempera ture/concentration environment in the first reactor different from that in rhe second reactor, the temperature in the first reactor is lower than the tempera ture in the second reactor, or 2) in the case of operating both reactors at the same conditions the reaction tem perature in the first reactor is the same as the reaction temperature in the second reactor, e) the initiator for the first reactor is identical with or different from the initiator for the second reactor and the quantity of initiator fed into each reactor is greater than 0.

Depending on the objective, this reaction system makes it possible to set reaction conditions which correspond either to a system having temperature, concentration and pressure gradients or to a constant environment system. Moreover, it is possible to change from one set of operating conditions to another or to change from the operating conditions with temperature, concentration and pressure gradients to the operating conditions of a constant environment system or vice versa, without having to shut down the installation or interrupt the reaction.

For it has now been found that a constant environment system can be achieved without providing alternative piping for connecting the autoclaves in parallel, thereby avoiding the disadvantages and risks associated therewith, such as the mounting of fittings, temporary close-down of parts of the piping which brings the danger of dead spaces, complex manipulations during changeovers and the associated disturbances.

The high process stability of the reaction system described, as compared with multichamber reactors or tube reactors, is ensured by both the practically complete back-mixing and the individual control instruments and protective instruments of each reactor and is not inferior to that of a well-mixed single autoclave.

The process according to the invention is thus distinguished by a high flexibility in respect of the product grades which can be manufactured by the process, high process stability and hence low environmental pol!u- tion, and by ease of management. Its mode of operation may be explained in more detail by reference to Figure 1.

The feed ethylene E, which can contain comonomers, such as vinyl acetate, butene- 1 and the like, and also molecular weight regulators, such as hydrogen, ethane, propane and the like, and which, depending on the requirement of the reaction, is under a pressure from 1,000 to 3,000 bars and at a temperature from - 200C to 1000C, is split, by a ratio controller 3, into the part streams Q1 and Q2 which pass to the reactors 1 and 2 respectively. (M designates the stirrer drives). The reaction mixture from reactor 1 passes via the connecting line 4 into the reactor 2 and the reaction mixture from the reactor 2 passes via the pressure controller 5 into collecting equipment for separating the reaction mixture.

The nature and quantity of the initiators 11 and 12, which are fed into the part streams Q1 and Q2 immediately before the latter enter into the reactors 1 and 2 respectively, are adapted to the particular requirements of the reaction.

Depending on the desired product properties the total reaction system can be operated either as a system with temperature and concentration gradients or as a constant environment system. For a reaction system with tem perature and concentration gradients the par tial streams Q1 and Q2 are adjusted in such a way that Q1 > Q2, and the reaction tem peratures T1 and T2 (in R1 and R2 respec tively) are selected according to requirements, T2 always being > T1. The nature and amount of the initiators I1 and I2 are adapted to the reaction temperatures T1 and T2 respectively. The changes in product properties which can be achieved by means of a temperature and concentration gradient are the more pronounced, the larger the difference between the reaction temperatures T1 and T2 and the smaller the part stream Q2. The association between these factors as well as the influence of the conversion gradient on the product properties is illustrated in Fig. 3.

Temperature ranges of 1200--220"C and 1900-2800C are preferred for T1 and T2, respectively. A change in the temperature difference between T2 and T1 or a change in the ratio of the part stream influences the spectrum of properties in differing ways. For a constant environment system the reaction temperature T1 and T2 are brought to the same value. Although the splitting of the part stream Q1 and Q2 is not critical in this case, for maximum initiator-efficiency it is preferably selected in such a way that the ratio Q1/ Q2 is equal to the volume ratio V1/V2 of the reactors. In that case, the initiators I1 and I2 are normally chemically identical.

The flexibility of the process according to the invention can be further extended by numerous possible variations. For example, the part streams Q1 and Q2 can be cooled or heated differently in order to exert an additional influence on the conversion gradient. A comparable influence can also be obtained by cooling the connecting line 4. Such measures allow additional variation of those differences between the properties of the products of reactors 1 and 2 which are able to be influenced by temperature and conversion gradients as well as of the relative quantities of the products formed in each reactor. In order to increase the differences in the properties of the products obtained in the two reactors the process according to the invention can also be operated with an adjustable pressure gradient between the two reactors (not shown in Figure 1), the pressure in the first reactor being higher than the pressure in the second reactor. All the variation possibilities mentioned can either be put into use independently or in combination with each other.

The examples which follow are intended to explain the process according to the invention in more detail. For all the examples two practically completely mixed stirrer autoclaves are employed, which are connected to one another according to the scheme shown in Figure 1.

In examples 1-5 and the first comparative example, two autoclaves of the same size and with an L/D ratio of 2.6 are used. In examples 6-10 and the second comparative example, the first reactor has a volume of 0.725 liters and an L/D of 1:1; the second reactor has a volume of 0.325 liters and an L/D of 1.0.

In the two comparative examples only the first reactor is used. The ethylene feed rate, E, is kept constant at 120 kg/h in examples 1-5 and in the first comparative example; in examples 6-10 and the second comparative example, E is 13 kg/h. The other reaction Darameters and the results are shown in tables I and II.

Example 1.

This example demonstrates that the process according to the invention can be operated as a constant environment reaction system. A comparison of the product properties from examDle 1, indicated in the summarizing table I, with those of the comparative example, confirms the equivalence of the products manufactured in the two examples by differing technologies.

Examples 2-5.

These examples show the influence of differing gradations of the reaction temperature temperature and differing ratios of the partial feed streams on the properties of the products.

Examples 6-10.

These examoles demonstrate that when the process according to the invention is operated under conditions producing a temperature/ conversion gradient, certain product properties differ substantially from those obtained with a single autoclave and approach, to a surprising degree, the values obtained using a tubular process. This effect is illustrated, using the relationship between extrusion pressure and Melt Index, in Figure 2, where data for commercial products made in tubular processes are compared with data for commercial products made in constant environment single-autoclave processes and with data from example d10 and 5. As can be seen the data for these examples fall on the same line as the data for the products from tubular reactors.

REACTION PARAMETERS

Proportion Concentration TE1) TR11) TR21) Pressure Q1 2) Initiator3 of conversion of molecular in R1 weight regulator4 Example C C C bars E 11 12 % mol-% comparison 20 260 - 1800 1,0 PN - 100 0,9 Pr 1 20 260 260 1800 0,5 PN PN 50 0,9 Pr 2 20 260 260 1800 0,6 PN PN 60 0,5 Pr 3 20 210 260 1800 0,6 PO PN 48 1,4 Pr 4 20 160 260 1800 0,6 PD PN 35 1,9 Pr 5 20 160 260 1800 0,95 PD PN 55 2,7 Pr 6 40 160 250 1450 1,0 PL PZ 57 1,5 Cy 7 40 160 250 1860 1,0 PL PZ 57 1,1 Pr 8 40 160 250 1860 1,0 PL PZ 57 2,5 Cy 9 85 170 250 1860 1,0 PL PZ 52 0,7 Pr 10 43 160 280 1860 1,0 PL PZ 49 1,2 Cy comparison 31 217 - 1860 1,0 PB - 100 1,1 Cy 1) TE, TR1, TR2 denote the feed temperatures and reaction temperatures in the autoclave 1 and 2 respectively.

2) Q1 = partial feed stream to autoclave 1; E = total feed stream.

3) PN = t-butyl pernonanoate ; PO = t-butyl peroctoate ; PD = t-butyl perneodecanoate ; PL = t-butyl perpivalate; PZ = di-t-butyl peroxide; pb = t-butyl perisobutyrate.

4) Pr = propylene, Cy = Cyclohexane; content of extraneous regulator in the feed stream 0.05 mol %.

PRODUCT PROPERTIES

Ml2 Density Melt5) Extrusion Gloss at Extensibility Surface8) Swell pressure6) 20 un or m/min 7) roughness Example g/10 min g/cm3 % bars % 9) m comparison 5,1 0,917 136 127 - - 0,186 1,6 1 5,0 0,917 137 126 - - 0,186 1,5 2 1,79 0,917 133 164 - 90 0,213 2,2 3 1,81 0,919 130 152 10 45 0,190 1,2 4 1,80 0,919 128 140 23 30 0,178 0,7 5 1,81 0,921 125 138 48 25 0,155 0,5 6 0,36 0,923 114 198 1,5 4,6 - 7 0,63 0,924 114 183 25 7,6 - 8 1,95 0,929 116 145 21 15 - 9 0,19 0,922 103 228 16 4,6 - 10 0,23 0,923 99 186 - 3,7 - comparison 2,0 0,928 114 169 42 12 - 5) Swelling of the molten extrudate during the MI measurement (large values denote a small "neck-in" in the extrusion of flat films).

6) Measured under standard conditions in an extrusiometer in examples 1 - 5 and during manufacture of blown film in examples 6 - 10.

7) Long-chain branches per 1,000 C atoms.

8) Measured under standard conditions on a strand from the high-pressure capillary viscometer.

9) In examples 2 - 5 the film extensibility was measured in m; in examples 6 - 10 and the second comparative example the draw rate was measured in m/min. until the break of the strand in the extrusiometer test.

Claims (8)

WHAT WE CLAIM IS:

1. A continuous process for the production of a polymer of ethylene in autoclave reactors with practically complete back-mixing under pressures of 800 to 3000 bars in the presence of a free-radical-generating initiator in which a) two autoclaves are connected in series such that all the reaction product from the first reactor flows to the second reactor, b) each autoclave has a ratio of Length to-Diameter (L/D) of between 1:1 and 3:1, c) the monomer(s) feed stream is divided between the two reactors such that 1) when operating under temperature, con centration and/or pressure conditions in the first reactor different from that in the second reactor, the feed stream to the first reactor is equal to or greater than the feed stream to the second reactor or 2) when operating both reactors at the same conditions, the monomer(s) feed stream is divided as desired, d) the reaction temperatures in the two autoclaves are set in such a way that 1) when operating with a temperature/ concentration environment in the first reactor different from that in the second reactor, the temperature in the first reactor is lower than the temperature in the second reactor or (2) when operating both reactors at the same conditions, the reaction tempera ture in the first reactor is the same as the reaction temperature in the second reactor, e) the initiator for the first reactor is iden tical with or different from the initiator for the second reactor and the quantity of initiator fed into each reactor is greater than zero.

2. A process according to Claim 1, in which the reaction is carried out under the conditions of c) 1) with the reaction temperature in the second reactor greater than the reaction temperature in the first reactor 1.

3. A process according to Claim 2, in which the temperature in the second reactor is 190 to 2800C and the temperature in the first reactor is 120 to 2200 C.

4. A process according to any of the foregoing claims, in which the reaction is carried out under the conditions of c) 1) with the monomer(s) feed stream to the first reactor 50 to 99 /O of the total feed stream.

5. A process according to any of the foregoing claims, in which the reaction is carried out under the conditions of c) 1) with the reaction pressure in the second reactor equal to or lower than the reaction pressure in the first reactor.

6. A process according to any of the foregoing claims, in which the reaction is carried out under the conditions of c) 1) with the ratio of the monomer(s) feed streams to the first and second reactor is equal to the ratio of the reaction volume of the first and second reactor.

7. A process according to Claim 1, when carried out substantially as described in any one of the Examples.

8. A polymer of ethylene when produced by the process of any of the foregoing claims.