EP 2 089 705 B1 . » el OAATION AE Ea (1) EP 2089 705 B1 (12) EUROPEAN PATENT SPECIFICATION (45) Date of publication and mention (51) inte ofthe grant ofthe patent: GOIN 39409 GOIN 25/1a°%=509 2409.2010 Bulletin 201012 Goin reo (21) Application number: 07866225.1 (86) Intemational application number: PorreP2o07/010281 (22) Date offing: 26.11.2007 (87) Intemational publication number: Wo 2008/064843 (05.06.2008 Gazette 2008/23) (64) METHOD OF RAPIDLY DETERMINING THE MFR IN THE HIGH-PRESSURE POLYMERIZATION OF ETHYLENE. VERFAHREN ZUR RASCHEN MAF-BESTIMMUNG BEI DER HOCHDRUCKPOLYMERISIERUNG VON ETHYLEN PROCEDE DE DETERMINATION RAPIDE DU DEBIT MASSIQUE DANS LA POLYMERISATION A HAUTE PRESSION DE ETHYLENE (84) Designated Contracting States: (72) Inventor: SCHONAICH, Dirk AT BE BG CH CY CZ DE DK EE ES FIFR GB GR 13620 Carry le Rouet (FR) HU IE IS IT LILTLU LV MC MT NL PL PT ROSE SISK TR (74) Representative: Seelert, Stefan Basell Polyoletine GmbH (20) Priory: 27.11.2008 DE 102006055853 Intellectual Property 21.12.2006 US 876445 P Industriopark Hoechst, B852, 3rd floor 65926 Frankfurt am Main (DE) (43) Date of publication of apolication: 19.08.2009 Bulletin 2009734 (66) References cited! EP-A-0.311 300 US-A1- 2003 192.367 (73) Proprietor: Basell Polyolefine GmbH 50389 Wesseling (DE) Note: Within nine months ofthe publication of the mention ofthe grant of the European patent In the European Patent Bulletin, any person may give notice to the European Patent Office of opposition to that patent, in accordance with the Implementing Regulations. Notice of opposition shall not be deemed to have been fled untl the opposition fee has been paid. (Art. 99(1) European Patent Convention),EP 2.089 705 B1 Description [0001] The invention relates to a method of rapidly determining the mett mass flow rate inthe high-pressure polym- cerization of ethylene, [0002] |The high-pressure polymerization process continues to be a valued process for preparing low density polyeth ylene (LOPE) which is carried out very successfully worldwide on an industrial scale in numerous plants. In the high pressure polymerization, the polymerization usually initiated by means of atmospheric oxygen orby means ofperoxides ‘or by means of other free-radical formers or by means of mixtures of these, [0003] in the high-pressure polymerization, the reaction gas is firstly heated to a temperature in the range from 90 to 200°C in order to start the strongly exothermic polymerization reaction. The heat of reaction which is then liberated in the actual polymerization Is removed by water cooling; generally up to 30% of the monomers used are converted Into polymer in one pass through the continuously operated tube reactor. [0004] _A very important parameter for characterizing LDPE is its melt mass flow rate (MFR) In accordance with ISO 1133, since itis, apart trom the density of the polyethylene and the amount and type of additives added, the feature Which is generally used for the specification of various LDPE sales products. Apart from the density and the additive content, itis the parameter which serves to define the various product types. The melt mass flow rate (MFR) is the most important directly influenceable quality parameter of low density polyethylene (LOPE). [0005] US2003/192367 discloses a method of determining the melt mass flow rate (MR) ofapolymermeltin areactor asit flows from a frst reactor to a second reactor (extruder, or moulding apparatus) ata predetermined flow rate; wherein the calculation of the MFR is based upon a pressure differential calculation with respect to a known flow rate and temperature difference. [0006] In existing plants, this parameter is determined in various ways: 1. Manual, discontinuous sampling of pellets between extruder and devolatilization hoppers and subsequent manual measurement, 2. Automatic, continuous sampling between extruder and devoltilzation hoppers and also automatic, continuous ‘measurement by means of calibrated viscosity measurement apparatuses. [0007] Both methods have the disadvantage that the MFR of the product actually produced can be determined only after a dead time of 20 - 45 minutes (residence time of the product between reactor outlet and sampling point plus ‘measurement time), [0008] During start-up ofthe plant and in the case of changes of product type, the MFR isthe parameter which, owing to the above-deserbed method, takes up the most time and thus determines the amount of ou-of-specifieation product produced. There is therefore a great need to determine the MFR of the product as close as possible to production [0009] itis an object ofthe present invention to provide a process forthe high-pressure polymerization of ethylene by ‘means of which quantities of out-of-specification polymer can be largely avoided, This also applies particulary to start up and product change procedures. [0010] "The present invention is based on the recognition that the heat transfer coetficientin the heat exchanger atthe reactor outlet (after-cooler) of a high-pressure reactor depends on the viscosity and thus also the melt mass flow rate of the product and can be correlated with this. [0011] Accordingly, a method of determining the melt mass flow rate of an ethylene polymer in the polymerization of ‘ethylene or of ethylene with further comonomers, which is carried out at temperatures of from 140°C to 350°C and pressures of from 40 MPa to 350 MPain a eactorto orm the ethylene polymer andthe ethylene polymer is subsequently cooled in a heat exchanger, which comprises the steps: 1a) establishment of a correlation between the thermal conductivity ofthe ethylene polymer in the heat exchanger {and the melt mass flow rate of the ethylene polymer, with the individual measurements each being carried out at an essentially constant melt mass flow rate, b) measurement ofthe thermal conductivity of an ethylene polymer in the heat exchanger, €) calculation of the melt mass flow rate from the thermal conductivity measured inb) and the correlation established ina), is provided. [0012] The method makes it possible to determine the melt mass flow rate (MFR) of the polymer as soon ast leaves the reactor and lows through the after-cooler.Itis nat necessary to walt forthe residence time ofthe productater leavingEP 2.089 705 B1 the aftercooler until the result of the analysis is obtained, [0013] In this way, the MFR can be determined significantly earlier and the quantly of out-of-specifcation product produced during start-up and changes of product type can thusbe significantly educed. Furthermore, it makes itpossible to recognize deviations of the MFR in steady-state operation of the plant significantly earlier than has hitherto been the ‘case. Itis possible to undertake corrective interventions before a deviation can be determined by conventional methods. This can likewise mean that the production of out-of-spectication product is avoided. In any case, it makes it possible to keep the MFR within a narrow range. [0014] In step a), a correlation between the thermal conductivity and the melt ass flow rate is firstly determined, This, ispreferably achieved by determining the heat transfer coefficient which represents measure ofthe thermal conductivity after equilbrium has been attained in the cooler. Inthe simplest case, its suficient to determine a correlation for the respective product types, Le. for each MER (at one and the same density) and for each density. Although the density itsef has no significant influence on the heat transfer coefficient, the correlation shifts a a result of higher densities being produced under other process conditions (pressure, temperature). The empirical formula resulting therefrom then ‘makes it possible to determine the MFR quantitatively by means of the heat transfer coefficient ofthe after-cooler. In the simplest case (coverage ofa relatively small bandwidth of varous MFR), alinear relationship between heat transfer coztticient and MFR can be used. In addition, all customary data correlation methods can in principle be used [0015] The melt mass flow rate (MFR) is determined here by conventional analysis, for example in accordance with ISO 1198, of the polymer product. The most frequently used concitions are 190°C and a weight of 2.16, 5 or 21.6 kg [0016] The heat transfer coetficient can be calculated, for example, from the thermal balance in accordance with the formula () Amidt cp (Te Tr!) = KA AT @ Without being restricted thereto, where the parameters have the following meanings: dist: mass flow ofthe reaction mixture pi —_~heat capacity ofthe reaction mixture K heat transfer coefficient A heat transfer surface area of after-cooler Tpit, Taott: temperature ofthe reaction mixture on entry into or exit from the cooler AT: mean temperature difference between the cooling medium (pressurized water) and the reaction mixture. [0017] To obtain information about the trendin the MFR, itis sufficient to monitorthe heat transfer coefficient by means ‘of measured values determined withthe aid of existing measuring instruments, e.g. temperature ofthe reaction mixture atthe inlet and outet of the after-cooler (Tp, Ta") and the temperature ofthe cooling water atthe inlet and outet of the after-cooler, witnoutcarrying outan absolute calculation. A decreasing heettransfer coefficient indicates a decreasing MFR and vice versa. [0018] Depending on the plant configuration, some variables can remain constantin the calculation ofthe heat transfer Coefficient, so that the correlation of the MFR can be carried out drectly withthe temperatures of the reaction mixture atthe inlet and outlet ofthe after-cooler. This simplifies the determination ofthe heat transfer coeficientwithout influencing the resut [0019] However, the aftercooler can display different behavior inthe long term, for example as a result of polymer deposit, whichis why itis advantageous also to check the correlation wit the MFR continually. This can be achieved, for example, by a regular set-actual comparison of the MFR sensor with the MFR measured in-ine on the product with atime delay, [0020] The method ofthe invention can be employed only when the efter-cooler has reached an at least approximately Constant operating temperature. This restriction means that the MFR which is determined by the method of the invention during the frst minutes after start-up of the polymerization does not coincide with the actual value. The reason for this is that the metal ofthe after-cooler stil takes up heat up to this point in time and therefore gives a false value deviating from the ectual heat transfer coefficient. Only after thermal equilrium has been established in the atter-cooler are reliable data produced, However, compared to the time taken to set the MFR after startup, this phase is so shor that its significance is negligible. [0021] The method ofthe invention is particularly advantageous when the polymerization initiator mixtures are metered in continuously and essentialy pulsation-ree, since a uniform initiator concentration profile is established under such conditions and no fluctuations interfere inthe optimization process. [0022] The method of the invention can be used both forthe homopolymerization of ethylene and for the copolymer:EP 2.089 705 B1 ization of ethylene with one or more other monomers, provided that these monomers are free-racicelly copolymerizable with ethylene under high pressure. Examples of suitable copolymerizable monomers are o,f-unsaturated C.-Cy car boxylic acids, in panicular maleic actd, fumaric acid, aconic acid, acrylic acid, methacrylic acid and crotonic aca, derivatives of u,J-unsaturated Cz-Cy-carooxylic acids, @.g, unsaturated Cy-C,,-carboxyiic esters, in particular esters of C}-Ce-alkanols, or anhydrides, in particular methyl methacrylate, ethyl methacrylate, n-butyl methacrylate or tert-buty methacrylate, methyl acrylate, ethyl acrylate, n-butyl acryate, 2-ethylhexyl acrylate, tert-butyl acrylate, methacrylic an: hydride, maleic anhycrie or itaconic anhydride, and ocolafins such as propene, ‘-butene, 1-pentene, 1-hexene, 1 ‘octene or 1-decene. In addition, vinyl carboxylates, particularly preferably Vinyl acetate, can be used as comonomers n-Butyl acrylate, acrylic acid or methacrylic acid are particularly advantageously used as comonomer. The proportion ‘of comonomer or comonomers in the reaction mixture is from 1 to 45% by weight, preferably from 3 to 80% by weight, based on the amount of ethylene. Inthe case of copolymerization, the further monomers are preferably fed in ata plurality of diferent points on the reactor. [0023] For the purposes of the present invention, polymers are all substances which are made up of at least two ‘monomer units. They are preferably LDPE polymers having a mean molecular weight M, of more than 20.000 g. However, the method of the invention can also be advantageously employed in the preparation of oligomers, waxes and polymers having a molecular weight M, of less than 20 000 g. [0024] Possible initiators for starting the polymerization in the respective reaction zone are, for example, air, oxygen, ‘azo compounds or peroxidic polymerization initiators. Initiation using organic peroxides or azo compounds represents a particularly preferred embodiment of the method of the invention. Examples of suitable organic peroxides are peroxy esters, peroxy ketals, peroxy ketones and peroxycarbonates, e.g, di(2-ethyhexy!) peroxydicarbonate, dioyclohexyl per ‘oxydicarbonate, diacetyl peroxydicarbonate, tert-butyl peroxyisopropylearbonate, di-tert-butyl peroxide, d-ter-amyl per oxide, dicumy! peroxide, 2,5-dimethyl-2,6-di-tert-butylperoxyhexane, tert-butyl cum peroxide, 2,5-dimethy'-2,5-al{tert butyiperoxy)hex-3-yne, 1,3- 1000, [0028] Inthe preparation of highly viscous products, e.g. MFR = 0.26 g/10 mi, itis advantageous to add the monomer ‘or monomers not only at the inlet of the reactor tube but to feed in monomers at a plurality of diferent points on the reactor, Ths is particularly preferably done atthe beginning ofthe further reaction zones. [0029] The initiators can be employedindividually oras amixture inconcentrations ofrom0.1 toSOmoltof polyethylene produced, n particular from 0.2 to 20 molt, in each zone. Itis often advantageous to use the initiators in the dissolved state. Examples of suitable solvents are ketones and aliphatic hydrocarbons, in particular octane, decane and isodo- decane and also other saturated Cy-Cyg-hydrocarons. The solutions comprise the intistor mixtures in proportions ot from 2 to 65% by welght, preferably trom 5 to 40% by weight and particulary preferably from 10 to 30% by weight. Panicular preference is given to using mixtures of initiators which have different decomposition temperatures. [0030] In the process ofthe invention, the molar mass of the polymers to be prepared can as usual be controlled by ‘addition of molecular weight regulators. Examples of suitable regulatorsare hydrogen, aliphaticandolefinichydrecaroone, 8, pentane, hexane, cyclohexane, propene, 1-pentene or 1-hexene, Ketones such as acetone, methyl ethyl Ketone (2- bbutanone), methyl isobuty! Ketone, methyl isoamyl ketone, diethyl Ketone or diamyl ketone, aldehydes such as formal: dehyde, acetaldehyde or propionaldehyde and saturated aliphatic aleohols such as methanol, ethanol, propanol, isoEP 2.089 705 B1 propanol or butanol. Particular preference is given to using saturated aliphatic aldehydes, in particular propionaldehyde ‘or e-olefins such as propene or 1-hexene. The molecular weight regulators are preferably added to the reaction mixture upstream of the tube reactor. It can also be introduced together with the polymerization initiator at one or more points along the tube reactor. The addition ofthe regulator can be incorporated into the optimization process. [0031] The process of the invention is usually carried out at pressures of from 1x108 - 4x10® Pa (1000 to 4000 bar), With pressures of rom 1.8x10® - 3.5x10® Pa (1800 to 3500 bar) being preferred and pressures of from 2x108- 3.3x108 a (2000 to 3300 bar) being particularly preferred. The temperatures are generally in the range from 100 to 350°C, preferably rom 140 to 340°C and very particularly preferably from 150°C to 320°C, In the case of copolymerization of ethylene with sensitive or strongly regulating comonomers, in particular free-radically polymerizable carboxylic esters, e.g, vinyl esters, the polymerization is preferably cerried out at temperatures below 280°C, In general, preference is given toa process in which the polymerization temperature is no higher than 320°C. [0032] The flowing reaction mixture generally comprises polyethylene in an amount in the range from 0 to 45% by ‘eight, based on the total weight of the monomers fed to the reactor, usually up to 40% by weight. [0033] Since MFR data which are very close to the process can be provided by means of the MFR monitor of the invention, tis particularly advantageous to couple the data obtained by means of the MFR monitor with a kinetic reaction ‘model which likewise allows prediction of product data such as the MFR. As a result of the rapid back coupling with data Which are close to the process, the prediction can be continually modified and the accuracy can thereby be improved considerably [0034] The basic principles of such a system are explained below. [0035] The implementation requires a model for the reaction system in order to be able to calculate the influence of the various polymerization initiators on the reaction temperature and thus also on the product properties. A system of differential equations which should replicate the reaction kinetics in the reactor and the mass and heat balances as accuretely as possible is usually used for this purpose. Such models are prior art and form the basis ofall advanced process controllers which operate on a “ist principles” basis, as have been used for some time for the control and regulation of chemical reaction processes. Such reaction models are described, for example, in *Modelierung der Ethylen-Polymerisation im Elenac-Hochdruck-Rohrreaktorvertahren’, FO. Malling, R. Klimesch, M. Schwibach, M. Buback, M. Busch, Chem. Ing. Tech. 71, 1801 (1999), G. Luft, H. Bitsch, H. Seidl, J. Macromol. Sci-Chem., A11 (6), 1089 (1977) and in M. Busch, Macromol. Theory Simul 10, 408 (2001), [0036] To calculate the optimal composition of the initiator mixture and the optimal reaction conditions, the use of a ‘computer-aided tool for rapialy solving the system of differential equations is indispensable. Many such tools are com: ‘mercially avalable, Further details regarding the computer program "Muscod I" used here may be found in Forschrits: bericht des VDI, Series 2: Verfahcenstechnik, No. 613, 1999, in which the principles of the program and examples ofits use are described in more detall [0037] A further prerequisite for the implementation of ficient regulation is sufficiently rapid determination of the physical parameters in the reactor. Temperature measurement is preferably carried out using a temperature sensor having short response times, as is described in WO 97/25601 [0038] Furthermore, the mass flow ofthe various initiators into the reactor should preferably be continuous and as Pulsationree as possible in order to achieve good regulation. For this purpose, itis possible to use, for example, the ‘metering device described in WO 00/7085, by means of which rapid uniform mixing of polymerization initiator fed in ‘and reaction mixture is ensured, [0039] While the molar mass M, can be influenced very well by means ofthe added regulators independently of the initiators, determination of the product properties, in particular the density and the melt mass flow rate (MFA), close to production is of great advantage or precise regulation. The method of the invention of determining the MFR is particularly useful for precisely this purpose, To determine the density, samples are taken and the density is calculated from an IR. ‘or Raman spectrum, On-line determination of the density is also possible, The methods mentioned are generally known to those skilled in the art or are described in DE 102 97 394, These possible ways of determining product properties Close to the process can also be incorporated into the regulation process, as a result ef which the proportion of reject product which does not meet the required specification can be reduced further [0040] In a further advantageous embodiment ofthe present invention, the tube reactor is preceded by an autoclave by means of which a prepolymerization at a lower temperature above about 80°C is possible. [0041] A detailed description ofa suitable regulation system may be found in WO 2004/078800. [0042] _Alldocuments citedare expressly incorporated by reference into the present patent application. All percentages in this patent application are by weight based on the total weight ofthe respective mixtures, unless indicated otherwise. [0043] The invention is illustrated below withthe aid of examples, without being restricted thereto, Examples [0044] The polymerization was carried out n a high-pressure tube reactor having a length of 2200 m and a diameterEP 2.089 705 B1 (of 76 mm. The reactor had 4 reaction zones and intator was metered into the reactor atthe beginning of each of these. ‘The pressure at the reactor inlet was 2.8x10? Pa (2800 bar) and the ethylene throughput was 116 th. The after-cooler had a length of 480 m and an intemal diameter of 76 mm. A process control system from Hartmann and Braun was used. Propionaldehyde served as regulator. [0045] The heat transfer costfcient was determined from the reaction parameters withthe aid of the formula (I) [0046] Figure 1 depicts the course of the MFR determined by means of the method of the invertion and the MFR determined by conventional means at @ change of product type. The determination by conventional means was carried ‘out on the finished pellets in accordance with ISO 1133. The sawtooth curve results from the bleeding, viz. a regular ‘opening of the pressure maintenance valve, which causes a sudden pressure drop and briefly allows the temperature to rise sharply. The use of suitable fters enables the curve to be smoothed. When the process is carried out without bleeding, this sight interference in the determination of the MFR does not occur. In any case, the Interference is incon- ‘sequential since it ocours for a few seconds once per minute; the changes in the MFR to be observed occur over significantly longer periods of tme. The sharp drops in the comparative curve determined by conventional means are attoutable to bret interruptions in operation ofthe measuring instrument, [0047] it can clearly be seen that the MFR values determined by the method of the invention response significantly eater to the changes in the composition in the reactor. The dead time can be significantly reduced by meane of the ‘method ofthe invention. Comparedto the MFR deterrined by the method ofthe invention, the conventional determination of the MFR has a time delay of about 30 minutes. [0048] In the example described, the method used reduces the production of out-of-specification material in three places: 1. To prevent out-of-specfcation product contaminating a batch, the product immediately after implementation of ‘product change is usualy discarded. The novel method enables the product tobe allocated tothe in-specification batch until the MFR is actually outside the specification. 2. The method presented makes it possible to recognize whether the MFA approximates the desired value earlier than hitherto. Here too, earlier corrective interventions reduce the amount of out-ot-specification product produced. 3, The method presented makes it possible to activate transfer to the new hopper for in-spacifcation product earlier than hitherto, since the correct MFR is recognized earlier. Itis merely necessary to wat forthe (short and known) residence time of the product between atter-cooler and pellet dryer. In this way, itis ensured that out-of-spectication product which is still present inthe lines and apparatuses downstream ofthe after-cooler does not contaminate the ‘new batch of in-specification product. Claims 1. A method of determining the melt mass flow rate of an ethylene polymer in the polymerization of ethylene or of tetylene with further comonomers, which is catried out at temperatures of from 140°C to 350°C and pressures of from 40 MPa to 960 MPa ina reactor to form the ethylene polymer and tne ethylene polymer is subsequently cooled ing heat exchanger, which comprises: a) establishment of correlation between the thermal conductivity of the ethylene polymerin the heat exchanger ‘and the melt mass flow rate ofthe ethylene polymer, with the individual measurements each being caried out at an essentially constant melt mass flow rate, ) measurement ofthe thermal conductivity of an ethylene polymer in the heat exchanger 6) calculation of the melt mass flow rate from the thermal conductivity measured in b) and the correlation established ina). 2. The method according to claim 1, wherein the steps b) and c) are carried out continuously. 3. The method according to either of the preceding claims, wherein step a) is carried out discontinuously at regular Intervals 4, The method according to any of the preceding claims, wherein step a) is caried out immediately before a change to‘ product having an altered melt mass flow rate, 5. process forthe polymerization of ethylene or of ethylene with further comonomers which comprises adeterminationEP 2.089 705 B1 ‘method according to claim 1 6. The process according to claim 5, wherein the tube reactor has a length-to-diameter ratio of more than 1000. Patentanspriiche 4. Verfahren 2ur Bestimmung der Schmelze-MassenilieGrate eines Ethylenpolymersin der Polymerisation von Ethylen Coder von Ethylen mit welteren Comonomeren, das bel Tempereturen von 140 °C bis 350 °C und Drilcken von 40, MPa bis 350 MPa in einem Reaktor zum Bilden des Ethylenpolymers durchgefUhrt wird und das Ethylenpolymer aanschileBend in einem Warmetauscher gekahlt wird, das Foigendes umfasst: «@) Enmitlung einer Korrelation zwischen der Warmeleltshigkelt des Ethylenpolymers im Warmetauscher und dor Schmelze-Massenfliefrate des Ethylenpolymers, wobei die individuellen Messungen jeweils bel einer im \Wesentichen konstanten Schmelz-MassenflieSrate durchgefhrt werden, ) Messung der Warmeleitfahigkeit eines Ethylenpolymers im Warmetauscher ©) Berechnung der Schmelze-MassenfiaBrate aus der in b) gemessenen Warmeletfhigkeit und der in a) ermittetten Korrelation, 2. Verfahren nach Anspruch 1, worin die Scheie b) und c) Kentinulerich durehgeflna werden. 3. Verfahren nach einem der vorangehenden Anspriiche, worin Schritt ) in regelmaBigen intervallen diskontinulertich durchgefihrt wird. 4. Verfahren nach einem der vorangehenden Anspriche, worin Schrit @) unmittelar vor einer Umstellung auf ein Produkt mit einer geanderten Schmelze-MassentlieBrate durchgettnet wird, Verfahren zur Polymerisation von Ethylen oder von Ethylen mit weiteren Comonomeren, weiches ein Bestimmungs- Verfahren nach Anspruch 1 umfasst 66. Verfanren nach Angpruch §, worin der Rohrreaktor ein Verhaltnis von Lange zu Durchmesser von mehr als 1000 autweist. Revendications Procédé pour déterminer indice de fuiaté massique dun polymere d'éthyléne dans la polymérisation ethylene ou dethylene eveod autres comonomeres, qul est etfectuée & des temperatures de 140°C 850°C et des pressions de 40 MPa & 350 MPa dans un réacteur pour former le polymere d'éthyléne et le polymere déthyiéne est ensuite refroiai dans un échangeur de chaleur, comprenant 2) Fétablissement d'une corréation entre la conductivté thermique du polymere d'éthyiéne dans échangeur de chaleur et indice de fuidté massique du polymére ethylene par des mesures individuelles, effectuées cchacune & un indice de Tuicté massique essentillement constant, b) mesure de la conduetivté thermique 'un polymare d'éthyléne dans Téchangeur de chaleur 6) caleul de indice de fuidité massique & partrde la conductive thermique mesurée dans b) et de la corrélation établie dans a), 2. Procédé selon la revendication 1, dans lequel les étapes b) etc) sont effectuées en continu Procédé selon lune quelconque des revendications préoédentes, dans lequel Iétape a) est effectuée de maniére discontinue & des intervalles régulers. 4. Procédé selon une quelconque des revendications précédentes, dans lequel Iétape a) est effectuée juste avant ln changement vers un produit présentant un indice de fuldté massique different. 5. Procédé de polymérisation d'éthyléne ou déthyléne avec dautres comenoméres qui comprend un procédé de détermination selon la revendication 1EP 2.089 705 B1 Procédé selon |a revendication 5, dans lequel le réacteurtubulaire présente un rapport longueuradiamétre supérieur 1000.EP 2.089 705 B1 50 5.0 45 45 40 40 35, 35 30 30 25: 25 20 20 15 15 1.0 1.0 05: 05 00 0.0 6:43h 9:07 14:3th «13:55h 16:19 18:43 24:07 Time Fig. 1 uaEP 2.089 705 B1 REFERENCES CITED IN THE DESCRIPTION This list of references cited by the applicant is for the reader's convenience only. It does not form part of the European patent document. Even though great care has been taken in compiling the references, errors or omissions cannot be excluded and the EPO disclaims all ability in this regard. Patent documents cited in the description + US 2008192867 A [0005] + wo 9726601 A [0037] + Wo 0077085 A [0038] Non-patent literature cited in the description --0. Mahling ; R. Klimesch ; M. Schwibach ; M. Buback ; M. Busch. Modelierung der Ethylen-Po- lymerisation im Elenac-Hochdruck-Roohrreaktorver- fahren. Chem. ing. Tech, 1999, vol, 71, 1301 [0035] + G Luft; H. Bitsch; H. Seidl. J. Macromol, ‘Sol-Chem., 1977, vol. A11 (6), 1089 [0035] 10 DE 10287394 [0039] Wo 2004078800 A 0041} M. Busch. Macromol. Theory Simul, 2001, vol. 10, 408 [0035] Forschitisbericht des VDI, Series 3: Verfahrenstech: rik. 1999 [0036]