CN104570724A - Polymerization process condition optimization method taking polyolefin microscopic quality as target - Google Patents

Abstract

The invention discloses a polymerization process condition optimization method taking polyolefin microscopic quality as a target, belonging to the field of manufacture of chemical products. The polymerization process condition optimization method is characterized by building a process mathematic model based on reaction mechanism for an olefin polymerization process, and solving an optimal polymerization process condition which meets requirements on quality of specific products by taking average molecular weight or molecular weight distribution as an index of the microscopic quality of the polyolefin. The polymerization process condition optimization method can be used for calculating to obtain the optimal process operation condition under the condition with the given average molecular weight and molecular weight distribution of the polymers; the method can be applied to the design of structures of polymer products and can be used for shortening time for developing a novel type polymer, so that energy conservation and emission reduction are implemented; the economic benefit is increased; the polymerization process condition optimization method is simple and clear in principle and easy to understand and implement; the implementation of computer programming is facilitated; the on-line control of the quality of the polymer products is facilitated.

Description

A kind of polymerization process condition optimization method that is target with polyolefine micro-quality

Technical field

The invention belongs to Chemicals and manufacture field, the present invention relates to a kind of polymerization process condition optimization method of polyolefin products more precisely.

Background technology

Polyolefin industry is often using melting index as the quality index of product, but melting index is a very macroscopical parameter, only reflects the molecular-weight average of polymkeric substance.The melting index of variant production is likely identical, but microtexture may vary, and it uses and processing characteristics also greatly differs from each other.

Molecular weight distribution is polyolefinic important micro-quality index, has important impact to the mechanical property of product and processing characteristics.The polyolefin products that design, development mechanics performance and processing characteristics balance, needs to be grasped the detailed information of molecular weight distribution.Usually, first exploitation new grades polyolefine need utilize pilot plant test to propose production data, is carried out the adjustment of production data by pilot experiment, finally attempts producing in full scale plant.But polyolefinic molecular weight distribution cannot on-line checkingi, just can show that whether polymerization process condition is correct after needing the structure off-line analysis to product.Sometimes, be also difficult to for a long time find suitable processing parameter to reach the quality index of new grades product.So not only wasted the production time but also created a large amount of transitional products, reducing the economic benefit of factory.In other words, the on-line measurement of polyolefin molecular weight distribution is very difficult, causes the real-time control of polymeric articles quality to realize.At present, the optimization of polymerization process and control depend on the whole process mathematical modeling based on strict reaction mechanism and thermodynamic equation of state.By can the mathematical model of Accurate Prediction molecular weight distribution, the hard measurement of molecular weight distribution can be realized.

US Patent No. 5687090 and US6093211 describe the modeling method can predicted polymer molecular structure such as molecular weight and distribution thereof, copolymerization composition and distribute, and establish associated analog device.But the method is confined to the product property calculating polymkeric substance at the prerequisite Imitating of known polymerization processes condition, and is not suitable for the exploitation of product innovation and the quality control of product.Because the polyolefinic quality index of new grades is known, it is required to determine that it select what kind of polymerization process condition to produce, this is that conventional polymerization process modeling is not accomplished.The method of polymericular weight and distribution thereof can increase the development time undoubtedly, improve Financial cost to adopt the operational condition of first setting device to judge again.Visible, if based on by the steady-state optimization to polymerization process condition, the polymerization process condition finding applicable new grades product quality indicator rapidly and accurately, then the economic benefit of experiment number, shortening trial time, minimizing transitional product, raising factory can be reduced.

Therefore, the present invention is directed to olefin polymerization process, set up the mathematical model based on strict reaction mechanism, propose, with the polyolefine micro-quality particularly molecular weight distribution optimal problem that is optimization aim, to use steady state optimization method to solve best process conditions.

Summary of the invention

The present invention seeks to, for developing the polyolefine product innovation length consuming time of specifying molecular weight distribution, the shortcoming that benefit is low in prior art, to provide a kind of polymerization process condition optimization method being target with polyolefine micro-quality.

Specifically, the present invention adopts following technical scheme to realize, and comprises the steps:

1) set up the mathematical model of olefin polymerization process, comprise the steps:

1-1) determine the form of the reactor of paradigmatic system, working method, each reactor is regarded as the combination of continuous stirred tank reactor or continuous stirred tank reactor;

1-2) according to form, the working method of the reactor determined, select the thermodynamical model of reactive system and determine its parameter;

1-3) on the basis of the form of the reactor determined, working method and thermodynamical model, determine polymerization kinetics mechanism and parameter thereof;

1-4) according to thermodynamical model and the polymerization kinetics mechanism of reactive system, set up material balance and the energy balance model of polymerization process, calculated by material balance and energy balance model comprise transient molecular amount and distribution thereof, copolymerization forms and distributes, accumulation molecular weight and distribution, copolymerization composition and be distributed in interior polyolefinic molecular structure information;

2) using the target that polyolefine molecular-weight average or molecular weight distribution are optimized as polymerization process condition, the initial value of setting polymerization process condition;

3) according to the value of polymerization process condition, the material balance utilizing step 1) to obtain and energy balance model calculate polyolefine molecular-weight average or molecular weight distribution;

4) the polyolefine molecular-weight average calculated as step 3) or molecular weight distribution meet the optimization tolerance preset, then enter step 5), otherwise the value of amendment polymerization process condition, returns step 3) and carries out iteration until calculate the polyolefine molecular-weight average or molecular weight distribution that meet the optimization tolerance preset;

5) using the best polymerization process condition that the value of polymerization process condition now obtains as optimization, method ends.

Of the present inventionly to be further characterized in that: described olefin polymerization process is alkene homopolymerization or polymerization process.

Of the present inventionly to be further characterized in that: described paradigmatic system at least comprises a reactor and related procedure equipment.

Of the present inventionly to be further characterized in that: the working method of described reactor is interval, semicontinuous or continuous.

Of the present inventionly to be further characterized in that: the implementation method of described olefin polymerization process is slurry phase, body phase or gas phase.

Of the present inventionly to be further characterized in that: described material balance and energy balance model are stable state or the dynamicmodel of single or multiple reactors in series.

Of the present inventionly to be further characterized in that: described material balance and energy balance model, can to calculate and follow the tracks of the polymer molecular structure information of reactive system random time, arbitrarily reactor.

Of the present inventionly to be further characterized in that: described polymerization process condition is the temperature of reactor, pressure, hydrogen and monomer mole ratio.

Of the present inventionly to be further characterized in that: described step 2) in, the value of hydrogen and monomer mole ratio in the reactor having trade mark product according to device, the initial value of hydrogen and monomer mole ratio in setting reactor.

Of the present inventionly to be further characterized in that: described optimization tolerance is 0.0001.

Beneficial effect of the present invention is: present invention achieves under the prerequisite of given Polymer average molecular weights and molecular weight distribution, calculates optimum process conditions; The present invention can be applied to the design of polymeric articles structure, shortens the time of exploitation new grades polymkeric substance, thus realizes energy-saving and emission-reduction, increase economic efficiency; Principle of the invention simple and clear, easy to understand is also implemented, and is convenient to realize computer programming, is conducive to the on-line Control of polymeric articles quality.

Accompanying drawing explanation

Fig. 1 is optimization method block diagram of the present invention

Fig. 2 is continuous stirred tank reactor model (CSTR) schematic diagram

The optimum result of Fig. 3 to be ethene slurry two still series polymerizations flow process with molecular-weight average be target

Fig. 4 is that ethene slurry two still series polymerizations flow process calculates the molecular weight distribution of gained and comparing of analytical value with Different Optimization target

Fig. 5 is that ethene slurry two still series polymerizations flow process calculates the residual error comparison diagram of gained with Different Optimization target.

Embodiment

Below in conjunction with drawings and Examples, the inventive method is described in detail.The parameters hereinafter occurred, its implication refers to nomenclature table.

One embodiment of the present of invention, its step is specific as follows:

1) set up the mathematical model of olefin polymerization process, step is as shown in the right block diagram of Fig. 1.

1. form, the working method of reactor is first determined.Commercial olefin polymerization device has upright tank reactor, annular-pipe reactor, horizontal type agitated bed reactor and fluidized-bed reactor etc., and operational stage has the equal multiple phase of slurry phase, body phase, gas phase, supercutical fluid.But dissimilar reactor all can regard continuous stirred tank reactor (CSTR) as shown in Figure 2, or the combination of CSTR.

2. select the thermodynamic equation of state of polymerization system and determine its parameter.The present embodiment adopts the physical property of chain disturbance statistics creation (PC-SAFT) state equation description system and balances each other.By pure component in reparameterization method determination polymeric system as the parameter of the PC-SAFT state equations such as ethene, hydrogen, nitrogen.On the basis accurately determining pure component model parameter, obtain the interaction parameter of the physical data determination binary composition according to document.

3. polymerization kinetics mechanism and parameter thereof is determined.Olefinic polymerization kinetics mechanism is as shown in table 1, comprises the activation of catalyzer, chain initiation, chainpropagation, chain tra nsfer, chain inactivation, k _aA, k _i, k _p, k _tM, k _tH, k _dfor the Kinetics Rate Constants By Using of corresponding elementary reaction, j is active sites numbering.In reactions steps, chainpropagation and chain tra nsfer are the deciding step determining polyolefin molecular weight, and therefore, the mol ratio of hydrogen and monomer is the design variable controlling polyolefin molecular weight.

Table 1 olefinic polymerization kinetics mechanism

4. the material balance of process of establishing and energy balance model.Below for single polymerization reactor, set up material balance and energy balance model equation is as follows:

Catalyst component:

$\frac{d\left(\right[C_p\left]V\right)}{\mathrm{Vdt}}=\frac{q_{\mathrm{in},\mathrm{cat}}{\left[C_p\right]}_f}{V}-\underset{j=1}{\overset{\mathrm{Ns}}{\mathrm{\Σ}}}{k}_{\mathrm{aA}}\left(j\right)\left[A\right]F\left(j\right)\left[C_p\right]-\underset{j=1}{\overset{\mathrm{Ns}}{\mathrm{\Σ}}}\frac{q_{\mathrm{out}}}{V}\·F\left(j\right)\left[C_p\right]$

$\frac{d\left(\right[A\left]V\right)}{\mathrm{Vdt}}=\frac{q_{\mathrm{in},A}{\left[A\right]}_f}{V}-\underset{j=1}{\overset{\mathrm{Ns}}{\mathrm{\Σ}}}{k}_{\mathrm{aA}}\left(j\right)\left[A\right]F\left(j\right)\left[C_p\right]-\frac{q_{\mathrm{out}}}{V}\left[A\right]$

$\begin{array}{c}\frac{d\left(\right[P_0\left(j\right)\left]V\right)}{\mathrm{Vdt}}=\frac{q_{\mathrm{in},\mathrm{cat}}{\left[P_0\left(j\right)\right]}_f}{V}+k_{\mathrm{aA}}\left(j\right)\left[C_p\left(j\right)\right]\left[A\right]-k_{\mathrm{iA}}\left(j\right)\left[P_0\left(j\right)\right]\left[M_A\right]-k_{\mathrm{iB}}\left(j\right)\left[P_0\left(j\right)\right]\left[M_B\right]+\\ k_{\mathrm{tHA}}\left(j\right)Y_A^0\left(j\right){\left[H_0\right]}^{0.5}+k_{\mathrm{tHB}}\left(j\right)Y_B^0\left(j\right){\left[H_0\right]}^{0.5}+(k_{\mathrm{tAA}}\left(j\right)Y_A^0\left(j\right)+k_{\mathrm{tBA}}\left(j\right)Y_B^0\left(j\right))\left[M_A\right]+\\ (k_{\mathrm{tAB}}\left(j\right)Y_A^0\left(j\right)+k_{\mathrm{tBB}}\left(j\right)Y_B^0\left(j\right))\left[M_B\right]-\frac{q_{\mathrm{out}}}{V}\left[P_0\left(j\right)\right]\end{array}$

Monomer and hydrogen:

$\frac{d\left(\right[M_A\left]V\right)}{\mathrm{Vdt}}=\frac{q_{\mathrm{in},M_A}{\left[M_A\right]}_f}{V}-\underset{j=1}{\overset{\mathrm{Ns}}{\mathrm{\Σ}}}\left(k_{\mathrm{iA}}\left(j\right)\right[P_0\left(j\right)]+k_{\mathrm{pAA}}\left(j\right)Y_A^0\left(j\right)+k_{\mathrm{pBA}}\left(j\right)Y_B^0\left(j\right)+k_{\mathrm{tAA}}\left(j\right)Y_A^0\left(j\right)+k_{\mathrm{tBA}}\left(j\right)Y_B^0\left(j\right))\left[M_A\right]-\frac{q_{\mathrm{out}}}{V}\left[M_A\right]$

$\frac{d\left(\right[M_B\left]V\right)}{\mathrm{Vdt}}=\frac{q_{\mathrm{in},M_B}{\left[M_B\right]}_f}{V}-\underset{j=1}{\overset{\mathrm{Ns}}{\mathrm{\Σ}}}{k}_{\mathrm{iB}}\left(j\right)\left[P_0\left(j\right)\right]+k_{\mathrm{pAB}}\left(j\right)Y_A^0\left(j\right)+k_{\mathrm{pBB}}\left(j\right)Y_B^0\left(j\right)+k_{\mathrm{tAB}}\left(j\right)Y_A^0\left(j\right)+k_{\mathrm{tBB}}\left(j\right)Y_B^0\left(j\right)\left[M_B\right]-\frac{q_{\mathrm{out}}}{V}\left[M_B\right]$

$\frac{d\left(\right[H_2\left]V\right)}{\mathrm{Vdt}}=\frac{q_{\mathrm{in},H_2}{\left[H_2\right]}_f}{V}-\underset{j=1}{\overset{\mathrm{Ns}}{\mathrm{\Σ}}}(k_{\mathrm{tHA}}\left(j\right)Y_A^0\left(j\right)+k_{\mathrm{tHB}}\left(j\right)Y_B^0\left(j\right)){\left[H_2\right]}^{0.5}-\frac{q_{\mathrm{out}}}{V}\left[H_2\right]$

0 rank square of polymers alive and dead polymers:

$\begin{array}{c}\frac{d\left(Y_A^0\left(j\right)V\right)}{\mathrm{Vdt}}=\frac{q_{\mathrm{in}}{Y}_A^0{\left(j\right)}_f}{V}+\left(k_{\mathrm{iA}}\left(j\right)\right[P_0\left(j\right)]+k_{\mathrm{pBA}}\left(j\right)Y_B^0\left(j\right))\left[M_A\right]\\ -\left(k_{\mathrm{pAB}}\left(j\right)\right[M_B]+k_{\mathrm{tHA}}\left(j\right){\left[H_2\right]}^{0.5}+k_{\mathrm{dA}}\left(j\right)+k_{\mathrm{tAA}}\left(j\right)[M_A]+k_{\mathrm{tAB}}\left(j\right)[M_B\left]\right)Y_A^0\left(j\right)-\frac{q_{\mathrm{out}}}{V}{Y}_A^0\left(j\right)\end{array}$

$\begin{array}{c}\frac{d\left(Y_B^0\left(j\right)V\right)}{\mathrm{Vdt}}=\frac{q_{\mathrm{in}}{Y}_B^0{\left(j\right)}_f}{V}+\left(k_{\mathrm{iB}}\left(j\right)\right[P_0\left(j\right)]+k_{\mathrm{pAB}}\left(j\right)Y_A^0\left(j\right))\left[M_B\right]\\ -\left(k_{\mathrm{pBA}}\left(j\right)\right[M_A]+k_{\mathrm{tHB}}\left(j\right){\left[H_2\right]}^{0.5}+k_{\mathrm{dB}}\left(j\right)+k_{\mathrm{tBA}}\left(j\right)[M_A]+k_{\mathrm{tBB}}\left(j\right)[M_B\left]\right)Y_B^0\left(j\right)-\frac{q_{\mathrm{out}}}{V}{Y}_B^0\left(j\right)\end{array}$

$\begin{array}{c}\frac{d\left(X^0\left(j\right)V\right)}{\mathrm{Vdt}}=\frac{q_{\mathrm{in}}{X}^0{\left(j\right)}_f}{V}+(k_{\mathrm{tHA}}\left(j\right){\left[H_2\right]}^{0.5}+k_{\mathrm{dA}}\left(j\right)+k_{\mathrm{tAA}}\left(j\right)[M_A]+k_{\mathrm{tAB}}\left(j\right)[M_B\left]\right)Y_A^0\left(j\right)\\ +(k_{\mathrm{tHB}}\left(j\right){\left[H_2\right]}^{0.5}+k_{\mathrm{dB}}\left(j\right)+k_{\mathrm{tBA}}\left(j\right)[M_A]+k_{\mathrm{tBB}}\left(j\right)[M_B\left]\right)Y_B^0\left(j\right)-\frac{q_{\mathrm{out}}}{V}{X}^0\left(j\right)\end{array}$

1 rank square of polymers alive and dead polymers:

$\begin{array}{c}\frac{d\left(Y_A^1\left(j\right)V\right)}{\mathrm{Vdt}}=\frac{q_{\mathrm{in}}{Y}_A^1{\left(j\right)}_f}{V}+\left(k_{\mathrm{iA}}\left(j\right)\right[P_0\left(j\right)]+k_{\mathrm{pAA}}\left(j\right)Y_A^0\left(j\right))\left[M_A\right]+k_{\mathrm{pBA}}\left(j\right)\left[M_A\right](Y_B^1\left(j\right)+Y_B^0\left(j\right))\\ -\left(k_{\mathrm{pAB}}\left(j\right)\right[M_B]+k_{\mathrm{tHA}}\left(j\right){\left[H_2\right]}^{0.5}+k_{\mathrm{dA}}\left(j\right)+k_{\mathrm{tAA}}\left(j\right)[M_A]+k_{\mathrm{tAB}}\left(j\right)[M_B\left]\right)Y_A^1\left(j\right)-\frac{q_{\mathrm{out}}{Y}_A^1\left(j\right)}{V}\end{array}$

$\begin{array}{c}\frac{d\left(Y_B^1\left(j\right)V\right)}{\mathrm{Vdt}}=\frac{q_{\mathrm{in}}{Y}_B^1{\left(j\right)}_f}{V}+\left(k_{\mathrm{iB}}\left(j\right)\right[P_0\left(j\right)]+k_{\mathrm{pBB}}\left(j\right)Y_B^0\left(j\right))\left[M_B\right]+k_{\mathrm{pAB}}\left(j\right)\left[M_B\right](Y_A^1\left(j\right)+Y_A^0\left(j\right))\\ -\left(k_{\mathrm{pBA}}\left(j\right)\right[M_A]+k_{\mathrm{tHB}}\left(j\right){\left[H_2\right]}^{0.5}+k_{\mathrm{dB}}\left(j\right)+k_{\mathrm{tBA}}\left(j\right)[M_A]+k_{\mathrm{tBB}}\left(j\right)[M_B\left]\right)Y_B^1\left(j\right)-\frac{q_{\mathrm{out}}{Y}_B^1\left(j\right)}{V}\end{array}$

$\begin{array}{c}\frac{d\left(X^1\left(j\right)V\right)}{\mathrm{Vdt}}=\frac{q_{\mathrm{in}}{X}^1{\left(j\right)}_f}{V}+(k_{\mathrm{tHA}}\left(j\right){\left[H_2\right]}^{0.5}{k}_{\mathrm{dA}}\left(j\right)+k_{\mathrm{tAA}}\left(j\right)[M_A]+k_{\mathrm{tAB}}\left(j\right)[M_B\left]\right)Y_A^1\left(j\right)\\ +(k_{\mathrm{tHB}}\left(j\right){\left[H_2\right]}^{0.5}+k_{\mathrm{dB}}\left(j\right)+k_{\mathrm{tBA}}\left(j\right)[M_A]+k_{\mathrm{tBB}}\left(j\right)[M_B\left]\right)Y_B^1\left(j\right)-\frac{q_{\mathrm{out}}{X}^1\left(j\right)}{V}\end{array}$

2 squares of polymers alive and dead polymers:

$\begin{array}{c}\frac{d\left(Y_A^2\left(j\right)V\right)}{\mathrm{Vdt}}=\frac{q_{\mathrm{in}}{Y}_A^2{\left(j\right)}_f}{V}+\left(k_{\mathrm{iA}}\left(j\right)\right[P_0\left(j\right)]+k_{\mathrm{pAA}}\left(j\right)(2Y_A^1\left(j\right)+Y_A^0\left(j\right))+k_{\mathrm{pBA}}\left(j\right)(Y_B^1\left(j\right)+2Y_B^1\left(j\right)+Y_B^0\left(j\right)))\left[M_A\right]\\ -\left(k_{\mathrm{pAB}}\left(j\right)\right[M_B]+k_{\mathrm{tHA}}\left(j\right){\left[H_2\right]}^{0.5}+k_{\mathrm{dA}}\left(j\right)+k_{\mathrm{tAA}}\left(j\right)+{[M}_A]+k_{\mathrm{tAB}}\left(j\right)\left[M_B\right])Y_A^2\left(j\right)-\frac{q_{\mathrm{out}}{Y}_A^2\left(j\right)}{V}\end{array}$

$\begin{array}{c}\frac{d\left(Y_B^2\left(j\right)V\right)}{\mathrm{Vdt}}=\frac{q_{\mathrm{in}}{Y}_B^2{\left(j\right)}_f}{V}+\left(k_{\mathrm{iB}}\left(j\right)\right[P_0\left(j\right)]+k_{\mathrm{pBB}}\left(j\right)(2Y_B^1\left(j\right)+Y_B^0\left(j\right))+k_{\mathrm{pAB}}\left(j\right)(Y_A^1\left(j\right)+2Y_A^1\left(j\right)+Y_A^0\left(j\right)))\left[M_B\right]\\ -\left(k_{\mathrm{pBA}}\left(j\right)\right[M_A]+k_{\mathrm{tHB}}\left(j\right){\left[H_2\right]}^{0.5}+k_{\mathrm{dB}}\left(j\right)+k_{\mathrm{tBA}}\left(j\right)+{[M}_A]+k_{\mathrm{tBB}}\left(j\right)\left[M_B\right])Y_B^2\left(j\right)-\frac{q_{\mathrm{out}}{Y}_B^2\left(j\right)}{V}\end{array}$

$\begin{array}{c}\frac{d\left(X^2\left(j\right)V\right)}{\mathrm{Vdt}}=\frac{q_{\mathrm{in}}{X}^2{\left(j\right)}_f}{V}+(k_{\mathrm{tHA}}\left(j\right){\left[H_2\right]}^{0.5}+k_{\mathrm{dA}}\left(j\right)+k_{\mathrm{tAA}}\left(j\right)[M_A]+k_{\mathrm{tAB}}\left(j\right)[M_B\left]\right)Y_A^2\left(j\right)\\ +(k_{\mathrm{tHB}}\left(j\right){\left[H_2\right]}^{0.5}+k_{\mathrm{dB}}\left(j\right)+k_{\mathrm{tBA}}\left(j\right)[M_A]+k_{\mathrm{tBB}}\left(j\right)[M_B\left]\right)Y_B^2\left(j\right)-\frac{q_{\mathrm{out}}{X}^2\left(j\right)}{V}\end{array}$

Energy balance:

$\begin{array}{c}V\·{10}^{-3}\·\left(\underset{i=1}{\overset{N_m}{\mathrm{\Σ}}}\right[M_i]{\mathrm{mw}}_i{C}_{p,L,i}+[H_2]{\mathrm{mw}}_{H_2}{C}_{p,L,H_2}+(Y_A^0+Y_B^0+X^0)\·\mathrm{Mw}\·C_{p,s})\frac{\mathrm{dT}}{\mathrm{dt}}\\ ={10}^{-3}\·(q_{\mathrm{in},\mathrm{cat}}{\mathrm{\ρ}}_{\mathrm{cat}}{C}_{p,\mathrm{cat}}+q_{\mathrm{in},A}{\mathrm{\ρ}}_A{C}_{p,A}+\underset{i=1}{\overset{N_m}{\mathrm{\Σ}}}{q}_{\mathrm{in},M_i}{\mathrm{\ρ}}_{M_i}{C}_{p,L,i}+q_{\mathrm{in},H_2}{\mathrm{\ρ}}_{H_2}{C}_{p,L,H_2})(T_{\mathrm{in}}-T_{\mathrm{ref}})\\ +(\underset{i=1}{\overset{N_m}{\mathrm{\Σ}}}{F}_{R,M_i}{C}_{p,L,i}+F_{R,H_2}{C}_{p,L,H_2})(T_R-T_{\mathrm{ref}})-(F_{G,M_A}+F_{G,M_B}+G_{G,H_2}){\mathrm{\ΔH}}_v\\ -(\underset{i=1}{\overset{N_m}{\mathrm{\Σ}}}{F}_{G,M_i}{C}_{p,G,i}+F_{G,H_2}{C}_{p,G,H_2})(T-T_{\mathrm{ref}})+V\underset{i=1}{\overset{N_m}{\mathrm{\Σ}}}{R}_i\·(-\mathrm{\Δ}{H}_{r,i})\\ -q_{\mathrm{out}}\·{10}^{-3}\·\left(\underset{i=1}{\overset{N_m}{\mathrm{\Σ}}}\right[M_i]{\mathrm{mw}}_i{C}_{p,L,i}+[H_2]{\mathrm{mw}}_{H_2}{C}_{p,L,H_2}+(Y_A^0+Y_B^0+X^0)\·\mathrm{Mw}\·C_{p,s})(T-T_{\mathrm{ref}})\end{array}$

According to above-mentioned model, the method for calculation of Polymer average molecular weights, average copolymerization composition can be summarized as follows:

Monomer forms: $f_1=\frac{\left[M_A\right]}{\left[M_A\right]+\left[M_B\right]}$

The instantaneous copolymerization composition of polymkeric substance: $F_1\left(j\right)=\frac{r_A\left(j\right){f_1}^2+f_1(1-f_1)}{r_A\left(j\right){f_1}^2+2f_1(1-f_1)+r_B\left(j\right){(1-f_1)}^2}$

Polymer accumulation copolymerization composition in j active sites:

The average copolymerization composition of polymkeric substance:

Number-average molecular weight: $\mathrm{MWN}=\frac{\underset{j=1}{\overset{\mathrm{Ns}}{\mathrm{\Σ}}}{Y}_A^1\left(j\right)+Y_B^1\left(j\right)+X^1\left(j\right)}{\underset{j=1}{\overset{\mathrm{Ns}}{\mathrm{\Σ}}}{Y}_A^0\left(j\right)+Y_B^0\left(j\right)+X^0\left(j\right)}\stackrel{\‾}{m}$

Weight-average molecular weight: $\mathrm{MWW}=\frac{\underset{j=1}{\overset{\mathrm{Ns}}{\mathrm{\Σ}}}{Y}_A^2\left(j\right)+Y_B^2\left(j\right)+X^2\left(j\right)}{\underset{j=1}{\overset{\mathrm{Ns}}{\mathrm{\Σ}}}{Y}_A^1\left(j\right)+Y_B^1\left(j\right)+X^1\left(j\right)}\stackrel{\‾}{m}$

Molecular weight distributing index:

Wherein, average monomer unit molecular weight is

According to above-mentioned model, molecular weight distribution method of calculation can be summarized as follows:

One-parameter Flory most Probable distrebution is utilized to describe the instantaneous chain length distribution of each active sites of olefin polymerization process:

w _j(r)＝rτ(j) ²exp(-rτ(j))

Wherein w (r) is the chain length distribution function of weight, and r is polymkeric substance chain length, and parameter τ is the ratio of total chain tra nsfer speed and total polymerization speed.

For homopolymerization process, the τ of single active sites and rate constant, the correlation of polymerizing condition is:

$\mathrm{\τ}\left(j\right)=\frac{k_{\mathrm{tM}}\left(j\right)\left[M\right]+k_{\mathrm{tH}}\left(j\right){\left[H_2\right]}^{0.5}}{k_p\left(j\right)\left[M\right]}$

For polymerization process, then have:

$\mathrm{\τ}\left(j\right)=\frac{k_{\mathrm{tMT}}\left[M_T\right]+{\left[H_2\right]}^{0.5}}{k_{\mathrm{pT}}\left[M_T\right]}$

Wherein: k _tMT=(k _tAA(j) φ ₁(j)+k _tBA(j) φ ₂(j)) f ₁+ (k _tAB(j) φ ₁(j)+k _tBB(j) φ ₂(j)) f ₂

k _tHT＝(k _tHA(j)φ ₁(j)+k _tHB(j)φ ₂(j))

k _pT＝(k _pAA(j)φ ₁(j)+k _pBA(j)φ ₂(j))f ₁+(k _pAB(j)φ ₁(j)+k _pBB(j)φ ₂(j))f ₂

[M _T]＝[M _A]+[M _B]

${\mathrm{\φ}}_1=\frac{k_{\mathrm{pBA}}{f}_1}{k_{\mathrm{pBA}}{f}_1+k_{\mathrm{pBA}}{f}_2},{\mathrm{\φ}}_2=1-{\mathrm{\φ}}_1$

Chain length distributionly can be write as the form that independent variable(s) is logarithmic coordinates by instantaneous by mathematic(al) manipulation:

w _j(r)＝ln10·r ²τ(j) ²exp(-rτ(j))

Total chain length distribution of polymkeric substance be each active sites generate the chain length distribution superposition of polymkeric substance, as shown in the formula:

$W\left(r\right)=\underset{j=1}{\overset{\mathrm{Ns}}{\mathrm{\Σ}}}m\left(j\right)w_j\left(r\right)$

Above formula being write as the conventional molecular weight distribution phraseology of testing gained with gel permeation chromatography is:

x＝log M _w

$\frac{\mathrm{dWt}}{d\left(\log M_w\right)}=2.3\·\underset{j=1}{\overset{\mathrm{Ns}}{\mathrm{\Σ}}}m\left(j\right)\frac{1}{\stackrel{\‾}{M_n}{\left(j\right)}^2}\exp (4.6x-\frac{1}{\stackrel{\‾}{M_n}\left(j\right)}\exp \left(2.3x\right))$

Copolymerization composition distribution calculation method is as follows:

Utilize Stockmayer bivariate distribution function to describe the instantaneous copolymerization composition single active sites generating multipolymer to distribute

$w_j\left(F\right)=\frac{3}{4\sqrt{2\mathrm{\β}\left(j\right)\mathrm{\τ}\left(j\right)}{[1+{(F-\stackrel{\‾}{F_i}\left(j\right))}^2/\left(2\mathrm{\β}\left(j\right)\mathrm{\τ}\left(j\right)\right)]}^{5/2}}$

Wherein, the definition of parameter τ is with identical above, and parameter beta following formula is obtained:

$\mathrm{\β}\left(j\right)=\stackrel{\‾}{F_i}\left(j\right)(1-\stackrel{\‾}{F_i}\left(j\right)){[1+4\stackrel{\‾}{F_i}\left(j\right)(1-\stackrel{\‾}{F_i}\left(j\right))(r_A\left(j\right)r_B\left(j\right)-1)]}^{0.5}$

F be i monomeric unit to account in j active sites generate the mole fraction of multipolymer, represent i monomeric unit and account for molar average point rate j active sites generating multipolymer, r _a(j) and r _bj () is the unexpectedly poly-rate of monomer A and monomers B in j active sites respectively.Total copolymerization composition distribution CCD of multipolymer is the superposition of the multipolymer copolymerization composition distribution that each active sites generates, as shown in the formula:

$W_t\left(F\right)=\underset{j=1}{\overset{\mathrm{Ns}}{\mathrm{\Σ}}}m\left(j\right)w_j\left(F\right)$

Generally speaking, the input variable of above-mentioned model is q _{in, cat}, q _{in, A}, q _out, q _in, q _{in, M}, q _{in, H2}, [A] _f, [C _p] _f, F (j), [H ₂] _f, [M] _f, m _wA, m _wB, Ns, the initial value of each kinetic constant and state variables.The value being directly output as each state variables of t of model, such as [C _p], [A], [H ₂], [M _a], [M _b], [P ₀(j)], X ⁰(j), Y _a ⁰(j), Y _b ⁰(j), X ^-1(j), Y _a ¹(j), Y _b ¹(j), X ^-2(j), Y _a ²(j), Y _a ²(j) etc.The final output of model is weight-average molecular weight MWW, number-average molecular weight MWN, the dispersion index PDI and molecular weight distribution MWD of polymkeric substance, average copolymerization composition and copolymerization composition distribution CCD.

2) objective function of polyolefine molecular-weight average or molecular chain conformation is set up.

1., time using the polyolefinic molecular-weight average of new grades as optimization aim, objective function is:

min(|MWW _cal-MWW _obj|/MWW _obj)

Wherein MWW _calfor calculating the polyolefine weight-average molecular weight of gained, MWW _objfor the polyolefinic weight-average molecular weight of the target trade mark, wherein the polyolefinic weight-average molecular weight of the target trade mark is known quantity.

2. time using the polyolefinic molecular chain conformation of new grades as optimization aim, molecular weight distribution curve is divided into N sheet, the sampling point number namely on molecular weight distribution curve is N, and the value of N number of point makes chain length be decile in logarithmic coordinates.Optimization object function is as follows:

$\min \underset{k=1}{\overset{N_{\mathrm{point}}}{\mathrm{\Σ}}}\left[\right|W_{\mathrm{cal}}\left(\log {M_W}^k\right)-W_{\mathrm{obj}}\left(\log {M_W}^k\right)|/W_{\mathrm{obj}}\left(\log {M_W}^k\right)]$

Wherein W _obj(logM _w ^k) represent the value of the polyolefinic molecular weight distribution curve of the target trade mark at k point place, be known quantity, W _cal(logM _w ^k) be the value of polyolefinic molecular weight distribution curve at k point place that model optimization obtains, k represents the numerical value of taken point on curve.N _pointrepresentative is always counted.

3) determine the constraint condition of polymerization technique, comprise the upper lower limit value of temperature of reaction, the upper lower limit value of reaction pressure, the upper lower limit value of vapor phase hydrogen and monomer mole ratio, reactive system removes higher limit of heat energy power etc.

4) the optimum polymerization process condition meeting new grades quality product and require is solved.Design variable is hydrogen and monomer mole ratio in reactor, first arranges the initial value of systematic polymerization processing condition, and polymerization process condition is the temperature of arbitrary stage reactor, pressure, hydrogen and monomer mole ratio.For hydrogen in reactor and monomer mole ratio, the design variable value of trade mark product can be had as its initial value according to device.Then, according to the value of polymerization process condition, obtain material balance before utilization and energy balance model calculates polyolefine molecular-weight average or molecular weight distribution.As the polyolefine molecular-weight average that calculates or molecular weight distribution meet the optimization tolerance preset, then using the value of polymerization process condition now as optimizing the best polymerization process condition that obtains, otherwise the value of amendment polymerization process condition, returns step 3) and carries out iteration until calculate the polyolefine molecular-weight average or molecular weight distribution that meet the optimization tolerance preset.The maximum iteration time of system solution device can be set as 100 times, and the tolerance that setting is optimized can be 0.0001, and iterative algorithm can adopt SQP(SQP) Algorithm for Solving.

Being below produce high density polyethylene(HDPE) (HDPE) for object with the two still series polymerizations reactor of ethene slurry, according to the steady-state model of present method process of establishing, take molecular-weight average as the optimum result of the ethene slurry polymerization two still series connection Production Flow Chart of target.Namely optimization aim is the poly weight-average molecular weight of new grades (105316gm/mol), and determine the hydrogen in recycle gas and ethylene molar ratio, optimum result is as follows:

1) objective function: min (| MWW _cal-MWW _obj|/MWW _obj)

Wherein MWW _calfor the polyethylene weight-average molecular weight of computation optimization gained, MWW _objfor target trade mark polyethylene weight-average molecular weight.

2) design variable: the mol ratio of hydrogen and ethene in the first reaction member and the second reaction member recycle gas.

Table 2 take molecular-weight average as the optimum result of the ethene slurry polymerization two still series connection Production Flow Chart of target

Table 2 is optimum result and the comparing of analytical value.As seen from table, the HDPE weight-average molecular weight after optimization and the relative error of analytical value only 0.0028%, but the relative error of dispersion index PDI and analytical value comparatively large (32.87%).Fig. 3 is the optimal value of molecular weight of polyethylene distribution curve and comparing of analytical value.As seen from the figure, the molecular weight distribution curve after optimization and the deviation of analytical value are comparatively large, and maximum deviation reaches 0.12.Although the weight-average molecular weight of product reaches requirement, dispersion index and molecular weight distribution and the target trade mark require to still have relatively large deviation.

Below changing with molecular weight distribution curve is target, and carry out polymerization process condition optimization to the said products, optimum result is as follows:

1) objective function: poly molecular weight distribution curve is divided into 100, the sampling point number namely on molecular weight distribution curve is that the value of 100,100 points makes chain length be decile in logarithmic coordinates.Optimization object function is as follows:

$\min \underset{k=1}{\overset{N_{\mathrm{point}}}{\mathrm{\Σ}}}\left[\right|W_{\mathrm{cal}}\left(\log {M_W}^k\right)-W_{\mathrm{obj}}\left(\log {M_W}^k\right)|/W_{\mathrm{obj}}\left(\log {M_W}^k\right)]$

Wherein W _obj(logM _w ^k) represent the value of the poly molecular weight distribution curve of the target trade mark at k point place, W _cal(logM _w ^k) be the value of poly molecular weight distribution curve at k point place that model optimization obtains, k represents the numerical value of the point that curve is got.N _pointrepresentative is always counted.

2) design variable: the set(ting)value of controller H2/C2H4_C1 and H2/C2H4_C2, the i.e. mol ratio of hydrogen and ethene in the first still and the second still circulation gas.

Table 3 take molecular weight distribution as the optimum result of the ethene slurry polymerization two still series connection Production Flow Chart of target

Table 3 is the optimum result of the ethene slurry polymerization two still series connection Production Flow Chart taking molecular weight distribution as target.As seen from table, during using molecular weight distribution as optimization aim, calculate the weight-average molecular weight of gained and molecular weight distributing index consistent with analytical value, in decision variable and recycle gas hydrogen and ethylene molar ratio and analytical value also more identical.Fig. 4 calculates the molecular weight distribution of gained and comparing of analytical value with Different Optimization target.From figure, during using molecular weight distribution as optimization aim, the analogue value of poly molecular weight distribution curve and analytical value are substantially identical.In addition, can find from Fig. 5, time in two kinds of optimization methods using molecular weight distribution as optimization aim, the residual error of gained is less, and maximum deviation only has 0.02.

Nomenclature table

Although the present invention with preferred embodiment openly as above, embodiment is not of the present invention for limiting.Without departing from the spirit and scope of the invention, any equivalence change done or retouching, belong to the protection domain of the present invention equally.Therefore the content that protection scope of the present invention should define with the claim of the application is standard.

Claims (10)

1., with the polymerization process condition optimization method that polyolefine micro-quality is target, it is characterized in that, comprise the following steps:

1) set up the mathematical model of olefin polymerization process, comprise the steps:

1-1) determine the form of the reactor of paradigmatic system, working method, each reactor is regarded as the combination of continuous stirred tank reactor or continuous stirred tank reactor;

1-2) according to form, the working method of the reactor determined, select the thermodynamical model of reactive system and determine its parameter;

1-3) on the basis of the form of the reactor determined, working method and thermodynamical model, determine polymerization kinetics mechanism and parameter thereof;

1-4) according to thermodynamical model and the polymerization kinetics mechanism of reactive system, set up material balance and the energy balance model of polymerization process, calculated by material balance and energy balance model comprise transient molecular amount and distribution thereof, copolymerization forms and distributes, accumulation molecular weight and distribution, copolymerization composition and be distributed in interior polyolefinic molecular structure information;

2) using the target that polyolefine molecular-weight average or molecular weight distribution are optimized as polymerization process condition, the initial value of setting polymerization process condition;

3) according to the value of polymerization process condition, the material balance utilizing step 1) to obtain and energy balance model calculate polyolefine molecular-weight average or molecular weight distribution;

4) the polyolefine molecular-weight average calculated as step 3) or molecular weight distribution meet the optimization tolerance preset, then enter step 5), otherwise the value of amendment polymerization process condition, returns step 3) and carries out iteration until calculate the polyolefine molecular-weight average or molecular weight distribution that meet the optimization tolerance preset;

5) using the best polymerization process condition that the value of polymerization process condition now obtains as optimization, method ends.

2. the polymerization process condition optimization method that is target with polyolefine micro-quality according to claim 1, it is characterized in that, described olefin polymerization process is alkene homopolymerization or polymerization process.

3. the polymerization process condition optimization method that is target with polyolefine micro-quality according to claim 1, it is characterized in that, described paradigmatic system at least comprises a reactor and related procedure equipment.

4. the polymerization process condition optimization method that is target with polyolefine micro-quality according to claim 1, is characterized in that, the working method of described reactor is interval, semicontinuous or continuous.

5. the polymerization process condition optimization method that is target with polyolefine micro-quality according to claim 1, it is characterized in that, the implementation method of described olefin polymerization process is slurry phase, body phase or gas phase.

6. the polymerization process condition optimization method that is target with polyolefine micro-quality according to claim 1, it is characterized in that, described material balance and energy balance model, is stable state or the dynamicmodel of single or multiple reactors in series.

7. the polymerization process condition optimization method that is target with polyolefine micro-quality according to claim 1, it is characterized in that, described material balance and energy balance model, can calculate and follow the tracks of the polymer molecular structure information of reactive system random time, arbitrarily reactor.

8. the polymerization process condition optimization method that is target with polyolefine micro-quality according to claim 1, is characterized in that, described polymerization process condition is the temperature of reactor, pressure, hydrogen and monomer mole ratio.

9. the polymerization process condition optimization method that is target with polyolefine micro-quality according to claim 8, it is characterized in that, described step 2) in, the value of hydrogen and monomer mole ratio in the reactor having trade mark product according to device, the initial value of hydrogen and monomer mole ratio in setting reactor.

10. the polymerization process condition optimization method that is target with polyolefine micro-quality according to claim 1, it is characterized in that, described optimization tolerance is 0.0001.