Intrinsically safe design method for polyethylene homopolymerization reaction process in combination with product property requirements

Abstract

The invention belongs to the technical field of design optimization of intrinsically safe chemical processes, polymerization product properties and reaction processes, and discloses a polyethylene homopolymerization reaction process intrinsically safe design method combining product property requirements. Giving polymer property constraint and quality constraint, respectively carrying out single-target optimization on the annual income and the chemical fire explosion index, and respectively taking annual income maximization and negative value maximization as targets to obtain the maximum annual income and the minimum chemical fire explosion index; and carrying out multi-objective optimization on negative values of the annual income and the chemical fire explosion index, and combining a plurality of objectives into a function by adopting a weight coefficient normalization strategy for each objective. A nonlinear programming NLP model is formed by each target, a property constraint and a quality constraint, and the NLP model is solved to finally obtain a polymer product capable of meeting the property requirement and a comprehensive optimal production condition capable of meeting the safety and economy requirements of a polymer production process.

Description

Technical Field

[0001] The present invention relates to the technical field of design optimization of intrinsically safe chemical processes, polymerization product properties and reaction processes, and in particular to an intrinsically safe design method for a polyethylene homopolymerization process in combination with product property requirements. Background Art

[0002] With the rapid development of social economy, people's demand for high-performance polymers for special application scenarios is increasing. On the basis of meeting the requirements of polymer properties, people also hope that its production process is safe and reliable, and the incidence of accidents such as combustion and explosion is reduced. Compared with a single low-molecular reaction process, the polymerization reaction process has a more complex reaction mechanism and kinetic relationship. When the operating conditions change, it will not only change the reaction rate of the polymerization reaction, but also affect the molecular weight distribution, copolymer composition and sequence distribution of the polymerization product. Due to the complexity of polymer structure, the production process research of traditional polymer new products is usually based on the corresponding reaction kinetics to model the reaction process, and obtain polymer products with target properties by adjusting various operating parameters. However, this method consumes manpower and material resources, and if products that cannot meet the needs are produced, it will also lead to waste of raw materials.

[0003] Slurry polymerization is a highly efficient polymerization process conducted in a liquid medium. Its well-researched methods are widely used in the chemical and petrochemical industries, particularly in the production of polyolefins such as polyethylene and polypropylene. Its advantages in large-scale industrial production make it one of the most commonly used polymerization methods. The five-active-site ethylene slurry polymerization reaction mechanism is a model describing how ethylene reacts with a catalyst in slurry polymerization and is ultimately converted into polyethylene. This mechanism involves multiple active sites on the catalyst surface, through which ethylene molecules react with the catalyst to form various intermediates and ultimately the polymer.

[0004] Intrinsically safe design eliminates or significantly reduces potential hazards and prevents accidents by optimizing engineering, equipment selection, and operating conditions. Its core concept is to fundamentally alter equipment or system design so that, under any circumstances, even if a malfunction or anomaly occurs, it will not cause dangerous events such as fire, explosion, or hazardous substance release. Design requires comprehensive consideration of materials, equipment, process conditions, and technical means to ensure safety during production, storage, and transportation. Intrinsically safe design in industries such as chemical and oil and gas reduces risks of fire and explosion by optimizing process flows, selecting appropriate materials and equipment, and adopting low-temperature and low-pressure reaction conditions.

[0005] Intrinsically safe design helps eliminate or reduce hazards at the source, optimize processes and equipment, and ensure safe and stable equipment operation. Especially in high-risk industries, it can significantly improve safety, reduce risks, and minimize long-term economic losses.

[0006] The Dow Chemical Fire and Explosion Index (F&EI) is an indicator used to assess the potential hazards and risks of chemical processes or plant facilities in the event of a fire or explosion. It is often used in intrinsically safe design. As an intrinsically safe indicator, it has clear and well-defined guidelines and can obtain a certain safety score. The F&EI has been widely used in the chemical industry. It is a comprehensive indicator that combines multiple factors, such as the flammability and explosion hazard of the substance, and the degree of hazardousness of equipment and facilities, to assess the potential risk of fire and explosion. Generally, a larger F&EI value indicates a greater risk of fire or explosion in the system or equipment. Because polymerization reactions are inherently exothermic and the raw materials are inherently toxic, flammable, and explosive, the hazards of the reaction process are primarily related to the exothermic nature of the reaction. The hazards of reactants and products are related to changes in operating conditions. Therefore, the Dow Chemical Fire and Explosion Index (F&EI) evaluation method can be incorporated into the intrinsic design of polymerization processes to reduce the hazards of the reaction process and reduce the investment in protective devices in subsequent production processes.

[0007] CN112384539B discloses a method for producing a polyethylene polymer suitable for injection or compression molded articles, particularly for the manufacture of lids or closures, and relates to the polyethylene polymer itself, injection or compression molded articles comprising the polymer, and the use of the polymer for the production of injection or compression molded articles (e.g., lids and closures). The method comprises three different polymerization stages to produce multimodal high-density polyethylene with a range of specific properties, which enables the formation of molded articles having advantageous characteristics in terms of good processability and resistance to cracking. Although this method has many advantages in terms of the design of polyethylene properties, it lacks consideration in terms of the intrinsic safety design of the polyethylene production process, resulting in a high investment in safety protection devices after actual production.

[0008] CN118313214A discloses an intrinsically safe design method for a plug flow reactor based on process safety time. This method uses a computer-aided method based on a nonlinear programming model of process safety time to design a chemical plug flow reactor. Maximizing the output of a target product is used as the objective function in a subsequent NLP model. An ε-constraint algorithm is used to balance conflicts between multiple objectives, and the multi-objective problem is converted into a series of single-objective problems. The Dow Chemical Fire and Explosion Index (F&EI) is transferred to the constraint condition as a safety evaluation indicator. The introduction of the Dow Chemical Fire and Explosion Index (F&EI) expands the intrinsically safer design range of the plug flow chemical reactor. Summary of the Invention

[0009] Therefore, the present invention aims to provide a polyethylene homopolymerization process intrinsic safety design method combined with product property requirements. The method is based on the relevant reaction mechanism and reaction kinetics of the polymerization reaction, focuses on the reaction process of polymer production, and introduces the Dow Chemical Fire and Explosion Index (F&EI) evaluation method while combining the polymer property prediction model to quantify the actual potential fire, explosion hazards, and reaction hazards of the polymerization reaction system process equipment and the materials contained therein. A polyethylene homopolymerization process intrinsic safety design method combined with product property requirements is proposed. Based on the given polymer property requirements, the annual return Φ and the Dow Chemical Fire and Explosion Index (F&EI) are firstly optimized by single objectives, with the objectives of maximizing the annual return Φ and maximizing the negative value of the F&EI, respectively, to obtain the maximum annual return Φ and the minimum Dow Chemical Fire and Explosion Index (F&EI). Then, the annual return Φ and the negative value of the Dow Chemical Fire and Explosion Index (F&EI) are optimized by multiple objectives. By adopting a weight coefficient for each objective along the normalization strategy, multiple objectives are combined into one function. The various objectives, quality constraints, and property constraints constitute a nonlinear programming NLP model. By solving the NLP model, we can ultimately obtain the comprehensive optimal production conditions that not only meet the property requirements of the polymer product, but also meet the safety and economic requirements of the polymer production process.

[0010] This method requires the simultaneous introduction of a polymer product design model and a reaction model to identify common parameters between the two models. Since slurry polymerization methods are well-established and widely applicable, the present invention establishes a reactor model based on existing polyolefin reaction kinetics. In combination with a polymer property prediction model, the number-average molecular weight is used as a variable to connect the polymer product design model and the reaction model. Furthermore, the Dow Chemical Fire and Explosion Index (F&EI) evaluation method is introduced into the reaction process design to establish an intrinsically safe design method for polyethylene homopolymerization processes that incorporates product property requirements. The present invention uses the polyethylene Ziegler-Natta homopolymerization process as an example to demonstrate the characteristics of this design method.

[0011] To achieve the above object, the technical solution of the present invention is as follows:

[0012] A method for designing intrinsic safety of a polyethylene homopolymerization process in accordance with product property requirements specifically comprises the following steps:

[0013] Step 1: Determine the reaction rate of each substance in the reaction process according to the reaction kinetics equation of the ethylene slurry homopolymerization process:

[0014] Based on the ethylene slurry polymerization reaction mechanism with five active sites, the reaction kinetic equation of the ethylene slurry homopolymerization process includes catalyst activation, chain initiation, chain growth, chain transfer and chain deactivation reactions. The specific equation is as follows:

[0015] C P(j)+A→P0(j) (1)

[0016] P0(j)+M→P1(j) (2)

[0017] P n (j)+M→P n+1 (j) (3)

[0018] P n (j)+M→P1(j)+D n (j) (4)

[0019] P n (j)+H2→P0(j)+D n (j) (5)

[0020] P n (j)+A→P0(j)+D n (j) (6)

[0021] P n (j)→P0(j)+D n (j) (7)

[0022] P n (j)→C d (j)+D n (j) (8)

[0023] P0(j)→C d (j) (9)

[0024] Among them, C P (j), P0(j), C d (j) are the potential active sites, empty active sites, and deactivated active sites on the jth active site on the catalyst, A is the co-catalyst, M is the homomonomer, and P is the n (j) is the active polymer with a chain length of n on the jth active site on the catalyst, D n (j) is the dead polymer with chain length n on the jth active site on the catalyst, k aA (j), k i (j), k p (j), k tM (j), k tH (j), k tA (j), k t (j), k d (j) Reaction kinetic constants corresponding to catalyst activation, chain initiation, chain extension, transfer to ethylene monomer, transfer to hydrogen, transfer to cocatalyst, transfer to β-hydride, and chain deactivation reactions, respectively;

[0025] The reaction kinetic constant is temperature-dependent and is calculated using the Arrhenius law:

[0026]

[0027] Where k0 is the pre-exponential factor, E a is the reaction activation energy, R is the ideal gas constant, T is the reactor temperature, T ref is the reaction reference temperature;

[0028] Step 2: Optimizing the ethylene homopolymerization process based on mass constraints and property constraints;

[0029] Step 3: Perform intrinsically safe design and optimize the ethylene homopolymerization process. First, single-objective optimization is performed on the annual return Φ and the Dow Chemical Fire and Explosion Index (F&EI), respectively, with the goals of maximizing the annual return Φ and maximizing the negative value of the F&EI, to obtain the maximum annual return Φ and the minimum Dow Chemical Fire and Explosion Index (F&EI). Multi-objective optimization is performed on the annual return Φ and the negative value of the Dow Chemical Fire and Explosion Index (F&EI). By applying a weight coefficient normalization strategy to each objective, multiple objectives are combined into a single function, and the maximum value of this function is optimized as the objective. Each objective, along with quality constraints and property constraints, forms a nonlinear programming (NLP) model. By solving the NLP model, the optimal production conditions are ultimately obtained that meet both the property requirements of the polymer product and the safety and economic requirements of the polymer production process:

[0030] The final goal is as follows:

[0031]

[0032] Among them Ф up and F&EI up are Φ and F&EI obtained when the goal is to maximize annual income Φ, Φ low and F&EI low are Φ and F&EI obtained when the negative value of F&EI is maximized, and α is the weight coefficient of multi-objective optimization;

[0033] Step 4: According to the set upper and lower limit constraints of the variables, the GAMS solver is used to solve the established NLP model.

[0034] The reaction processes involved in the ethylene slurry homopolymerization process all consider the following assumptions:

[0035] (1) The CSTR is operated with a full reactor, with only liquid phase in the reactor, and the effect of gas-liquid phase equilibrium on the reaction process is ignored;

[0036] (2) The reactor is an ideal reactor, and the components are evenly distributed in the reactor during the reaction;

[0037] (3) Ignoring the diffusion effect in the reactor, the temperature in the reactor is uniform;

[0038] (4) Under the equal-length assumption of polymerization reaction, the reaction kinetics of chain growth, chain transfer and chain deactivation are independent of chain length.

[0039] The reaction rates of each substance in the reaction process are as follows:

[0040] Define the combined reaction rate constant of chain transfer and chain deactivation: K TD (j)

[0041] K TD (j) = k tM (j)[M]+k tH (j)([H2]) 0.5 +k tA (j)[A]+k t (j)+k d (j) (12)

[0042] Where [] represents the concentration of each substance in the reactor;

[0043] The reaction rate of living polymer and dead polymer with chain length 1:

[0044]

[0045] where Y 0 (j) is the zero-order moment of the living polymer on the jth active site on the catalyst;

[0046] Reaction rate of living polymer with dead polymer whose chain length is greater than 2:

[0047]

[0048] The reaction rates of the zero-order, first-order, and second-order moments of the living polymer and the dead polymer at the j-th active site on the catalyst are:

[0049]

[0050] The reaction rate equation of potential active sites, empty active sites, and deactivated active sites on the jth active site of the catalyst is:

[0051]

[0052] The mass conservation constraints of each substance in the ethylene slurry homopolymerization process are as follows:

[0053] The total amount of solvent in the input and output materials remains unchanged, and no mass conservation constraints are imposed on the solvent;

[0054] Mass conservation constraints for ethylene monomer, hydrogen, and co-catalyst:

[0055]

[0056] Where V is the volume of the reactor, F in and F are the total amount of input and output materials respectively;

[0057] Mass conservation constraints for potential active sites, empty active sites, and deactivated active sites on the jth active site on the catalyst:

[0058]

[0059] The λ-order moment mass conservation constraint for living and dead aggregates:

[0060]

[0061] Property constraints include molecular weight constraints, glass transition temperature constraints, and melt index constraints:

[0062] Number average molecular weight constraint:

[0063]

[0064] Among them, M n is the number average molecular weight, mw is the relative molecular mass of ethylene, 28.4 g / mol.

[0065] Weight average molecular weight constraint:

[0066]

[0067] Among them, M w is the weight average molecular weight;

[0068] Polyethylene mass flow constraints:

[0069]

[0070] Where: W PE is the mass flow rate of polyethylene product:

[0071] Glass transition temperature constraints:

[0072]

[0073] Melt index constraints:

[0074] MFR=α1·(M w ) -β (38)

[0075] Among them, T g (∞) is the ultimate glass transition temperature, K gis a constant related to the type of polymer. For polyethylene, its value is 15000K·g / mol. α1 and β1 are constants. For polyethylene, their values ​​are 4.195×10 19 The sum is -3.9252.

[0076] The single target calculation in step 3 is as follows:

[0077] The calculation formula for annual return Φ is as follows:

[0078]

[0079] Where a is the selling price of polyethylene, b is M 、b H2 、b c 、b A are the unit prices of ethylene monomer, hydrogen, catalyst, and co-catalyst respectively; ΔW M , ΔW H2 , ΔW C , ΔW A are the consumption of each raw material; t is the production time, 8000h / year, c V The cost of the reactor unit;

[0080] Dow Chemical Fire and Explosion Index (F&EI) calculation equation:

[0081] F&EI=MF·F3 (40)

[0082] Calculation equation for process unit risk factor F3:

[0083] F3=F1·F2 (41)

[0084] Among them, F1 is generally the process hazard factor, and F2 is the special process hazard factor.

[0085] Step 4 specifically involves building an NLP model in GAMS, using the CONOPT solver to solve the problem, and obtaining the results of each single-objective and multi-objective optimization.

[0086] This intrinsically safe design approach will provide the following information:

[0087] The values ​​of reactor volume, polyethylene product output, annual revenue Φ and F&EI.

[0088] The beneficial effects of the present invention are as follows: Typically, research on the production process of new polymer products is based on reaction kinetics to model the reaction process, and polymer products with target properties are obtained by adjusting various operating parameters. In short, this is a sequential simulation-testing-verification process. This method is usually time-consuming and labor-intensive, and the uncertain products also result in waste of raw materials. The beneficial effects of the present invention are that computer-aided molecular design is used to reduce the time required for developing new polymer brands and the waste of reaction raw materials. At the same time, the introduction of the safety index as one of the final optimization goals improves the safety and reliability of the reaction system. The present invention establishes a reactor model based on the existing polyolefin reaction kinetics mechanism, combines it with a polymer property prediction model, and uses the number-average molecular weight as a variable connecting the polymer product design model and the reaction model. At the same time, the Dow Chemical Fire and Explosion Index (F&EI) evaluation method is introduced into the reaction process design to quantify the actual potential fire, explosion, and reaction hazards of the polymerization reaction system process equipment and the materials contained therein. A method for designing the intrinsically safe polyethylene homopolymerization process that combines product property requirements is proposed. After optimization, the optimal production conditions are ultimately obtained that meet both the property requirements of the polymer product and the safety and economic requirements of the polymer production process. BRIEF DESCRIPTION OF THE DRAWINGS

[0089] Figure 1 Schematic diagram of the intrinsically safe design method for the polyethylene homopolymerization process in combination with product property requirements. DETAILED DESCRIPTION

[0090] like Figure 1 As shown, a method for designing intrinsic safety of a polyethylene homopolymerization process in combination with product property requirements specifically includes the following steps:

[0091] Step 1: Determine the reaction rate of each substance in the reaction process according to the reaction kinetics equation of the ethylene slurry homopolymerization process;

[0092] The present invention is based on the ethylene slurry polymerization reaction mechanism of five active sites. The reaction kinetic equation of the ethylene slurry homopolymerization process includes catalyst activation, chain initiation, chain growth, chain transfer and chain deactivation reactions, wherein the chain transfer reaction includes the transfer reaction from the living polymer to the monomer, hydrogen, co-catalyst, and β-hydride. The specific equation is as follows:

[0093] Catalyst activation reaction:

[0094] C P (j)+A→P0(j) (1)

[0095] Among them C P (j) is the potential active site on the jth active site of the catalyst, A is the co-catalyst, P0(j) is the empty active site on the jth active site of the catalyst, and the reaction kinetic constant of the catalyst activation reaction is recorded as kaA (j).

[0096] Chain initiation reaction:

[0097] P0(j)+M→P1(j) (2)

[0098] Where M is the homomonomer of the polymerization reaction, P1(j) is the living polymer with a chain length of 1 on the jth active site of the catalyst, and the reaction kinetic constant of the chain initiation reaction is denoted as k i (j).

[0099] Chain growth reaction:

[0100] P n (j)+M→P n+1 (j) (3)

[0101] Among them, P n (j) is the living polymer with chain length n on the jth active site on the catalyst, P n+1 (j) is the living polymer with a chain length of n+1 on the jth active site of the catalyst, and the reaction kinetic constant of the chain growth reaction is recorded as k p (j).

[0102] Transfer reaction to ethylene monomer:

[0103] P n (j)+M→P1(j)+D n (j) (4)

[0104] Among them D n (j) is the dead polymer with chain length n on the jth active site of the catalyst, and the reaction kinetic constant of the transfer reaction to ethylene monomer is recorded as k tM (j).

[0105] Shift reaction to hydrogen:

[0106] P n (j)+H2→P0(j)+D n (j) (5)

[0107] Where H2 is hydrogen, the reaction kinetic constant of the transfer reaction to hydrogen is recorded as k tH (j).

[0108] Transfer reaction to co-catalyst:

[0109] P n (j)+A→P0(j)+D n (j) (6)

[0110] The reaction kinetic constant of the transfer reaction to the co-catalyst is denoted as k tA (j).

[0111] β-Hydride transfer reaction:

[0112] P n (j)→P0(j)+D n (j) (7)

[0113] The kinetic constant of the transfer reaction to β-hydride is denoted as k t (j).

[0114] Chain inactivation reaction:

[0115] P n (j)→C d (j)+D n (j) (8)

[0116] P0(j)→C d (j) (9)

[0117] Among them C d (j) is the deactivated active site on the jth active site of the catalyst, and the reaction kinetic constant of the chain deactivation reaction is recorded as k d (j).

[0118] The reaction kinetic constant is related to temperature and can be calculated by Arrhenius law. The calculation equation is as follows:

[0119]

[0120] Where k0 is the pre-exponential factor, E a is the activation energy of the reaction, R is the ideal gas constant, T is the reactor temperature, T ref is the reference temperature of the reaction; in the case analysis of the present invention, the reference temperature of the reaction is 358.15K.

[0121] Step 2: Optimizing the ethylene homopolymerization process based on mass constraints and property constraints;

[0122] Step 3: Perform intrinsic safety design and optimize the ethylene homopolymerization process. The present invention first performs single-objective optimization on the annual return Φ and the Dow Chemical Fire and Explosion Index F&EI, respectively, with the goals of maximizing the annual return Φ and maximizing the negative value of F&EI, to obtain the maximum annual return Φ and the minimum Dow Chemical Fire and Explosion Index F&EI. Then, multi-objective optimization is performed on the annual return Φ and the negative value of the Dow Chemical Fire and Explosion Index F&EI. By using a weight coefficient for each target along a normalized strategy, multiple targets are combined into a function, and the maximum value of the function is used as the target for optimization. Each target, quality constraint, and property constraint constitute a nonlinear programming NLP model. By solving the NLP model, the optimal production conditions that can meet both the property requirements of the polymer product and the safety and economic requirements of the polymer production process are finally obtained;

[0123] The final goal is as follows:

[0124]

[0125] Among them Ф up and F&EI up are Φ and F&EI obtained when the goal is to maximize annual income Φ, Φ low and F&EI low are Φ and F&EI obtained when the negative value of F&EI is maximized, and α is the weight coefficient of multi-objective optimization.

[0126] Step 4: Based on the set upper and lower limit constraints of the variables, the GAMS solver is used to solve the established NLP model;

[0127] Furthermore, this method selects an ethylene slurry polymerization mechanism based on five active sites, the comonomer does not participate in the reaction, and a CSTR (ideal continuous stirred tank) is selected as the reactor. To simplify the actual process, the following assumptions are considered:

[0128] (1) The CSTR is operated with a full reactor, with only liquid phase in the reactor, and the effect of gas-liquid phase equilibrium on the reaction process is ignored;

[0129] (2) The reactor is an ideal reactor, and the catalyst, solvent, feed and other components are evenly distributed in the reactor during the reaction;

[0130] (3) Ignoring the diffusion effect in the reactor, the temperature in the reactor is uniform;

[0131] (4) Under the equal-length assumption of polymerization reaction, the reaction kinetics of chain growth, chain transfer and chain deactivation are independent of chain length.

[0132] The reaction rates of each substance in the reaction process are as follows:

[0133] According to the above polyethylene reaction kinetic equation and reaction kinetic constant, the comprehensive reaction rate constant of chain transfer and chain deactivation is defined as K TD (j), as shown in Equation (12), the reaction rates of living polymers, dead polymers and catalyst sites can be determined from the combined reaction rate constants of chain transfer and chain deactivation.

[0134] K TD (j) = k tM (j)[M]+k tH (j)([H2]) 0.5 +k tA (j)[A]+k t (j)+k d (j) (12)

[0135] Wherein [M] is the concentration of monomer in the reactor, [H2] is the concentration of hydrogen in the reactor, and [A] is the concentration of co-catalyst in the reactor.

[0136] The reaction rate r of the living polymer chain length is 1 P1(j) :

[0137]

[0138] The reaction rate r of the dead polymer chain length is 1 D1(j) :

[0139]

[0140] Among them, [P o (j)] is the concentration of empty active sites on the jth active site of the catalyst in the reactor, [P1(j)] is the concentration of active polymers with a chain length of 1 on the jth active site of the catalyst in the reactor, and Y 0 (j) is the zero-order moment of the active polymer at the jth active site of the catalyst.

[0141] When the chain length is greater than or equal to 2, the reaction rate r Pn(j) :

[0142]

[0143] When the chain length is greater than or equal to 2, the reaction rate of the dead polymer r Dn(j) :

[0144]

[0145] Among them [P n-1 (j)] is the concentration of the living polymer with chain length n-1 on the jth active site of the catalyst in the reactor, [P n (j)] is the concentration of the living polymer with chain length n on the jth active site of the catalyst in the reactor.

[0146] Since the composition range of polymer chains is relatively wide, directly using the above rate calculation equation will bring about a huge amount of calculation. At the same time, through the analysis of the properties of polyethylene products, it is found that most product indicators are related to the molecular weight average of the number average molecular weight and weight average molecular weight. Therefore, the present invention uses the average molecular weight instead of the molecular weight distribution for calculation, and uses the moment method to describe the distribution of different chains during the polymerization reaction, ignoring the specific content information of each chain length, and adding and merging the reaction rate equations of different chain lengths step by step to quickly obtain the statistical information of the polymer and reduce the amount of calculation. The present invention uses the moment method to calculate the number average molecular weight and weight average molecular weight of the polymer. For a given distribution function f(n), its λ-order moment μ λ The basic form is as follows:

[0147]

[0148] In the present invention, the λ moment of the active polymer is expressed as equation (18):

[0149]

[0150] where Y λ (j) represents the λ moment of the active polymer at the jth active site of the catalyst.

[0151] The λ moment of dead aggregate is shown in equation (19):

[0152]

[0153] where X λ (j) represents the λ moment of the dead aggregate at the jth active site of the catalyst.

[0154] According to the definition of the above moment equation, the reaction rate equations of the zero-order, first-order and second-order moments of living polymers and dead polymers can be defined:

[0155] The reaction rates of the zero-order, first-order, and second-order moments of the active polymer at the j-th active site of the catalyst are as follows:

[0156]

[0157] Among them, the zero-order moment reaction rate of the living polymer represents the rate of change of the total number of living polymer chains, that is, the generation or consumption of active centers in the system; the first-order moment reaction rate of the living polymer represents the rate of change of the total molecular weight in the living chain, which is closely related to the chain growth process; the second-order moment reaction rate of the living polymer represents the rate of change of the sum of the squares of the molecular weight of the living chain, which is directly related to the change of the molecular weight distribution.

[0158] The reaction rates of the zero-order, first-order, and second-order moments of the dead polymer at the j-th active site of the catalyst are shown in the following equations:

[0159]

[0160] Among them, the zero-order moment reaction rate of dead polymers represents the rate of change of the total number of terminated chains, that is, the concentration change of all inactive polymer chains; the first-order moment reaction rate of dead polymers represents the rate of change of the sum of the molecular weights in the terminated chains; and the second-order moment reaction rate of dead polymers represents the rate of change of the sum of the squares of the molecular weights of dead polymers.

[0161] According to the definition of the moment equation above, the reaction rate equations for potential active sites, empty active sites, and deactivated active sites on the jth active site of the catalyst can also be defined as follows:

[0162] The reaction rate equation of potential active sites:

[0163]

[0164] The reaction rate equation for the empty active site is:

[0165]

[0166] The reaction rate equation for deactivating active sites is:

[0167]

[0168] Furthermore, the mass conservation constraints of each substance in the ethylene slurry homopolymerization process are as follows:

[0169] To simplify the model, this paper uses an ideal CSTR reactor with one CSTR. Ethane is typically used as the reaction solvent in slurry processes. The solvent's role in the reaction is to dissolve the reactants while keeping the catalyst suspended in the system. Therefore, the feed contains solvent in addition to the raw materials, but the solvent does not participate in the reaction. This paper assumes that the total amount of solvent in the feed and discharge remains constant, and does not impose mass conservation constraints on the solvent.

[0170] The mass conservation constraint equation for the reaction monomer ethylene is:

[0171]

[0172] The mass conservation constraint equation for hydrogen is:

[0173]

[0174] Mass conservation constraint equation of the co-catalyst:

[0175]

[0176] The mass conservation constraint equation for potential active sites on the catalyst is:

[0177]

[0178] The mass conservation constraint equation for the empty active sites on the catalyst is:

[0179]

[0180] The mass conservation constraint equation for the deactivated active sites on the catalyst is:

[0181]

[0182] The λ-order moment mass conservation constraint equation for living polymers is:

[0183]

[0184] The λ-order moment mass conservation constraint equation for dead aggregates is:

[0185]

[0186] Among them, the subscript in represents the concentration of each component in the feed, F in and F represent the total amount of input and output, respectively, and V represents the volume of the reactor.

[0187] Furthermore, property constraints include molecular weight constraints, glass transition temperature constraints, and melt index constraints:

[0188] Since the average molecular weight of the polymer has a great influence on the properties of the polymer, the present invention takes into account the influence of the average molecular weight on the properties when predicting the properties of the polymer. The average molecular weight will be calculated based on the reaction rate equation and the mass conservation constraint equation using the moment equation method. The average molecular weight can be calculated through the reaction process of the polymerization reaction. At the same time, the calculated average molecular weight can participate in the next step of polymer property calculation. Therefore, the average molecular weight becomes a key parameter connecting the polymer reaction process and product design. According to the definition of the moment method, the number average molecular weight M of the polymer n and weight average molecular weight M w The constraints are shown in equations (37) and (38):

[0189] Number average molecular weight constraint:

[0190]

[0191] Among them, M n is the number average molecular weight, mw is the relative molecular mass of ethylene, 28.4 g / mol.

[0192] Weight average molecular weight constraint:

[0193]

[0194] Among them, Mw is the weight average molecular weight.

[0195] According to the definition of average molecular weight, the weight average molecular weight is greater than or equal to the number average molecular weight.

[0196] The molecular weight distribution of polymers obtained by polymerization is usually very dispersed. If different polymers have the same number average molecular weight, their molecular weight distributions are generally different. In order to accurately reflect the molecular weight distribution of the polymer, the polydispersity coefficient is used to reflect the molecular weight distribution width of the polymer with reference to the expectation and variance of the standard normal distribution. The polydispersity coefficient is defined as the ratio of the weight average molecular weight to the number average molecular weight, as shown in equation (39):

[0197]

[0198] From the definition of polydispersity index, we can know that the larger the PDI value, the more diversified the molecular weight of the polymer chain and the wider the molecular weight distribution; when PDI = 1, all polymer chains have the same molecular weight, and the number average molecular weight is equal to the weight average molecular weight; when PDI is minimum, the polymer system is uniform.

[0199] The mass flow constraint for polyethylene is determined by the first-order moments of living and dead polymers:

[0200]

[0201] Where: W PE is the mass flow rate of polyethylene product.

[0202] Polymer property constraints:

[0203] In order to meet users' performance requirements for polyethylene products, determine the property indicators of polyethylene products according to product needs, and optimize the best production plan, the present invention will introduce a property prediction model based on a complete reactor model to relax the degree of freedom of the property end, so that this method can determine the property constraint range according to demand and optimize the best production plan that meets the property requirements.

[0204] The present invention uses two polymer properties, the glass transition temperature affected by the number average molecular weight and the melt index affected by the weight average molecular weight, to constrain the molecular weight distribution of the polymer, while demonstrating the applicability of the present invention.

[0205] The glass transition temperature (GTT) is a key property of polymers. When the temperature rises above the GTT, amorphous polymers transform from a glassy state to a highly elastic state. After polymer production, subsequent processing and use have strict requirements on the polymer's state, so determining the GTT is crucial. The GTT constraints employed in this invention are as follows:

[0206]

[0207] Where T g is the glass transition temperature of the polymer, T g (∞) is the limit value of the glass transition temperature when the number average molecular weight approaches infinity, K g It is a constant related to the type of polymer. According to relevant data, the limiting property T of polyethylene g (∞) is 176K, constant K g It is 15000K·g / mol.

[0208] Melt flow index (MFR, also known as melt flow rate) is an important physical parameter of polymer materials, used to characterize the flow properties of thermoplastics in the molten state. MFR refers to the mass (or volume) of molten polymer passing through a standard capillary die within a certain period of time (usually 10 minutes) under specified conditions, and is expressed in g / 10min. A high MFR value indicates that the polymer has good fluidity after melting and low melt viscosity, which is suitable for blow molding and injection molding processes; a low MFR value indicates that the polymer has poor fluidity after melting and high melt viscosity, which is suitable for extrusion or calendering processes. The melt index directly reflects the processability of the polymer. MFR is related to the weight-average molecular weight of the polymer, and the melt index constraint is shown in Equation (42):

[0209] MFR=α1·(M w ) -β (42)

[0210] Among them, α1 and β1 are constants. For polyethylene, their values ​​are 4.195×10 19 The sum is -3.9252.

[0211] Furthermore, the single target in step 3 is calculated as follows:

[0212] The calculation formula for annual return Φ is as follows:

[0213]

[0214] Where a is the selling price of polyethylene, b is M 、b H2 、b c 、b A are the unit prices of ethylene monomer, hydrogen, catalyst, and co-catalyst respectively; ΔW M , ΔW H2 , ΔW C , ΔW A are the consumption of each raw material; t is the production time, 8000h / year, c V The cost of the reactor unit.

[0215] The present invention uses the Dow Chemical Fire and Explosion Index (F&EI) as the optimization target for polymerization process design. The main calculation steps are to select the process unit, determine the material coefficient, calculate the general process hazard coefficient F1 and the special process hazard coefficient F2, determine the process unit hazard coefficient F3, and finally determine the fire and explosion index. The Dow Chemical Fire and Explosion Index (F&EI) is obtained by multiplying the material coefficient MF and the process unit hazard coefficient F3, as shown in formula (44):

[0216] F&EI=MF·F3 (44)

[0217] The material factor MF is a characteristic of the inherent rate of release of potential energy from a fire or explosion caused by combustion or chemical reactions. This safety design method selects the most hazardous material in the process unit, taking into account the worst-case scenario that may occur in the actual process. The process unit hazard factor F3 is obtained by multiplying the general process hazard factor F1 by the special process hazard factor F2, as shown in formula (45):

[0218] F3=F1·F2 (45)

[0219] Among them, the general process hazard factor F1 is the main factor in determining the magnitude of accident damage, including the following six items: (1) exothermic chemical reaction; (2) endothermic reaction; (3) material handling and transportation; (4) closed or indoor process unit; (5) channel; (6) emission and leakage control. These six factors that cause harm are applicable to most process situations. Generally, no penalty is imposed on all factors, but these issues play an important role in fire and explosion incidents. Their calculation is based on equation (46):

[0220]

[0221] Among them, this optimization method only considers the penalty item of the general process hazard factor F1: exothermic chemical reaction.

[0222] Like the general process hazard factor F1, the special process hazard factor F2 is the main factor affecting the probability of accidents. Special process conditions are the main cause of fire and explosion accidents. Special process hazards include the following twelve items: (1) toxic substances; (2) negative pressure operation (<500mmHg); (3) operation in or near the combustion range; (4) dust explosion; (5) pressure release; (6) low temperature; (7) the amount of flammable and unstable chemicals; (8) corrosion and abrasion; (9) leakage - joints and packing; (10) use of open flame equipment; (11) hot oil heat exchange system; (12) rotating equipment, as shown in equation (47):

[0223]

[0224] Among them, since some penalty items do not work in the process units of this method, this method does not consider negative pressure operation, dust explosion, pressure release, low temperature, corrosion and wear, leakage and joints, the use of open flame equipment, hot oil heat exchange system and rotating equipment.

[0225] When calculating the penalty for the quantity of flammable and unstable chemicals for the special process hazard item F2, the inventory of flammable materials is the independent variable and can be estimated by the material volume V of the reactor.

[0226] The penalty is calculated according to equation (48):

[0227]

[0228] where Y is the penalty value, TB is the process unit energy content, which is a function of the reactor volume V, density ρ, and the heat of combustion Hc of the substance, as shown in equation (49):

[0229] TB=f(V,ρ,H c ) (49)

[0230] Furthermore, step 4 specifically involves establishing an NLP model in GAMS, using the CONOPT solver to solve the problem, and obtaining the results of each single-objective and multi-objective optimization.

[0231] This intrinsically safe design approach will provide the following information:

[0232] The values ​​of reactor volume, polyethylene product output, annual revenue Φ and F&EI.

[0233] Case Study:

[0234] This case studies the production of polyethylene products with a glass transition temperature between 175-175.50K and a melt index between 0.139-0.157g / min. Since polymerization reactions require strict temperature control, the reaction temperature is fixed at 358.15K. The operating conditions for this case are fixed feed conditions, total feed flow rate, and feed composition. Single-objective optimization is performed on the annual return Φ and the negative value of the Dow Chemical Fire and Explosion Index (F&EI). Multi-objective optimization is then performed on the annual return Φ and the negative value of the Dow Chemical Fire and Explosion Index (F&EI). By applying a weight coefficient to each objective along a normalized strategy, multiple objectives are combined into a single function. The optimization results are shown in the following table:

[0235] Table 1

[0236]

[0237] From the above table, we can see that when annual income is used as the objective function, the maximum annual income is 4505.46*10 4$ / year, the F&EI value is 112.31, which is a moderate risk. When the negative value of F&EI is used as the objective function, the minimum value of F&EI is 72.044, and the annual return is negative 72.044*10 4 $ / year; when multi-objective optimization is performed, the annual income is 3918.59*10 4 $ / year, which is 13% lower than the maximum profit, and the F&EI is 96.4, which is 14.2% lower than the maximum value. It can be seen that the intrinsically safe design method of the present invention reduces the danger of the polyethylene production process to a considerable extent while ensuring considerable profits.

Claims

1. A method for designing intrinsic safety of a polyethylene homopolymerization process in combination with product property requirements, characterized in that: The specific steps include: Step 1: Determine the reaction rate of each substance in the reaction process according to the reaction kinetics equation of the ethylene slurry homopolymerization process: Based on the ethylene slurry polymerization reaction mechanism with five active sites, the reaction kinetic equation of the ethylene slurry homopolymerization process includes catalyst activation, chain initiation, chain growth, chain transfer and chain deactivation reactions. The specific equation is as follows: C P (j)+A→P0(j) (1) P0(j)+M→P1(j) (2) P n (j)+M→P n+1 (j) (3) P n (j)+M→P1(j)+D n (j) (4) P n (j)+H2→P0(j)+D n (j) (5) P n (j)+A→P0(j)+D n (j) (6) P n (j)→P0(j)+D n (j) (7) P n (j)→C d (j)+D n (j) (8) P0(j)→C d (j) (9) Among them, C P (j), P0(j), C d (j) are the potential active sites, empty active sites, and deactivated active sites on the jth active site on the catalyst, A is the co-catalyst, M is the homomonomer, and P is the n (j) is the active polymer with a chain length of n on the jth active site on the catalyst, D n (j) is the dead polymer with chain length n on the jth active site on the catalyst, k aA (j), k i (j), k p (j), k tM (j), k tH (j), k tA (j), k t (j), k d (j) Reaction kinetic constants corresponding to catalyst activation, chain initiation, chain extension, transfer to ethylene monomer, transfer to hydrogen, transfer to cocatalyst, transfer to β-hydride, and chain deactivation reactions, respectively; The reaction kinetic constant is temperature-dependent and is calculated using the Arrhenius law: Where k0 is the pre-exponential factor, E a is the reaction activation energy, R is the ideal gas constant, T is the reactor temperature, T ref is the reaction reference temperature; Step 2: Optimizing the ethylene homopolymerization process based on mass constraints and property constraints; Step 3: Perform intrinsically safe design and optimize the ethylene homopolymerization process. First, perform single-objective optimization on the annual return Φ and the Dow Chemical Fire and Explosion Index (F&EI), respectively, with the goals of maximizing the annual return Φ and maximizing the negative value of the F&EI, to obtain the maximum annual return Φ and the minimum Dow Chemical Fire and Explosion Index (F&EI). Then, perform multi-objective optimization on the annual return Φ and the negative value of the Dow Chemical Fire and Explosion Index (F&EI). By applying a weight coefficient to each objective along a normalized strategy, multiple objectives are combined into a function, and the maximum value of this function is optimized as the objective. Each objective, quality constraint, and property constraint constitute a nonlinear programming NLP model. By solving the NLP model, we can ultimately obtain the optimal production conditions that not only meet the property requirements of the polymer product, but also meet the safety and economic requirements of the polymer production process: The final goal is as follows: Among them Ф up and F&EI up are Φ and F&EI obtained when the goal is to maximize annual income Φ, Φ low and F&EI low are Φ and F&EI obtained when the negative value of F&EI is maximized, and α is the weight coefficient of multi-objective optimization; Step 4: According to the set upper and lower limit constraints of the variables, the GAMS solver is used to solve the established NLP model.

2. The intrinsic safety design method for a polyethylene homopolymerization process according to claim 1, characterized in that: The reaction processes involved in the ethylene slurry homopolymerization process all consider the following assumptions: (1) The CSTR is operated with a full reactor, with only liquid phase in the reactor, and the effect of gas-liquid phase equilibrium on the reaction process is ignored; (2) The reactor is an ideal reactor, and the components are evenly distributed in the reactor during the reaction; (3) Ignoring the diffusion effect in the reactor, the temperature in the reactor is uniform; (4) Under the equal-length assumption of polymerization reaction, the reaction kinetics of chain growth, chain transfer and chain deactivation are independent of chain length.

3. The intrinsic safety design method for a polyethylene homopolymerization process according to claim 1, characterized in that: The reaction rates of each substance in the reaction process are as follows: Define the combined reaction rate constant of chain transfer and chain deactivation: K TD (j) K TD (j)=k tM (j)[M]+k tH (j)([H2]) 0.5 +k tA (j)[A]+k t (j)+k d (j) (12) Where [] represents the concentration of each substance in the reactor; The reaction rate of living polymer and dead polymer with chain length 1: where Y 0 (j) is the zero-order moment of the living polymer on the jth active site on the catalyst; Reaction rate of living polymer with dead polymer whose chain length is greater than 2: The reaction rates of the zero-order, first-order, and second-order moments of the living polymer and the dead polymer at the j-th active site on the catalyst are: The reaction rate equation of potential active sites, empty active sites, and deactivated active sites on the jth active site of the catalyst is:

4. The intrinsic safety design method for a polyethylene homopolymerization process according to claim 1, characterized in that: The mass conservation constraints of each substance in the ethylene slurry homopolymerization process are as follows: The total amount of solvent in the input and output materials remains unchanged, and no mass conservation constraints are imposed on the solvent; Mass conservation constraints for ethylene monomer, hydrogen, and co-catalyst: Where V is the volume of the reactor, F in and F are the total amount of input and output materials respectively; Mass conservation constraints for potential active sites, empty active sites, and deactivated active sites on the jth active site on the catalyst: The λ-order moment mass conservation constraint for living and dead aggregates:

5. The intrinsic safety design method for a polyethylene homopolymerization process according to claim 1, characterized in that: Property constraints include molecular weight constraints, glass transition temperature constraints, and melt index constraints: Number average molecular weight constraint: Among them, M n is the number average molecular weight, mw is the relative molecular mass of ethylene, 28.4 g / mol. Weight average molecular weight constraint: Among them, M w is the weight average molecular weight; Polyethylene mass flow constraints: Where: W PE is the mass flow rate of polyethylene product: Glass transition temperature constraints: Melt index constraints: MFR=α1·(M w ) -β (38) Among them, T g (∞) is the ultimate glass transition temperature, K g is a constant related to the type of polymer. For polyethylene, its value is 15000K·g / mol. α1 and β1 are constants. For polyethylene, their values ​​are 4.195×10 19 The sum is -3.9252.

6. The intrinsic safety design method for a polyethylene homopolymerization process according to claim 1, characterized in that: The single target calculation in step 3 is as follows: The calculation formula for annual return Φ is as follows: Where a is the selling price of polyethylene, b is M 、b H2 、b c 、b A are the unit prices of ethylene monomer, hydrogen, catalyst, and co-catalyst respectively; ΔW M , ΔW H2 , ΔW C , ΔW A are the consumption of each raw material; t is the production time, 8000h / year, c V The cost of the reactor unit; Dow Chemical Fire and Explosion Index (F&EI) calculation equation: F&EI=MF·F3 (40) Calculation equation for process unit risk factor F3: F3=F1·F2 (41) Among them, F1 is generally the process hazard factor, and F2 is the special process hazard factor.

7. The method for designing intrinsic safety of a polyethylene homopolymerization process according to claim 1, characterized in that: Step 4 specifically involves building an NLP model in GAMS, using the CONOPT solver to solve the problem, and obtaining the results of each single-objective and multi-objective optimization. This intrinsically safe design approach will provide the following information: The values ​​of reactor volume, polyethylene product output, annual revenue Φ and F&EI.

Images