Method, system and application of multi-stage separation unit initialization based on virtual transient model

Abstract

The application discloses a multi-stage separation unit initialization method and system based on a virtual transient model and application, relates to the technical field of chemical process simulation, and comprises the following steps: constructing virtual transient differential equations of a multi-stage separation unit; constructing virtual differential variable correlation equations of the multi-stage separation unit; constructing a virtual transient model of the whole multi-stage separation unit; solving the virtual transient model; and outputting initial estimated values. The application adopts the multi-stage separation unit initialization method and system based on the virtual transient model and application, avoids using complex hydraulic correlation equations and structural information of the separation unit when constructing a dynamic process differential algebraic equation model, the time-varying term is virtual, and the virtual transient model does not need to be based on a real physical process, but is completely equivalent to an algebraic equation model when being integrated to a steady state, so that the multi-stage separation unit initialization process based on the virtual transient model is easier to solve, and has the advantages of high calculation efficiency and high stability.

Description

Technical Field

[0001] This invention relates to the field of chemical process simulation technology, and in particular to a method, system and application for initializing multi-level separation units based on a virtual transient model. Background Technology

[0002] In the field of chemical process simulation technology, steady-state simulations of multi-stage separation units (such as distillation, extraction, and absorption) are typically based on algebraic models constructed using the MESH equations (mass conservation, phase equilibrium, compositional summation, and energy conservation). Due to the highly nonlinear, large-scale, and strongly coupled nature of these models, they are extremely sensitive to initial values ​​when used to describe the separation of non-ideal systems or multi-stage separation units with complex structures (such as distillation columns with multiple feeds, multiple side-stream extractions, and mid-stage recirculation), making convergence difficult. Different initialization algorithms are generally required for different systems (such as narrow-boiling-point, wide-boiling-point, and non-ideal systems). Currently, commercial process simulation software employs the following initialization techniques during the solution process:

[0003] For narrow boiling point systems, the Boston initialization algorithm is used by default;

[0004] For systems with wide boiling points, the Boston initialization algorithm or the Boston initialization algorithm combined with the one-step bubble point method can be used.

[0005] For systems with extremely wide boiling points (such as petroleum fractions), initial values ​​can be obtained using the Edmister method or a simplified calculation based on the constant molar flow assumption.

[0006] For strongly nonideal systems, the dedicated initialization algorithms developed by various commercial software can only guarantee simulation convergence under finite conditions.

[0007] However, the above algorithm has the following drawbacks:

[0008] Poor versatility: Existing initialization algorithms require different methods to be selected based on the system and tower structure, which cannot adapt to various complex situations (such as multiple feeds, multiple side lines, mid-section circulation, etc.), resulting in a limited scope of application of the algorithms.

[0009] Unstable initial value quality: The quality of the initial estimates provided by existing algorithms cannot be guaranteed, which can easily lead to divergence or convergence to non-physical solutions in steady-state simulation iterations. This is especially true for complex structures or wide-boiling-point systems (such as water-containing petroleum hydrocarbon systems), where the convergence failure rate is extremely high.

[0010] Narrow convergence region: Existing initialization algorithms are based on simplification assumptions or estimate initial values ​​for specific systems, which cannot guarantee that the obtained initial values ​​are reasonable and close to the steady-state solution. For simulation and solution algorithms such as Newton-type algorithms that have strict requirements on the convergence region, they cannot achieve convergence. Summary of the Invention

[0011] The purpose of this invention is to provide a method, system, and application for initializing multi-level separation units based on a virtual transient model, thereby solving the problems mentioned in the background art.

[0012] To achieve the above objectives, this invention provides a method for initializing multi-level separation units based on a virtual transient model, comprising the following steps:

[0013] Step S1: Construct the virtual transient differential equations of the multi-level separated units;

[0014] Step S2: Construct the virtual differential variable correlation equations for the multi-level separated units;

[0015] Step S3: Based on the virtual transient differential equation and the virtual differential variable correlation equation, construct a virtual transient model of the entire tower of multi-stage separation units;

[0016] Step S4: Solve the virtual transient model;

[0017] Step S5: Output the initial estimated value.

[0018] Preferably, step S1 includes: setting virtual dynamic variables for each tray in the multi-stage separation unit, the virtual dynamic variables being the virtual feed amount and virtual heat holding of each component on the tray, and constructing the mass conservation equation and heat conservation equation of the equilibrium stage of the multi-stage separation unit into differential equation form as the dynamic mass and heat balance relationship.

[0019] Preferably, for the first stage within a multi-stage separation unit Grade, Component The mass conservation differential equation is:

[0020] ;

[0021] For the first in a multi-level separation unit The differential equation for the conservation of heat at stage A is:

[0022] ;

[0023] in, Indicates the first Primary components Virtual material holding capacity Indicates the first Virtual heat storage level Indicates the first -1 level liquid phase flow rate, Indicates the first +1 stage vapor phase flow rate, Indicates the first Phase liquid flow rate, Indicates the first Phase flow rate of steam, Indicates the first Primary components liquid phase mole fraction, Indicates the first Primary components Vapor phase mole fraction, Indicates the first -1 grade components liquid phase mole fraction, Indicates the first +1 grade components Vapor phase mole fraction, Indicates the first Components in the multistage mixed feed composition mole fraction, Indicates the first Enthalpy of vapor phase, Indicates the first Enthalpy of liquid phase Indicates the first +1 level vapor phase enthalpy value, Indicates the first -1 level liquid phase enthalpy value, Indicates the first The enthalpy of the multistage mixed feed , They represent the first The flow rate of the vapor-liquid phase side sampling is as follows: Indicates the first The heat exchange between the stage and the external environment, Indicates the first The molar flow rate of the primary mixed feed.

[0024] Preferably, step S2 includes establishing simplified correlation equations between the virtual dynamic variable and other variables within the tower, as shown below:

[0025] ;

[0026] ;

[0027] in, It is a constant, usually taken as 1800h. -1 .

[0028] Preferably, step S3 includes: combining the mass conservation differential equation, the heat conservation differential equation, the simplified correlation equation and phase equilibrium equation of the virtual dynamic variable and other variables in the tower, and the composition summation equation to form a differential-algebraic equation system, thereby constituting a virtual transient model of the multi-stage separation unit.

[0029] Preferably, the phase equilibrium equation is as follows:

[0030] ;

[0031] The composition summation equation is as follows:

[0032] ;

[0033] in, Indicates the first Grade, Component The phase equilibrium constant, This indicates the total score for each group.

[0034] Preferably, step S4 includes: integrating the virtual transient model using a numerical integration algorithm until it approaches a steady state.

[0035] Preferably, step S5 includes: extracting the temperature, gas flow rate, liquid flow rate, and component mole fraction of each theoretical stage when the virtual transient model approaches steady state, as the initial estimates for iterative solution by the rigorous steady-state simulation algorithm.

[0036] A system based on a multi-level separation unit initialization method of a virtual transient model includes a computer-readable storage medium located on a computer device, which stores a model building module, a solution module, and an output module.

[0037] This invention also proposes the application of a multi-stage separation unit initialization method based on a virtual transient model, which is applied to conventional distillation columns, extraction columns, absorption columns, stripping columns, extractive distillation columns, azeotropic distillation columns, and complex distillation columns with multiple feeds, multiple side streams, mid-section circulation, and thermal coupling structures.

[0038] Therefore, the present invention employs the above-mentioned multi-level separation unit initialization method, system, and application based on a virtual transient model, which has the following beneficial effects:

[0039] (1) Virtual transient modeling of multi-stage separation unit: When constructing the differential algebraic equation model of dynamic process, the complex hydraulic correlation and structural information of separation unit are avoided. The time-varying terms are virtual and do not need to be based on the real physical process. However, when integrated to the steady state, it is completely equivalent to the algebraic equation model. Therefore, the initialization process of multi-stage separation unit based on virtual transient model is easier to solve and has the advantages of high computational efficiency and greater robustness.

[0040] (2) Robust initialization method: This method is very robust. On the one hand, the DEAs model based on virtual transients decouples the large-scale nonlinear equation system into a series of small-scale nonlinear or even linear equation systems, which are easy to solve. On the other hand, during the integration process, the results from the previous round of integration can provide good initial values ​​for the next round of integration. Therefore, the method is robust.

[0041] (3) Significantly improves the convergence of multi-level separated unit simulation: The initial values ​​obtained by virtual transient modeling are closer to the real steady-state state. Simulation experiments show that the initial values ​​provided by this method can effectively reduce the risk of iterative divergence in steady-state simulation and improve the convergence success rate.

[0042] (4) High versatility: This method is applicable to various systems (ideal, non-ideal, petroleum fractions, etc.) and tower structures (multiple feed, multiple side streams, mid-section circulation, etc.), without the need to adjust the algorithm for different situations. Compared with the existing technology, this method eliminates the dependence on a specific initialization algorithm.

[0043] (5) Improved computational efficiency: Based on the good initial values ​​obtained from the virtual transient model, the number of iterations in the subsequent steady-state simulation is greatly reduced, resulting in improved overall computational efficiency.

[0044] The technical solution of the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. Attached Figure Description

[0045] Figure 1 This is a flowchart illustrating the multi-level separation unit initialization method, system, and application embodiments based on the virtual transient model of the present invention.

[0046] Figure 2 This is a schematic diagram of the structure of the primary distillation column according to an embodiment of the present invention. Detailed Implementation

[0047] The technical solution of the present invention will be further described below with reference to the accompanying drawings and embodiments.

[0048] Unless otherwise defined, the technical or scientific terms used in this invention shall have the ordinary meaning understood by one of ordinary skill in the art to which this invention pertains. The terms "first," "second," and similar terms used in this invention do not indicate any order, quantity, or importance, but are merely used to distinguish different components. Terms such as "comprising" or "including" mean that the element or object preceding the word encompasses the elements or objects listed following the word and their equivalents, without excluding other elements or objects. Terms such as "connected" or "linked" are not limited to physical or mechanical connections, but can include electrical connections, whether direct or indirect. Terms such as "upper," "lower," "left," and "right" are used only to indicate relative positional relationships; when the absolute position of the described object changes, the relative positional relationship may also change accordingly.

[0049] Example

[0050] Please see Figures 1-2This invention provides an initialization method for multi-stage separation units based on a virtual transient model, applicable to the simulation and solution of multi-stage separation units for component separation in ideal systems, non-ideal systems, and complex petroleum fractions. Specifically, it includes the following steps:

[0051] Step S1: Construct virtual transient differential equations for multi-stage separation units. These units include conventional distillation, extraction, absorption, stripping, extractive distillation, azeotropic distillation, and complex distillation columns with multiple feeds, multiple side streams, mid-section circulation, and thermal coupling structures. Specifically, this involves setting virtual dynamic variables for each tray in the multi-stage separation unit. These virtual dynamic variables are the virtual feed holding capacity and virtual heat holding capacity of each component on the tray. The mass conservation equations and heat conservation equations for the equilibrium stage of the multi-stage separation unit are constructed into differential equations, representing the dynamic mass and heat balance relationship. This method introduces virtual feed holding capacity and heat holding capacity to construct a virtual transient differential algebraic model for the entire column, adapting to complex structures (multiple feeds, multiple side streams, etc.). It achieves efficient dynamic approximation of the steady-state solution from coarse initial conditions without the need for separate column sections, thereby obtaining high-quality initial values.

[0052] For the first in a multi-level separation unit Grade, Component The mass conservation differential equation is: ;

[0053] For the first in a multi-level separation unit The differential equation for the conservation of heat at stage A is:

[0054] ;

[0055] in, Indicates the first Primary components Virtual material holding capacity Indicates the first Virtual heat storage level Indicates the first -1 level liquid phase flow rate, Indicates the first +1 stage vapor phase flow rate, Indicates the first Phase liquid flow rate, Indicates the first Phase flow rate of steam, Indicates the first Primary components liquid phase mole fraction, Indicates the first Primary components Vapor phase mole fraction, Indicates the first -1 grade components liquid phase mole fraction, Indicates the first +1 grade components Vapor phase mole fraction, Indicates the first Components in the multistage mixed feed composition mole fraction, Indicates the first Enthalpy of vapor phase, Indicates the first Enthalpy of liquid phase Indicates the first +1 level vapor phase enthalpy value, Indicates the first -1 level liquid phase enthalpy value, Indicates the first The enthalpy of the multistage mixed feed , They represent the first The flow rate of the vapor-liquid phase side sampling is as follows: Indicates the first Heat exchange between stage and the external environment (positive value indicates stage 1) The primary environment transfers heat to the multi-stage separation unit; a negative value indicates that the primary environment transfers heat to the multi-stage separation unit. (Multi-stage separation unit transfers heat to the environment) Indicates the first The molar flow rate of the primary mixed feed.

[0056] Step S2: Constructing the virtual differential variable correlation equations for the multi-stage separation unit: Establish simplified correlation equations between the virtual dynamic variables and other variables within the tower (such as flow rate, composition, and enthalpy), as shown below:

[0057] ；

[0058] ;

[0059] in, It is a constant, usually taken as 1800h. -1 .

[0060] Step S3: Based on the virtual transient differential equation and the virtual differential variable correlation equation, construct a virtual transient model of the entire tower of the multi-stage separation unit: combine the mass conservation differential equation, the heat conservation differential equation, the simplified correlation equation of the virtual dynamic variable with other variables in the tower, the phase equilibrium equation, and the composition summation equation to form a system of differential-algebraic equations (DAEs), and thus construct a virtual transient model of the multi-stage separation unit.

[0061] The phase equilibrium equations are as follows:

[0062] ;

[0063] The above components are summed into the following equation:

[0064] ;

[0065] in, Indicates the first Grade, Component The phase equilibrium constant, This indicates the total score for each group.

[0066] Step S4: Solve the virtual transient model: Use a numerical integration algorithm (such as explicit Euler method, implicit Euler method or hybrid integration method) to integrate the virtual transient model until it approaches the steady state, that is, the virtual time is large enough and the variable changes tend to be gradual.

[0067] Step S5: Output initial estimates: Extract the temperature, gas flow rate, liquid flow rate and component mole fraction of each theoretical stage when the virtual transient model is close to steady state, and use them as initial estimates for iterative solution by rigorous steady-state simulation algorithms (such as flow rate summation method, Newton-type algorithm, inner and outer layer method, etc.).

[0068] A system based on a multi-level separation unit initialization method of a virtual transient model includes a computer-readable storage medium located on a computer device, which stores a model building module, a solution module, and an output module.

[0069] This invention also proposes the application of a multi-stage separation unit initialization method based on a virtual transient model, which is applied to conventional distillation columns, extraction columns, absorption columns, stripping columns, extractive distillation columns, azeotropic distillation columns, and complex distillation columns with multiple feeds, multiple side streams, mid-section circulation, and thermal coupling structures.

[0070] like Figure 1 The flowchart shown is a flowchart of this method, in which... Indicates the round.

[0071] Example 1

[0072] The initial values ​​output by this method improve the convergence success rate of traditional steady-state solution methods (such as Newton's method) by about 50%. In the atmospheric pressure tower example, the initial values ​​after more than 100 cycles of mixed integrals enable Newton's method to converge quickly, while the initial values ​​of Boston's method cause divergence. This proves that this method can improve the initialization quality.

[0073] Example 2

[0074] like Figure 2 As shown, the application of this embodiment is the primary distillation column in crude oil separation.

[0075] Table 1. Feed Specifications for the Primary Distillation Column

[0076]

[0077] In addition to the table above, the entire tower consists of 10 stages. The pressure of the condenser is 275 kPa, the pressure at the top of the tower is 289 kPa, the pressure drop of the entire tower is 21 kPa, the physical property method is BK10, and the specifications for the heating furnace include: furnace temperature of 478.15 K and furnace pressure of 345 kPa.

[0078] For the above applications, the iteration count results using the classic initialization algorithm are shown in Table 2 below.

[0079] Table 2. Number of iterations and inner / outer layer errors for calculations using the inner and outer methods with initialization values ​​provided by the classical initialization algorithm.

[0080]

[0081] The iteration count of this algorithm is shown in Table 3 below. The numerical integration method uses the explicit Euler integral method, with the step size set to 0.005 and the maximum number of iterations set to 10.

[0082] Table 3. Number of iterations and inner / outer layer errors for calculations using the inner and outer methods based on the initial values ​​provided by the explicit Euler algorithm integration using the virtual transient model.

[0083]

[0084] As can be seen from the comparison of Tables 2 and 3, this algorithm significantly improves computational efficiency. Based on the good initial values ​​obtained from the virtual transient model, the number of iterations in the subsequent steady-state simulation is reduced by 70%.

[0085] Therefore, the present invention adopts the above-mentioned multi-level separation unit initialization method, system and application based on virtual transient model. When constructing the dynamic process differential algebraic equation model, it avoids the use of complex hydraulic correlations and structural information of separation units. The time-varying terms are virtual and do not need to be based on real physical processes, but are completely equivalent to algebraic equation models when integrated to steady state. Therefore, the multi-level separation unit initialization process based on virtual transient model is easier to solve and has the advantages of high computational efficiency and greater robustness.

[0086] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit them. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can still be made to the technical solutions of the present invention, and these modifications or equivalent substitutions cannot cause the modified technical solutions to deviate from the spirit and scope of the technical solutions of the present invention.

Claims

1. A method for initializing multi-level separation units based on a virtual transient model, characterized in that, Includes the following steps: Step S1: Construct the virtual transient differential equations of the multi-level separated units; Step S2: Construct the virtual differential variable correlation equations for the multi-level separated units; Step S3: Based on the virtual transient differential equation and the virtual differential variable correlation equation, construct a virtual transient model of the entire tower of multi-stage separation units; Step S4: Solve the virtual transient model; Step S5: Output the initial estimated value; Step S1 includes: setting virtual dynamic variables for each tray in the multi-stage separation unit, the virtual dynamic variables being the virtual feed amount and virtual heat holding of each component on the tray, and constructing the mass conservation equation and heat conservation equation of the equilibrium stage of the multi-stage separation unit into differential equation form as a dynamic mass and heat balance relationship. In step S1, for the first stage within the multi-stage separation unit Grade, Component The mass conservation differential equation is: ； For the first in a multi-level separation unit The differential equation for the conservation of heat at stage 1 is: ； in, Indicates the first Primary components Virtual material holding capacity Indicates the first Virtual heat storage level Indicates the first -1 level liquid phase flow rate, Indicates the first +1 stage vapor phase flow rate, Indicates the first Phase liquid flow rate, Indicates the first Phase flow rate of steam, Indicates the first Primary components liquid phase mole fraction, Indicates the first Primary components Vapor phase mole fraction, Indicates the first -1 grade components liquid phase mole fraction, Indicates the first +1 grade components Vapor phase mole fraction, Indicates the first Components in the multistage mixed feed composition mole fraction, Indicates the first Enthalpy of vapor phase, Indicates the first Enthalpy of liquid phase Indicates the first +1 level vapor phase enthalpy value, Indicates the first -1 level liquid phase enthalpy value, Indicates the first The enthalpy of the multistage mixed feed , They represent the first The flow rate of the vapor-liquid phase side sampling is as follows: Indicates the first The heat exchange between the stage and the external environment, Indicates the first The molar flow rate of the primary mixed feed.

2. The multi-level separation unit initialization method based on a virtual transient model according to claim 1, characterized in that, Step S2 includes establishing simplified correlation equations between the virtual dynamic variable and other variables within the tower, as shown below: ； ； in, It is a constant, usually taken as 1800h. -1 .

3. The multi-level separation unit initialization method based on a virtual transient model according to claim 2, characterized in that, Step S3 includes: combining the mass conservation differential equation, the heat conservation differential equation, the simplified correlation equation and phase equilibrium equation of the virtual dynamic variable with other variables in the tower, and the composition summation equation to form a differential-algebraic equation system, thereby constituting a virtual transient model of the multi-stage separation unit.

4. The multi-level separation unit initialization method based on a virtual transient model according to claim 3, characterized in that, The phase equilibrium equations are as follows: ； The composition summation equation is as follows: ； in, Indicates the first Grade, Component The phase equilibrium constant, This indicates the total score for each group.

5. The multi-level separation unit initialization method based on a virtual transient model according to claim 4, characterized in that, Step S4 includes: integrating the virtual transient model using a numerical integration algorithm until it approaches a steady state.

6. The multi-level separation unit initialization method based on a virtual transient model according to claim 5, characterized in that, Step S5 includes: extracting the temperature, gas flow rate, liquid flow rate, and component mole fraction of each theoretical stage when the virtual transient model is close to steady state, as the initial estimates for the iterative solution of the rigorous steady-state simulation algorithm.

7. A system applied to the multi-level separation unit initialization method based on the virtual transient model as described in claims 1-6, characterized in that: This includes a computer-readable storage medium located on a computer device, which stores a model building module, a solution module, and an output module.

8. The application of a multi-level separation unit initialization method based on a virtual transient model, characterized in that: The multi-stage separation unit initialization method based on the virtual transient model described in any one of claims 1-6 is applied to conventional distillation columns, extraction columns, absorption columns, stripping columns, extractive distillation columns, azeotropic distillation columns, and complex distillation columns with multiple feeds, multiple side streams, mid-section circulation, and thermal coupling structures.

Images