A method and system for predicting NOx at furnace outlet based on numerical simulation

Abstract

The present invention relates to a furnace outlet NOx prediction system based on numerical simulation, including a prediction model establishment part and a real-time prediction part; and also relates to a prediction method, including: processing DCS data, obtaining boundary conditions required for numerical simulation calculation, establishing a three-dimensional geometric model of the furnace, calculating the furnace outlet NOx concentration data under different working conditions, and determining a database; according to the database, using the support vector method to establish a furnace outlet NOx prediction model; finally, according to the real-time furnace inlet parameters under actual working conditions, the furnace outlet NOx concentration is predicted in real time. The method of the present invention can predict the change of NOx at the furnace outlet and the NOx distribution in the furnace when the combustion working conditions change, that is, when the operating parameters such as wind / powder are adjusted, and provide advance guidance for SCR ammonia injection, which is of great significance to improving the denitration efficiency of coal-fired power plants, the safety of air preheater operation, and improving the economy.

Description

Technical Field

[0001] The invention relates to the technical field of NOx prediction at a furnace outlet of a coal-fired boiler, and in particular to a method and system for predicting NOx at a furnace outlet based on numerical simulation. Background Art

[0002] At present, coal-fired power plants mainly use selective catalytic reduction (SCR) flue gas denitrification technology to control NOx emissions, using reducing agents (NH3 or urea, etc.) to selectively convert NOx in flue gas into N2 and water. However, SCR flue gas denitrification ammonia injection control is difficult to adapt to the AGC load instructions and deep peak regulation requirements that change greatly and frequently, often resulting in excessive ammonia injection, excessive ammonia escape rate, and blockage and corrosion of air preheater components. The main reasons include: the lag and low accuracy of flue gas NOx concentration measurement are important factors leading to poor ammonia injection control effects. Abnormalities and failures of the continuous flue gas emission monitoring system (CEMS) occur from time to time, mainly sampling abnormalities and data measurement deviations, such as primary component failure, flue gas sampling probe blockage and pipeline leakage, standard gas failure, back-purge air water content, heating abnormalities of the heating pipeline, and pretreatment equipment failure, which lead to flue gas parameter inaccuracy (flow, temperature and pressure), flue gas composition analysis system abnormalities (data inaccuracy and analyzer failure), etc. Therefore, if the changes in NOx at the furnace outlet and the NOx distribution in the furnace can be predicted when the combustion conditions change, that is, when the operating parameters such as air / powder are adjusted, this can serve as a supplement to the basis for SCR ammonia injection and guide SCR ammonia injection in advance, which is of great significance to improving the denitrification efficiency of coal-fired power plants, the safety of air preheater operation, and improving economic efficiency.

[0003] In the prior art, the patent with application number 202010916787.1 and titled "A system and method for predicting NOx concentration at the inlet of SCR based on an exponential prediction model" uses the historical data of power plants and the idea of ​​data modeling to design a prediction system including a weight transmitter, a speed transmitter, a primary air volume transmitter, a secondary air volume transmitter, an oxygen transmitter, a primary air pressure transmitter, a controller and a distributed control system, and a method for predicting NOx concentration based on relevant parameters. This method still has disadvantages such as nonlinearity, delay, distributed parameters and many disturbance factors. Summary of the invention

[0004] In view of the shortcomings of the prior art, the present invention provides a furnace outlet NOx prediction method and system based on numerical simulation, aiming to solve the problem of delayed and inaccurate NOx concentration measurement, predict NOx emissions from the combustion process, and thus improve the ammonia injection control effect. The NOx concentration distribution in the furnace is calculated by numerical simulation, and a furnace combustion monitoring system is established based on the numerical simulation results.

[0005] The technical solution adopted by the present invention is as follows:

[0006] A method for predicting furnace outlet NOx based on numerical simulation includes the following steps:

[0007] 1) Extracting furnace combustion boundary conditions required for numerical simulation calculation from the DCS system, wherein the boundary conditions include furnace combustion inlet parameter data;

[0008] 2) Numerical simulation calculation is performed according to the furnace combustion boundary conditions to obtain the furnace outlet NOx concentration under different working conditions, specifically including:

[0009] According to the actual size of the furnace, establish the physical model of the furnace, including the water-cooled wall, burner, overburnt air, and superheater of the furnace;

[0010] Construct furnace combustion model, including: Realizable k-ε turbulence model, particle motion model, gas phase combustion model, volatilization analysis model, coke combustion model, radiation heat transfer model and NOx generation and reduction model;

[0011] 3) Establish a database of furnace combustion inlet parameter data and corresponding furnace outlet NOx concentration data under different operating conditions;

[0012] 4) Taking the furnace combustion inlet parameter data as the independent variable and the furnace outlet NOx concentration data as the dependent variable, the support vector machine method was used to establish the furnace outlet NOx prediction model;

[0013] 5) Obtain the real-time furnace inlet parameters under actual working conditions;

[0014] 6) Using the NOx prediction model, the NOx concentration at the furnace outlet is predicted in real time according to the real-time furnace inlet parameters.

[0015] Further technical solutions are:

[0016] In step 2), the gas phase combustion model adopts a non-premixed combustion model, and constructing and solving the non-premixed combustion model includes the following steps:

[0017] Based on the atomic mass fraction, the mixing fraction is defined as:

[0018]

[0019] In the formula, Z i is the element mass fraction of element i; the subscript ox represents the value at the oxidant inlet, and the subscript fule represents the value at the fuel inlet;

[0020] The time-averaged mixture fraction equation is established based on the mixture fraction:

[0021]

[0022] In the formula, S m is the mass of coal powder particles transferred into the gas phase; S user For custom source items;

[0023] Solve the conservation equation for the mean square value of the average mixture fraction using the following equation:

[0024]

[0025] In the formula, σ i =0.85, C g =286, C d =2.0.

[0026] In step 2), the coke combustion model uses a diffusion-dynamic controlled reaction rate model, taking into account the effects of diffusion and reaction kinetics on the particle surface reaction rate, and combining the rate at which oxygen diffuses to the coke surface and the chemical reaction kinetics rate on the coke surface, and the weighted calculation to obtain the coke surface reaction combustion rate is:

[0027]

[0028] In the above formula, m p , A p are the mass and area of ​​the droplet respectively; ρ is the gas density; R is the gas molar constant; T ∞ is the ambient medium temperature; Y ox is the mass fraction of the oxidant in the local gas; D0 is the diffusion rate coefficient; is the kinetic rate; M W，OX is the molar mass of the oxidant.

[0029] In step 2), the particle motion model adopts a random particle trajectory model and is established according to the following formula:

[0030]

[0031] and pulse speed; m p , A p are the particle mass and surface area, respectively; ρ g is the local gas density; g k is the acceleration due to gravity; C D is the drag coefficient;

[0032] The volatile analysis model adopts a single-step pyrolysis model and is established according to the following formula:

[0033]

[0034] In the formula, mp 表 indicates the particle mass, f v，0is the initial volatile mass fraction of the particles, f w，0 is the mass fraction of the evaporable and boiling substance in the particles; m p，0 is the initial mass of the particle; k is the reaction rate constant;

[0035] In step 2), the radiation heat transfer model adopts the P-1 model, and the local incident radiation solution equation is:

[0036]

[0037] Where n is the refractive index of the medium, σ is the Stefan-Boltzmann constant, and S G is the user-defined radiation source, a is the absorption coefficient, G is the incident radiation, T is the temperature, and Γ is the parameter introduced to describe the equation, which is:

[0038]

[0039] In the formula, σ s is the scattering coefficient; C is the anisotropic phase function coefficient.

[0040] In step 3), the operating conditions range from the operating conditions when the boiler is operating at a 30-100% BMCR load; the furnace combustion inlet parameter data includes boiler load data, primary air volume data, primary air temperature data, secondary air volume data, secondary air temperature data, burnout air volume data, burnout air temperature data, coal powder amount data, furnace inlet pressure data at the burner, and element content data of C, H, O, N, and P of the burned coal based on the received basis and low calorific value data of coal powder.

[0041] In step 5), the actual operating condition is the operating condition when the boiler is operated at a BMCR load of 30 to 100%.

[0042] A furnace outlet NOx prediction system based on numerical simulation, comprising a DCS data processing unit, a calculation simulation unit, a database unit, a NOx prediction model establishment unit, a furnace actual working condition acquisition unit and a NOx prediction unit;

[0043] The DCS data processing unit is used to extract the furnace combustion boundary conditions required for numerical simulation calculation from the DCS system data;

[0044] The calculation simulation unit is used to perform numerical simulation calculation according to the furnace combustion boundary conditions to obtain the furnace outlet NOx concentration under different working conditions;

[0045] The database unit includes furnace combustion inlet parameter data and furnace outlet NOx concentration data under different working conditions;

[0046] The NOx prediction model establishment unit is used to establish a furnace outlet NOx prediction model using a support vector method according to furnace combustion inlet parameter data and furnace outlet NOx concentration data;

[0047] The furnace actual working condition acquisition unit is used to acquire the real-time inlet parameters of the furnace under actual working conditions;

[0048] The NOx prediction unit is used to predict the NOx concentration at the furnace outlet in real time according to the real-time furnace inlet parameters and the NOx prediction model.

[0049] The calculation simulation unit includes a furnace model establishment unit and a numerical calculation unit; the furnace model establishment unit is used to establish a 1:1 physical model of the furnace according to the actual size of the furnace; the furnace combustion models used by the numerical calculation unit include: Realizable k-ε turbulence model, particle motion model, gas phase combustion model, volatile analysis model, coke combustion model, radiation heat exchange model and NOx generation and reduction model.

[0050] The beneficial effects of the present invention are as follows:

[0051] The prediction method provided by the present invention can predict the change of NOx at the furnace outlet and the NOx distribution in the furnace according to the real-time inlet parameters of the furnace under actual working conditions when the combustion conditions change, that is, when the operating parameters such as air / powder are adjusted, so as to provide advance guidance for SCR ammonia injection, which is of great significance to improving the denitrification efficiency of coal-fired power plants, the safety of air preheater operation, and improving the economy. BRIEF DESCRIPTION OF THE DRAWINGS

[0052] Figure 1 It is a structural diagram of a prediction system according to a specific embodiment of the present invention.

[0053] Figure 2 This is a schematic diagram of a NOx prediction model established by a support vector machine according to a specific embodiment of the present invention.

[0054] Figure 3 This is a diagram of the user interface of the prediction system according to a specific embodiment of the present invention.

[0055] In the figure: 101, DCS data processing unit; 102, calculation simulation unit; 103, database unit; 104, NOx prediction model establishment unit; 105, furnace actual working condition acquisition unit; 106, NOx prediction unit. DETAILED DESCRIPTION

[0056] The specific implementation of the present invention is described below with reference to the accompanying drawings.

[0057] like Figure 1As shown, a furnace outlet NOx prediction system based on numerical simulation in this embodiment includes a prediction model part and a real-time prediction part. The prediction model part includes a DCS data processing unit 101, a calculation simulation unit 102, a database unit 103 and a NOx prediction model establishment unit 104, and the real-time prediction part includes a furnace actual working condition acquisition unit 105 and a NOx prediction unit 106;

[0058] The DCS data processing unit 101 is used to extract the furnace combustion boundary conditions required for numerical simulation calculation from the distributed control system DCS (Distributed Control System, DCS) of the coal-fired power plant;

[0059] The calculation simulation unit 102 is used to perform numerical simulation calculations according to the furnace combustion boundary conditions to obtain the furnace outlet NOx concentration under different working conditions;

[0060] Database unit 103, including furnace combustion inlet parameter data and furnace outlet NOx concentration data under different working conditions;

[0061] The NOx prediction model building unit 104 is used to build a furnace outlet NOx prediction model using a support vector method according to the furnace combustion inlet parameter data and the furnace outlet NOx concentration data;

[0062] A furnace actual working condition acquisition unit 105 is used to acquire the real-time inlet parameters of the furnace under actual working conditions;

[0063] The NOx prediction unit 106 is used to predict the NOx concentration at the furnace outlet in real time according to the real-time furnace inlet parameters and the NOx prediction model.

[0064] The user interface of the above prediction system when it is implemented is as follows: Figure 3 The prediction system can predict the NOx concentration at the furnace outlet in real time according to the real-time furnace inlet parameters under actual working conditions, and provide guidance to users.

[0065] Among them, the calculation simulation unit 102 includes a furnace model building unit and a numerical calculation unit; the furnace model building unit is used to build a 1:1 physical model of the furnace according to the actual size of the furnace; the furnace combustion models used by the numerical calculation unit include: Realizable k-ε turbulence model, particle motion model, gas phase combustion model, volatile analysis model, coke combustion model, radiation heat exchange model and NOx generation and reduction model.

[0066] Based on the above-mentioned furnace outlet NOx prediction system based on numerical simulation, this embodiment also proposes a furnace outlet NOx prediction method based on numerical simulation, comprising the following steps:

[0067] 1) Extract the furnace combustion boundary conditions required for numerical simulation calculation from the DCS system data, and the boundary conditions include the furnace combustion inlet parameter data;

[0068] 2) Numerical simulation calculations are performed based on the furnace combustion boundary conditions to obtain the furnace outlet NOx concentration under different working conditions, including:

[0069] According to the actual size of the furnace, a 1:1 physical model of the furnace is established, including the water-cooled wall, burner, overburnt air, and superheater of the furnace;

[0070] Construct a furnace combustion model to describe the combustion process of coal in a large pulverized coal furnace, including particle movement, pyrolysis of pulverized coal, gas-phase turbulent combustion, combustion of coke particles, heat transfer between gas and solid phases, and radiation heat transfer. The furnace combustion model specifically includes: Realizable k-ε turbulence model, particle movement model, gas-phase combustion model, volatile analysis model, coke combustion model, radiation heat transfer model, and NOx generation and reduction model;

[0071] Among them, the gas phase combustion model adopts a non-premixed combustion model. Constructing and solving the non-premixed combustion model includes the following steps:

[0072] Based on the atomic mass fraction, the mixing fraction is defined as:

[0073]

[0074] In the formula, Z i is the element mass fraction of element i; the subscript ox represents the value at the oxidant inlet, and the subscript fuel represents the value at the fuel inlet;

[0075] The time-averaged mixture fraction equation is established based on the mixture fraction:

[0076]

[0077] In the formula, S m is the mass of coal powder particles transferred into the gas phase; S user For custom source items;

[0078] Solve the conservation equation for the mean square value of the average mixture fraction using the following equation:

[0079]

[0080] In the formula, σ i =0.85, C g =286, C d =2.0.

[0081] The above model is solved by solving the transport equations of one or two conserved quantities, and the concentration of a single component is solved by the predicted mixture fraction distribution. There is no need to solve a large number of component transport equations, thereby improving efficiency.

[0082] The coke combustion model uses the diffusion-dynamic controlled reaction rate model, taking into account the effects of diffusion and reaction kinetics on the particle surface reaction rate, and combining the rate of oxygen diffusion to the coke surface and the chemical reaction kinetics rate on the coke surface, the weighted calculation results in the coke surface reaction combustion rate:

[0083]

[0084] In the above formula, m p , A p are the mass and area of ​​the droplet respectively; ρ is the gas density; R is the gas molar constant; T ∞ is the ambient medium temperature; Y ox is the mass fraction of the oxidant in the local gas; D0 is the diffusion rate coefficient; is the kinetic rate; M W，OX is the molar mass of the oxidant;

[0085] Among them, the particle motion model uses a discrete phase model to describe the gas-solid two-phase flow in the boiler, treating the gas as a continuous phase and the pulverized coal particles as discrete phases; a random particle orbit model is used to establish the particle instantaneous equations under Lagrangian coordinates, thereby calculating the motion and orbits of different pulverized coal particles. Since the movement of pulverized coal particles is accompanied by chemical reaction processes such as inhomogeneous combustion with phase change, the Lagrangian random orbit model has a good effect in describing the process, and is established according to the following formula:

[0086]

[0087] In the formula, u ip is the velocity component of the particle in a certain direction; u′ ig are the average velocity and pulse velocity of the gas phase in a certain direction respectively; m p , A p are the particle mass and surface area, respectively; ρ g is the local gas density; g k is the acceleration due to gravity; C D is the drag coefficient;

[0088] Among them, the volatilization analysis model adopts a single-step pyrolysis model and is established according to the following formula:

[0089]

[0090] In the formula, m p represents the particle mass, fv，0 is the initial volatile mass fraction of the particles, f w，0 is the mass fraction of the evaporable and boiling substance in the particles; m p，0 is the initial mass of the particle; k is the reaction rate constant;

[0091] Among them, the radiation heat transfer model adopts the P-1 model, which can better describe the radiation heat transfer in the furnace, and its local incident radiation solution equation is:

[0092]

[0093] Where n is the refractive index of the medium, σ is the Stefan-Boltzmann constant, and S G is the user-defined radiation source, a is the absorption coefficient, G is the incident radiation, T is the temperature, and Γ is a parameter introduced to describe the equation. The formula is:

[0094]

[0095] In the formula, σ s is the scattering coefficient; C is the anisotropic phase function coefficient.

[0096] Among them, the NOx generation and reduction model only considers thermal NOx and fuel NOx when NOx is generated by coal combustion. Thermal NOx is described by the extended Zeldovich mechanism, and the generation path of fuel NOx is described by the De Scoete mechanism. After the thermal calculation is completed, a good flow field and temperature field are obtained, and then the post-processing method is used in the numerical simulation software (Fluent software) to calculate the NOx distribution in the furnace. Finally, the outlet plane is intercepted to calculate the average NOx concentration at the furnace outlet.

[0097] 3) Establish a database for furnace combustion inlet parameter data and furnace outlet NOx concentration data under different working conditions;

[0098] Specifically, the operating condition ranges from the operating condition when the boiler operates at 30 to 100% BMCR load.

[0099] Specifically, the furnace combustion inlet parameter data include boiler load data, primary air volume data, primary air temperature data, secondary air volume data, secondary air temperature data, burnout air volume data, burnout air temperature data, coal powder amount data, furnace inlet pressure data at the burner, and element content data of C, H, O, N, and P of the burned coal based on the received basis and low calorific value data of coal powder.

[0100] 4) Based on the furnace combustion inlet parameter data and furnace outlet NOx concentration data in the database, a furnace outlet NOx prediction model is established using the support vector machine method;

[0101] Specifically, the independent variable is the furnace combustion inlet parameter data, the dependent variable is the furnace outlet NOx concentration data, and the regression function and the normal vector of the regression function are:

[0102]

[0103]

[0104] In the formula, α i , α i * is the Lagrange multiplier and b is the displacement term.

[0105] The Gaussian kernel function with strong nonlinear mapping ability is selected as the kernel function of the support vector machine to adapt to the complex combustion environment inside the boiler. The schematic diagram of the NOx prediction model established based on the support vector machine is shown in Figure 2 shown.

[0106] 5) Obtain the real-time furnace inlet parameters under actual working conditions;

[0107] Specifically, the actual operating condition is the operating condition when the boiler operates at 30-100% BMCR load.

[0108] 6) Use the NOx prediction model to predict the NOx concentration at the furnace outlet in real time based on the real-time furnace inlet parameters.

[0109] The prediction method provided by the present invention can predict the change of NOx at the furnace outlet and the NOx distribution in the furnace when the combustion conditions change, that is, when the operating parameters such as air / powder are adjusted, and provide advance guidance for SCR ammonia injection, which is of great significance to improving the denitrification efficiency of coal-fired power plants, the safety of air preheater operation, and improving the economy.

Claims

1. A method for predicting furnace outlet NOx based on numerical simulation, characterized in that: The following steps are involved: 1) Extracting furnace combustion boundary conditions required for numerical simulation calculation from the DCS system, wherein the boundary conditions include furnace combustion inlet parameter data; 2) Numerical simulation calculation is performed according to the furnace combustion boundary conditions to obtain the furnace outlet NOx concentration under different working conditions, specifically including: According to the actual size of the furnace, establish the physical model of the furnace, including the water-cooled wall, burner, overburnt air, and superheater of the furnace; Construct furnace combustion model, including: Realizable k-ε turbulence model, particle motion model, gas phase combustion model, volatilization analysis model, coke combustion model, radiation heat transfer model and NOx generation and reduction model; 3) Establish a database of furnace combustion inlet parameter data and corresponding furnace outlet NOx concentration data under different operating conditions; 4) Taking the furnace combustion inlet parameter data as the independent variable and the furnace outlet NOx concentration data as the dependent variable, the support vector machine method was used to establish the furnace outlet NOx prediction model; 5) Obtain the real-time furnace inlet parameters under actual working conditions; 6) using the NOx prediction model to predict the NOx concentration at the furnace outlet in real time according to the furnace real-time inlet parameters; In step 2), the gas phase combustion model adopts a non-premixed combustion model, and constructing and solving the non-premixed combustion model includes the following steps: Based on the atomic mass fraction, the mixing fraction is defined as: In the formula, Z i is the element mass fraction of element i; the subscript ox represents the value at the oxidant inlet, and the subscript fuel represents the value at the fuel inlet; The time-averaged mixture fraction equation is established based on the mixture fraction: In the formula, S m is the mass of coal powder particles transferred into the gas phase; S user For custom source items; Solve the conservation equation for the mean square value of the average mixture fraction using the following equation: In the formula, σ i =0.85, C g =286, C d =2.0; In step 2), the coke combustion model uses a diffusion-dynamic controlled reaction rate model, taking into account the effects of diffusion and reaction kinetics on the particle surface reaction rate, and combining the rate at which oxygen diffuses to the coke surface and the chemical reaction kinetics rate on the coke surface, and the weighted calculation to obtain the coke surface reaction combustion rate is: In the above formula, m p , A p are the mass and area of ​​the droplet respectively; ρ is the gas density; R is the gas molar constant; T ∞ is the ambient medium temperature; Y ox is the mass fraction of the oxidant in the local gas; D0 is the diffusion rate coefficient; is the kinetic rate; M W,OX is the molar mass of the oxidant; In step 2), the particle motion model adopts a random particle trajectory model and is established according to the following formula: In the formula, u ip is the velocity component of the particle in a certain direction; u i ' g are the average velocity and pulse velocity of the gas phase in a certain direction respectively; m p , A p are the particle mass and surface area, respectively; ρ g is the local gas density; g k is the acceleration due to gravity; C D is the drag coefficient; The volatile analysis model adopts a single-step pyrolysis model and is established according to the following formula: In the formula, m p represents the particle mass, f v,0 is the initial volatile mass fraction of the particles, f w,0 is the mass fraction of the evaporable and boiling substance in the particles; m p,0 is the initial mass of the particle; k is the reaction rate constant; In step 2), the radiation heat transfer model adopts the P-1 model, and the local incident radiation solution equation is: Where n is the refractive index of the medium, σ is the Stefan-Boltzmann constant, and S G is a user-defined radiation source, a is an absorption The absorption coefficient, G is the incident radiation, T is the temperature, and Г is the parameter introduced to describe the equation. The formula is: In the formula, σ s is the scattering coefficient; C is the anisotropic phase function coefficient.

2. The method for predicting furnace outlet NOx based on numerical simulation according to claim 1, characterized in that: In step 3), the operating conditions range from the operating conditions when the boiler is operating at a 30-100% BMCR load; the furnace combustion inlet parameter data includes boiler load data, primary air volume data, primary air temperature data, secondary air volume data, secondary air temperature data, burnout air volume data, burnout air temperature data, coal powder amount data, furnace inlet pressure data at the burner, and element content data of C, H, O, N, and P of the burned coal based on the received basis and low calorific value data of coal powder.

3. The method for predicting furnace outlet NOx based on numerical simulation according to claim 1, characterized in that: In step 5), the actual operating condition is the operating condition when the boiler is operated at a BMCR load of 30 to 100%.

4. A furnace outlet NOx prediction system based on numerical simulation, characterized in that: It includes DCS data processing unit, calculation simulation unit, database unit, NOx prediction model establishment unit, furnace actual working condition acquisition unit and NOx prediction unit; The DCS data processing unit is used to extract the furnace combustion boundary conditions required for numerical simulation calculation from the DCS system data; The calculation simulation unit is used to perform numerical simulation calculation according to the furnace combustion boundary conditions to obtain the furnace outlet NOx concentration under different working conditions, specifically including: establishing a physical model of the furnace according to the actual size of the furnace, including the water-cooled wall, burner, burnout air, and superheater of the furnace; constructing a furnace combustion model, including: Realizable k-ε turbulence model, particle motion model, gas phase combustion model, volatilization analysis model, coke combustion model, radiation heat exchange model and NOx generation and reduction model; The database unit includes furnace combustion inlet parameter data and furnace outlet NOx concentration data under different working conditions; The NOx prediction model establishment unit is used to establish a furnace outlet NOx prediction model using a support vector machine method according to the furnace combustion inlet parameter data and the furnace outlet NOx concentration data; The furnace actual working condition acquisition unit is used to acquire the real-time inlet parameters of the furnace under actual working conditions; The NOx prediction unit is used to predict the NOx concentration at the furnace outlet in real time according to the furnace real-time inlet parameters and the NOx prediction model; The gas phase combustion model adopts a non-premixed combustion model, and constructing and solving the non-premixed combustion model includes the following steps: Based on the atomic mass fraction, the mixing fraction is defined as: In the formula, Z i is the element mass fraction of element i; the subscript ox represents the value at the oxidant inlet, and the subscript fuel represents the value at the fuel inlet; The time-averaged mixture fraction equation is established based on the mixture fraction: In the formula, S m is the mass of coal powder particles transferred into the gas phase; S user For custom source items; Solve the conservation equation for the mean square value of the average mixture fraction using the following equation: In the formula, σ i =0.85, C g =286, C d =2.0; The coke combustion model adopts the diffusion-dynamic controlled reaction rate model, taking into account the influence of diffusion and reaction kinetics on the particle surface reaction rate, and combining the rate of oxygen diffusion to the coke surface and the chemical reaction kinetics rate on the coke surface, the coke surface reaction combustion rate is obtained by weighted calculation: In the above formula, m p , A p are the mass and area of ​​the droplet respectively; ρ is the gas density; R is the gas molar constant; T ∞ is the ambient medium temperature; Y ox is the mass fraction of the oxidant in the local gas; D0 is the diffusion rate coefficient; is the kinetic rate; M W,OX is the molar mass of the oxidant; The particle motion model adopts a random particle trajectory model and is established according to the following formula: In the formula, u ip is the velocity component of the particle in a certain direction; u i ' g are the average velocity and pulse velocity of the gas phase in a certain direction respectively; m p , A p are the particle mass and surface area, respectively; ρ g is the local gas density; g k is the acceleration due to gravity; C D is the drag coefficient; The volatile analysis model adopts a single-step pyrolysis model and is established according to the following formula: In the formula, m p represents the particle mass, f v,0 is the initial volatile mass fraction of the particles, f w,0 is the mass fraction of the evaporable and boiling substance in the particles; m p,0 is the initial mass of the particle; k is the reaction rate constant; The radiation heat transfer model adopts the P-1 model, and the local incident radiation solution equation is: Where n is the refractive index of the medium, σ is the Stefan-Boltzmann constant, and S G is a user-defined radiation source, a is an absorption The absorption coefficient, G is the incident radiation, T is the temperature, and Г is the parameter introduced to describe the equation. The formula is: In the formula, σ s is the scattering coefficient; C is the anisotropic phase function coefficient.

Images