Optimization method and device for simulating SNCR process of thermal power plant based on CFD

Abstract

The invention relates to the technical field of thermal power plant denitration, in particular to a thermal power plant SNCR (selective non-catalytic reduction) process based on CFD (computational fluid dynamics) simulation for building a physical geometric model of a cyclone separator, optimizing parameters such as injection flow, injection speed and nozzle distribution to realize optimal mixing uniformity and denitration efficiency, and determining that the optimal ammonia nitrogen ratio NSR of SNCR is equal to 1.25 when the SNCR reaction temperature is 972 DEG C; by simulating four injection source arrangement schemes, the optimal injection source position is determined, the optimal mixing uniformity and denitration efficiency and the lowest ammonia escape amount are achieved, the denitration efficiency of the optimized cyclone separator SNCR denitration system reaches 81.2%, emission of NOx in pellet flue gas is effectively reduced, pollution to the environment is effectively reduced through the optimized injection scheme, and the economic benefits of the cyclone separator SNCR denitration system are improved. And the optimal injection source position is determined, the actual engineering error value after transformation is smaller than 0.03, and the environment-friendly requirement for flue gas emission is met.

Description

Technical Field

[0001] The present invention relates to the technical field of flue gas denitrification, and particularly relates to an optimization of related parameters such as injection flow rate, injection speed, and nozzle distribution of an SNCR system added to a cyclone separator of a thermal power plant based on CFD simulation. Background Art

[0002] Nitrogen oxides (NO x ), mainly including nitric oxide (NO) and nitrogen dioxide (NO2), are components of air pollutants that particularly seriously affect the environment and human health; they can not only cause respiratory diseases, but also participate in the formation of acid rain, photochemical smog, and even damage the ozone layer; especially coal-fired power plants produce a large amount of nitrogen oxides due to high-temperature combustion during the energy conversion process, becoming an important pollution source.

[0003] Currently, the mainstream technologies for controlling nitrogen oxide emissions in thermal power plants include low-nitrogen combustion technology and flue gas denitrification technology; low-nitrogen combustion technology reduces the generation of nitrogen oxides by optimizing combustion conditions, such as air staging and fuel staging; while the flue gas denitrification technology selective non-catalytic reduction (SNCR) uses a catalyst or a chemical reducing agent to convert NO x into harmless nitrogen and water; although the nitrogen oxide emissions are reduced to a certain extent, there are still some limitations. For example, low-nitrogen combustion may affect the combustion efficiency, and the SNCR technology faces many challenges such as high operating costs, reducing agent supply and treatment problems.

[0004] In the SNCR process of thermal power plants, the flue gas denitrification technology faces: denitrification efficiency, reaction temperature window, ammonia slip, equipment wear and maintenance, cost control, operation complexity, environmental regulation restrictions, by-product treatment, and space layout restrictions; although the SNCR technology has relatively low initial investment, its denitrification efficiency is usually inferior to that of the SCR technology, especially in the case of high nitrogen oxide concentration or low flue gas temperature.

[0005] In the SNCR process of thermal power plants, the reaction of the flue gas denitrification technology must be carried out within a narrow temperature window of 900°C to 1100°C, and exceeding this range will significantly reduce the efficiency.

[0006] The selective non-catalytic reduction (SNCR) technology has become a feasible option for pellet flue gas denitrification due to its effectiveness within a wide temperature range and the absence of a catalyst requirement; however, there are also some problems in the actual application of the SNCR technology, such as the reduction efficiency being affected by factors such as temperature, residence time, and mixing uniformity, as well as the ammonia slip problem.

[0007] Regarding the current efficiency and device problems, especially the emission of NO x , an effective SNCR optimization scheme and good removal efficiency are required. Summary of the Invention

[0008] In view of the problems existing in the prior art, the present invention proposes to set up an SNCR reaction system in a cyclone separator, optimize parameters such as injection flow rate, injection speed and nozzle distribution, and at the same time optimize the actual engineering transformation according to the simulated SNCR system configuration.

[0009] The reaction temperature window of SNCR is 950 - 1100 °C, and the optimized optimal temperature is 972 °C.

[0010] The SNCR denitration method for pellet flue gas with different ammonia-nitrogen ratios (NSR) is optimized, and ammonia slip is reduced at the same time.

[0011] By dispersing the flow rate of a single spray gun to multiple spray guns, reducing the flow rate of a single spray gun, and reducing the disturbance of the droplet beam on the flue gas flow field; arranging the spray guns on both sides of the cyclone separator inlet, making the injection direction orthogonal to the flue gas flow direction, and the two clusters of droplet flows collide to offset the kinetic energy and reduce the disturbance of the flue gas flow field; using the rotating flow field of the cyclone separator, arranging the spray guns inside the cyclone separator to improve the mixing efficiency of the reducing agent and the flue gas, and obtaining the optimal NSR = 1.25.

[0012] Four spray gun arrangement schemes are designed. Six spray guns are arranged on the boiler outlet plane, the injection angle is 60°, the flow velocity is 1 m / s, and the flow rate is 0.24 kg / s. It is simulated that Scheme 3 is the best, and the spray guns are arranged on the cyclone separator outlet plane to improve the mixing uniformity by using the rotating flow field of the cyclone separator.

[0013] Select the optimal NSR and spray source arrangement scheme, and compare the simulated and actual working condition data. There is a 2% difference in denitration efficiency, and the difference in ammonia escape concentration is 1.6 mg / m 3

[0014] Use Design Modeler to establish a 1:1 model of the cyclone separator. The 1st, 2nd, and 3rd separators are along the Y-axis direction. Use the MESH module to mesh the model and conduct mesh independence verification. Finally, it is divided into 239,849 units.

[0015] The simulations include: (1) the flow of flue gas in the fluid domain; (2) the injection, atomization, evaporation and pyrolysis of the reducing agent into a gaseous reducing agent; (3) the mixing of flue gas and gaseous reducing agent; (4) the reaction of flue gas and reducing agent. The mathematical models and control equations used include turbulence models, component transport and reaction models, discrete phase models, etc.

[0016] The initial boundary conditions for simulation are: velocity 1 m / s, turbulence intensity 10%, hydraulic diameter 14.56 m, temperature 927 °C, outlet pressure 0, and the mass fractions of each component refer to the actual flue gas composition.

[0017] The specific reaction mechanism involved in SNCR.

[0018] The concentration distribution simulated by CFD.

[0019] Compared the weighted uniformity of three cross-sectional areas of K for 3 injection sources × injection rate of 0.1 kg / s and 6 injection sources × injection rate of 0.05 kg / s.

[0020] Designed four schemes by changing the spray gun direction and relative position: Arrange 6 spray guns on the outlet plane of the cyclone separator, with a spray angle of 60°, a flow rate of 1 m / s, a flow rate of 0.24 kg / s, and the spray directions are orthogonal and parallel to the flue gas direction respectively to optimize the mixing uniformity of the reducing agent and the flue gas.

[0021] The simulated module is based on the CFD simulation of the three-dimensional model to obtain the simulation results corresponding to the test scheme.

[0022] The average flue gas value at the outlet of the SNCR reactor under the actual working conditions after transformation on a certain day. Description of the Drawings

[0023] Figure 1 It is a flow chart of the optimization method for the SNCR process of the cyclone separator in a thermal power plant based on CFD simulation provided by the present invention.

[0024] Figure 2 It is a physical geometry model diagram constructed by Design Modeler.

[0025] Figure 3 It is a physical geometry model size diagram constructed by Design Modeler.

[0026] Figure 4 It is a NO concentration distribution diagram without injecting the reducing agent.

[0027] Figure 5 It is a diagram of the change of NO and NH3 with NSR at the outlet.

[0028] Figure 6 It is a velocity flow field diagram inside the model at different incident velocities.

[0029] Figure 7 It is a diagram of the NH3 distribution at the central cylinder cross-section and the K distribution diagram at the separator inlet cross-section for 3 injection sources (a, c) and 6 injection sources (b, d).

[0030] Figure 8 It is a diagram of the injection source distribution for 4 schemes. Detailed Implementation Manner

[0031] The following are only some embodiments of the present invention. Other technicians can obtain other equivalent embodiments according to the principles and methods of the present invention without creative work, and all of them fall within the protection scope of the present invention.

[0032] Figure 1 It is a method flow chart of an embodiment of the cyclone SNCR system based on CFD simulation provided by the present invention. As Figure 1 shown, the optimization method of the cyclone based on CFD simulation includes:

[0033] Use Design Modeler software to build a 1:1 physical geometry model of the cyclone as Figure 2 shown.

[0034] The models adopted include SST turbulence model, volume reaction model, EDC model, NO x generation model, SNCR reaction model, DPM and other mathematical models, component transport and reaction model mechanism, discrete phase model, etc.

[0035] When dealing with three-dimensional turbulent reacting flows of gases in the computational domain, two-equation models are widely used in industrial flow simulations. These models determine the length and time scales of turbulence by solving two independent transport equations respectively; in ANSYS Fluent, there are two series of two-equation models, and they each have their own characteristics in dealing with turbulent flows.

[0034] The standard two-equation model (such as the k-ε model) has limitations in dealing with certain flow situations, such as the strong sensitivity of the free stream values inside and outside the shear layer, resulting in a decrease in prediction accuracy under complex flow conditions. Therefore, the standard model is usually not recommended in ANSYS Fluent.

[0035] To overcome these limitations, the BSL (Baldwin-Barth) and SST (Shear Stress Transport) models have been developed. The BSL model reduces the sensitivity of the standard model to the free stream by combining elements of the k-ε model and the k-ω model. The SST model is specially calibrated based on the k-ω model to better handle the cases of adverse pressure gradients and flow separation. The advantage of the SST model is that it combines the performance of the k-ε model in the region far from the wall and the advantages of the k-ω model in the near-wall region, enabling high prediction accuracy in a wide range of flow regions.

[0036] Since the SNCR (Selective Non-Catalytic Reduction) reaction computational domain involves a variable cross-section flow region, the streamline curvature changes greatly, which is prone to generate vortices or other additional strains, resulting in enhanced turbulent diffusion. The standard model cannot effectively describe the flow characteristics in this case. Therefore, turbulent simulation is carried out through the SST model to obtain more accurate results.

[0037] Turbulent kinetic energy (k) equation:

[0038] Specific dissipation rate (ω) equation:

[0039] where ρ is the fluid density, kg / m 3 ; t is time, s; k is the kinetic energy per unit mass, J / kg; ω is the specific diffusivity, s -1 ; u i is the velocity in the i direction, m / s; x i and x j are the displacements in the i and j directions respectively, m; Γ k and Γ ω are the effective diffusivities of k and ω respectively; G k and G ω represent the generation of k and ω respectively; Y k and Y ω represent the dissipation of k and ω respectively; D ω represents the cross-diffusion term; S k and S ω are user-defined source terms; G k and G ωb address the buoyancy term.

[0040] The species transport model is that when choosing to solve the conservation equation of chemical substances, ANSYS Fluent predicts the local mass fraction Y of each substance by solving the convection-diffusion equation of the i-th species i .

[0041] The general form of the conservation equation adopted is:

[0042] The finite rate model is the Eddy-dissipation-concept (EDC) model: Considering the time scales of turbulence and kinetics, detailed chemical kinetics is incorporated into turbulence.

[0043] When performing CFD (Computational Fluid Dynamics) simulation on the SNCR (Selective Non-Catalytic Reduction) denitration reaction mechanism, there are three widely recognized reaction mechanisms, namely:

[0044] Thermal DeNO x mechanism: Using ammonia (NH3) as a reducing agent, reacting it with nitrogen oxides (NO x ) at high temperature to achieve the effect of denitration; Ammonia has a high NO x reduction ability, but due to its toxicity and storage safety issues, as well as concerns about colorless and toxic NH3 leakage, it is often replaced by other reducing agents in practical applications.

[0045] NO x OUT mechanism: Using urea (CO(NH2)2) as a reducing agent, ammonia is generated through the injection and decomposition of an aqueous urea solution, and then reacts with NO x Reaction. Urea is safer than ammonia and is easier to store and transport, so it is more common in industrial applications.

[0046] RapreNO x Mechanism: Using cyanuric acid (C3H2N3O3) as a reducing agent, the ammonia and cyanate ions generated by its thermal decomposition react with NO x Reaction to achieve denitrification.

[0047] Numerical simulation calculations were carried out using ANSYS 2021R1 software. The reaction mechanism of its SNCR calculation module includes a two-step urea decomposition mechanism and a seven-step reduction kinetics mechanism, which effectively simulate urea decomposition and NO x Reduction process.

[0048] By inputting the parameter data, the SNCR denitrification effect under different working conditions can be simulated, providing a scientific basis for the design and optimization of the SNCR device.

[0049] Table 1: The reaction mechanism includes a two-step urea decomposition mechanism and a seven-step reduction kinetics mechanism.

[0050]

[0051] Where A is the pre-exponential factor; b is the temperature exponent; E is the activation energy of the reaction; M is an unknown inert component that does not participate in the simplified kinetic mechanism.

[0052] In CFD simulation, the interaction of the reducing agent particles injected into the fluid domain can be ignored, and the volume fraction of the particles relative to the entire fluid domain is very small. The Lagrangian discrete phase model of ANSYS Fluent combines the Euler method (for the continuous fluid phase) and the Lagrangian method (for the discrete particle phase) to handle.

[0053] The simulation describes the continuous fluid phase by solving the Navier-Stokes equation, and simulates the discrete particle phase by tracking the motion of each particle, where the particle phase exchanges momentum, mass, and energy with the continuous phase, and is considered through the corresponding physical models and boundary conditions.

[0054] Using the Lagrangian discrete phase model to simulate fluid flow macroscopically and simultaneously track and calculate the behavior of particles microscopically, effectively simulating the injection process and chemical reactions.

[0055] The vector expression of the N-S equation involved in the discrete phase model (DPM) is:

[0056] Among them, V is the velocity vector; g is the acceleration due to gravity; μ is the dynamic viscosity; ρ is the fluid density; p is the pressure.

[0057] Boundary conditions are the mathematical and physical conditions that the flow field variables on the calculation boundary must satisfy. Together with the initial conditions, they constitute the definite solution conditions to ensure the existence and uniqueness of the flow field solution. In the Fluent software, the initial conditions are set during the initialization process, while the fluid inlet and outlet conditions and wall conditions (boundary conditions) need to be set separately.

[0058] In the setting of this model, the inlet condition is set as a velocity inlet, with a velocity of 1 m / s, a turbulence intensity of 10%, a hydraulic diameter of 14.56 m, and a temperature of 927 °C. The specific numerical values of the mass fractions of each component are shown in Table 4. The outlet condition is set as a pressure outlet, with a gauge pressure of 0.

[0059] Table 2 shows the components of the inlet flue gas.

[0060]

[0061] Figure 4 It means that under this boundary condition, without setting an injection source, after initializing the model and performing 1500 steps of iterative calculation, the NO distribution map at the flue gas inlet and the central axis section of the cyclone separator is obtained, and the area-weighted average NO concentration of each surface in the output result is filled into Table 3.

[0062] Table 3 shows the area-weighted average NO concentration of each surface.

[0063]

[0064] Among them, the NO concentration at the inlet is similar to the distribution in the boiler without taking SNCR measures. The calculated NO flow rate at the inlet is 134 mg / s. The NO concentration distribution in the three cyclone separators is uniform, so an equal-flow reductant injection source is set in each separator.

[0065] The normalized stoichiometric ratio (NSR) is obtained from the actual molar ratio of urea to initial NO and the stoichiometric molar ratio of urea to initial NO. A urea solution with a mass concentration of 10% is selected as the reductant. When the NO flow rate is 0.06 kg / s and NSR = 1, the total reductant flow rate should be 0.6 kg / s, and 0.2 kg / s of reductant is required in each cyclone separator.

[0066] When the NSR is between 0.5 and 2, it is more appropriate. Therefore, the NSR is selected to be 0.5, 0.75, 1, 1.25, 1.5, 1.75, and 2 gradient flow rates of the reducing agent injection to investigate the influence of the injection volume on the reduction efficiency and ammonia escape; a jet source is set at the top of the intake pipe of each cyclone separator, which is set to be conical, numbered 1, 2, and 3 respectively. The jet source parameters are shown in Table 4; the injection speed is 1 m / s, and the results under different flow rates are as Figure 5 shown.

[0067] Table 4 shows the jet source parameters.

[0068]

[0069] From Figure 5 observation, it can be seen that as the flow rate of the jet source increases, the NO concentration at the outlet decreases significantly. When the flow rate reaches 0.25 kg / s (NSR = 1.25), the decreasing trend of the NO concentration slows down, and even a slightly rising phenomenon appears.

[0070] During the process of the flow rate of the jet source increasing from 0.1 kg / s to 0.25 kg / s, the NH3 concentration at the outlet increases sharply, and then the rising trend slows down. It shows that when the flow rate is less than 0.25 kg / s, the denitrification efficiency is proportional to the ammonia-nitrogen ratio, the reduction reaction of NO dominates, and a large amount of NH3 is consumed at the same time.

[0071] When the NSR exceeds 1.25, the NO reduction efficiency reaches the limit, and the growth rate of NH3 generated by the decomposition of the urea solution exceeds the consumption rate of NH3 for reducing NO, resulting in an abnormal increase in the NO concentration at the outlet; therefore, the optimal ammonia-nitrogen ratio should be selected as NSR = 1.25. At this time, the denitrification efficiencies at the three outlets are 81.00%, 81.63%, and 82.74% respectively, and the ammonia escape amounts are 1.76 mg / s, 2.08 mg / s, and 1.09 mg / s, all within a reasonable range.

[0072] Different injection volumes lead to changes in the flow field in the reaction zone. Therefore, the differences in the internal velocity flow field of the model when the single jet source flow rate is 0 kg / s and 0.25 kg / s are compared. See Figure 6 .

[0073] Figure 6 Note: The jet source flow rates of a and b are 0 kg / s and 0.25 kg / s respectively

[0074] From Figure 7 comparison, it is obtained that due to the influence of the jet source on the flow field, the streamlines near and inside the inlet of the cyclone separator become uneven, and there is a more obvious difference in the ammonia escape amount. Therefore, an additional jet source is added to the side of each cyclone separator intake channel far from the center line, numbered 4, 5, and 6 respectively, and the cone angle is set to 60°. The position parameters of the three jet sources are shown in Table 5.

[0075] Table 5 shows the position and injection direction parameters of a single injection source on the side.

[0076]

[0077] By increasing the number of injection sources, the flow rate of a single injection source is reduced, thereby reducing the influence of the reducing agent droplet beam on the flow field. When NSR = 1, the flow rates of injection sources No. 1 - 6 are evenly distributed at 0.05 kg / s, and the injection flow velocities are all 1 m / s. Based on the distribution of the turbulent intensity K of the cylindrical cross-section at the inlet of the cyclone separator, the influence degree of the injection source on the flow field turbulence is judged. The area-weighted uniformity K1, K2, and K3 of K at the three cross-sections are compared in Table 6, and the NH3 distribution and K distribution diagrams are shown in Figure 7.

[0078] Table 6 shows the area-weighted uniformity of the turbulent intensity at the inlet cross-section of the cyclone separator under different numbers of injection sources.

[0079]

[0080] According to Figure 7 the comparison between a and b shows that the increase in the injection source reduces the NH3 concentration gradient at the inlet and the central cylinder, indicating an improvement in the mixing effect. NH3 diffuses more widely and has a longer residence time in the lower part of the cyclone separator; Figure 7 c and d in

[0081] Figure 7 show that the reduction in the injection amount reduces the droplet kinetic energy and causes less disturbance to the flue gas flow. At the same time, the orthogonal injection sources interact with each other, reducing the kinetic energy and further reducing the disturbance to the flow field. The increase in the number of injection sources reduces the difference in the inlet turbulent intensity, making the distribution more reasonable and the NH3 distribution more uniform.

[0082] Four designed injection direction schemes are provided to reduce the disturbance of the injection source to the flue gas flow field and improve the mixing uniformity of the reducing agent. Each cyclone separator is equipped with 2 injection sources, a total of 6, with an injection angle of 60°, a flow velocity of 1 m / s, and a flow rate of 0.24 kg / s. The other parameters are the same as those in Table 4. The specific distribution positions are shown in Figure 8 .

[0083] To avoid the reaction between the NH3 decomposed from urea and other gas components in the model, which may affect the accuracy of the simulation results, the nitrogen oxide generation model is turned off, the inlet boundary conditions are modified so that the gas entering is only N2 that will not react with the reducing agent, and the calculation load is reduced to make the iterative calculation converge faster.

[0084] The four distribution schemes set take the area mass integral uniformity index of NH3 at the inlet of the central cylinder as the evaluation standard for the mixing uniformity of the reducing agent, and the simulation results of the four groups are shown in Table 7.

[0085] Table 7 shows the injection source positions and injection directions of the four schemes.

[0086]

[0087] Figure 8 Note: a is Scheme 1, b is Scheme 2, c is Scheme 3, and d is Scheme 4.

[0088] Table 8 shows the NH3 uniformity of each central cylinder under different injection source distributions.

[0089]

[0090] Through comprehensive analysis, compared with Scheme 1, the NH3 distribution uniformity obtained by the other injection source arrangement methods has been improved to a certain extent. It is proved that when the injection direction is consistent with the flue gas flow direction, the mixing effect of the reducing agent and the flue gas is stronger than when they are orthogonal; while the uniformity of Scheme 4 is slightly lower than that of Scheme 2 and Scheme 3 because the two injection sources in the inlet pipe of the same cyclone separator are on the same Y-axis. Therefore, at the same diffusion speed, the reducing agent injected by the two injection sources located on different Y-axes has a larger distribution range in the Z direction and higher uniformity.

[0091] After comprehensive comparison, the optimal injection source arrangement is obtained as follows: NSR = 1.25, the flow rate of each of the 6 injection sources is 0.0781 kg / s, and the position distribution is Scheme 3. They are respectively applied to numerical simulation and the actual boiler, and the data results are shown in Table 9.

[0092] Table 9 shows the comparison of numerical simulation and actual boiler data.

[0093]

[0094] The best design is Scheme 3, the simulated denitration efficiency obtained is 83.4%, the denitration efficiency under actual conditions is 81.2%, and the error is less than 0.03.

[0095] After the optimized design, the denitration efficiency of the boiler has been significantly improved and the ammonia escape amount meets the ultra-low emission standards.

[0096] Those skilled in the art can understand that all or part of the processes of implementing the methods in the above embodiments can be used to guide the transformation of corresponding similar projects through the above steps.

[0097] The above has introduced in detail the optimization method for the injection flow rate, injection velocity, and nozzle distribution of the SNCR system in a cyclone separator based on CFD simulation. In this article, specific examples are used to elaborate on the principle and implementation manner of the present invention. The description of the above embodiments is only used to help understand the method and its core idea of the present invention; at the same time, for those skilled in the art, according to the idea of the present invention, there will be changes in the specific implementation manner and application scope. In summary, the content of this specification should not be construed as a limitation to the present invention.

Claims

1. A method for optimizing the injection flow rate, injection velocity and nozzle distribution of a cyclone separator SNCR system in a thermal power plant based on CFD simulation, characterized in that: The method comprises the following steps: step 1: using Design Modeler software to construct a 1:1 physical geometry model of a cyclone separator; step 2: using mathematical models such as an SST turbulence model, a volume reaction model, an EDC model, a NOx generation model, a SNCR reaction model, and a DPM model for simulation and calculation; step 3: optimizing the SNCR denitrification method of pelletized flue gas with different ammonia nitrogen ratios (NSR) to reduce ammonia slip; step 4: dispersing the flow of a single spray gun to multiple spray guns, reducing the flow of a single spray gun, and improving the mixing efficiency of the reducing agent and the flue gas; step 5: utilizing the rotating flow field of the cyclone separator to improve the mixing uniformity; step 6: designing four spray gun layout schemes, selecting the best NSR and injection source layout scheme for simulation and comparison with actual operating condition data.

2. The method according to claim 1, characterized in that The spray flow dispersion is achieved by arranging the spray guns on both sides of the cyclone separator inlet so that the spray direction is orthogonal to the flue gas flow direction.

3. The method according to claim 1, characterized in that The jet speed is 1 m / s. In the jet source arrangement scheme, the spray gun is arranged on the outlet plane of the cyclone separator, utilizing the rotating flow field of the cyclone separator.

4. The method according to claim 1, characterized in that The process includes at least one cyclone separator, and a plurality of spray guns are arranged inside the cyclone separator according to an optimization scheme.

5. The device according to claim 4, characterized in that The building module guides the optimization method of actually transforming the cyclone separator SNCR denitration system according to the best solution 3, and the best denitration efficiency obtained is 81.2%.

6. A SNCR process system for a thermal power plant using the method or device according to any one of claims 1 to 6.

Images