[Energy people are watching, click on the top right corner to add 'Follow']
The distribution of components of the combustion process varies greatly from reaction region to reaction area, so there is no need to use the same chemical reaction mechanism throughout the simulation. The dynamic adaptive mechanism simplification algorithm can simplify the detailed mechanism in real time in the local reaction area to an accurate sub-mechanism, so that at each local location of the reaction, only the finite rate reaction integration operation needs to be coupled to the simplified mechanism of the smallest fraction and the number of reactions in the reaction region to achieve computational acceleration. At present, there are numerical simulations of gas and liquid fuel combustion coupled with dynamic adaptive mechanism simplification algorithms, while numerical simulations of pulverized coal combustion based on dynamic adaptive reactions have rarely been reported.
In order to improve the simulation accuracy and computational efficiency of pulverized coal flameless combustion and study the nitrogen conversion mechanism of pulverized coal flameless combustion fuel, Associate Professor Li Pengfei of Huazhong University of Science and Technology first introduced the principle of dynamic adaptive mechanism simplification algorithm, and then coupled the nitrogen-containing skeleton mechanism developed independently and adopted the dynamic adaptive mechanism simplification algorithm to simulate the limited rate of nitrogen conversion of pulverized coal flameless combustion fuel, and after the simulation results were verified by the system, the nitrogen conversion kinetics analysis of pulverized coal flameless combustion fuel was carried out.
summary
Flameless combustion is one of the new high-efficiency clean combustion technologies that have attracted wide attention in recent years, with volumetric low reaction rate combustion zone and typical medium and low temperature combustion characteristics, which need to be coupled with detailed reaction mechanism and consider turbulence and chemical reaction interaction to improve the numerical simulation accuracy of flameless combustion and its NO generation. Based on the dynamic adaptive reaction mechanism, the flameless combustion and NO generation characteristics of pulverized coal were studied in high-fidelity numerical simulation. By using the dynamic adaptive mechanism simplification algorithm, the simulation process locally simplifies the autonomously developed nitrogen-containing skeleton mechanism in real time. The evaluation found that compared with the skeleton mechanism simulation alone, the dynamic adaptive reaction can obtain about 3 times the computational acceleration without sacrificing the calculation accuracy, and the prediction accuracy of no no generation in the furnace is significantly better than that of the traditional NO post-processing simulation method. Based on the experimental simulation results, the furnace distribution of typical nitrogen-containing precursors such as HCN and NH3 was also obtained, and the key nitrogen conversion information such as the nitrogen conversion path, active components and active reactions of pulverized coal flameless combustion fuels were further analyzed. The results show that the no-generation of pulverized coal flameless combustion mainly depends on NH3, HCN and N2O intermediates, and NCO and HNO are the more critical intermediate components. HCN intermediates generate NO mainly through HNCO/CN and NCO pathways. NH3 intermediates are generated by HNCO and further converted into NH2 and HNO, eventually no. The N2O path mainly participates in NO restoration and contributes less to NO generation. CH3CN is also an important intermediate component for generating NO, which can be generated via the NCO path.
1 Numerical simulation method based on dynamic adaptive response
Figure 1 ISAT-DAC calculation process
Fig. 2 IFRF furnace geometry
The research object OFF coal flameless combustion test adopts high-speed direct injection of primary and secondary wind jets, and there is a strong jet diffusion and coiling effect of the high momentum jet, which causes large-scale flue gas circulation in the entire combustion area, and the reaction mixture is diluted by the recycled flue gas and heated to exceed the auto-ignition point, achieving flameless combustion.
Due to the symmetrical nature of the test furnace, only 1/4 of the furnace is simulated to save simulation consumption. Using a three-dimensional hexahedral structured grid, the total number of selected grids is about 600,000 after the grid independence analysis.
This calculation uses the detailed reaction mechanism of nitrogen content and couples the combustion oxidation process with the nitrogen conversion process to simulate the formation of pulverized coal flameless combustion NOx. The author evaluates, develops and simplifies a variety of widely used nitrogen-containing detailed reaction mechanisms, and finds that the PG2018 mechanism has significant advantages in nitrogen conversion simulation accuracy compared with other nitrogen-containing detailed mechanisms. Under the condition of ensuring the simulation accuracy, based on the development and simplification of the PG2018 mechanism, the skeletal reaction mechanism containing only 35 components and 259 steps of the reaction was obtained. In the numerical simulation of this study, the high-precision skeleton reaction mechanism is used and the DAC algorithm is coupled to achieve computational acceleration for application to the flameless combustion simulation of pulverized coal.
Numerical simulations are based on the Fluent platform. The turbulence model adopts the standard k-ε model, and the model coefficient Cε1 is corrected from 1.44 to 1.60 to improve the prediction accuracy of the circular tube jet. The chemical osmotic devocation vocation (CPD) model was used to simulate the volatilization analysis. The radiation transfer equation is solved using the discrete coordinate method (DO), and the gray gas weighted and (WSGG) gas radiation model is introduced, and the spatial change of the total emissivity in the WSGG model is a function of gas composition and temperature. The eddy dissipation conceptual model (EDC) coupled with the PG2018 nitrogen-containing skeleton mechanism (35 components and 259-step reaction) independently developed by the author's team was used to simulate homogeneous combustion and fuel nitrogen conversion, and the ISAT algorithm was used to accelerate the calculation combined with the DAC algorithm. Velocity-pressure coupling uses the SIMPLE algorithm, and equation discrete uses the high-order QUICK format.
Different from the approximate simulation of fuel-based NOx generation by traditional semi-empirical reprocessing method, this paper considers the volatile fraction and coke fuel nitrogen precipitation and combines the fuel nitrogen conversion mechanism (i.e., PG2018 nitrogen-containing skeleton mechanism) to carry out a detailed reaction mechanism simulation of the limited rate of coupled combustion oxidation reaction and fuel nitrogen conversion. The volatile components of pulverized coal are calculated using cpD models and considered as six components: CH4, H2, CO2, CO, NO and HCN. Volatile nitrogen is considered to be released in the form of HCN, and coke nitrogen is released in the form of NO. The coke burn-out model uses a kinetic/diffusion control model that assumes that the reaction rate on the surface of coke is influenced by kinetic or diffusion rate, and the particle size remains unchanged and the density changes during combustion.
2 Simulation results and discussion
<h3>2.1 Flameless combustion simulation and experimental verification of pulverized coal based on dynamic adaptive reaction</h3>
By comparing the simulation results of the coupled ISAT-DAC algorithm with the experimental data, the accuracy and applicability of the ISAT-DAC algorithm in the numerical simulation of pulverized coal flameless combustion can be verified. This simulation and test were compared and verified based on the axial speed, furnace temperature, O2 concentration, CO2 concentration, CO concentration and NO concentration, and flue gas emission data.
Overall, the speed and temperature field simulation results are in good agreement with the test. The speed simulation deviation mainly occurs at 0zx/d=1.2), and the axial velocity should be maintained at the initial velocity of the jet (about 65 m/s), so the test measurement is significantly lower. The prediction results of O2, CO2 and CO component concentrations were generally consistent with the experimental values.
Fig. 3 Comparison of the prediction results of axial speed and temperature in the furnace with the test data
Figure 4 Comparison of O2, CO2 and CO concentrations in the furnace with the test data
The comparison of the noo prediction results of the dynamic adaptive reaction based on the ISAT-DAC algorithm and the conventional post-processing simulation and test data shows that the no prediction result based on the post-processing method is not high, and the NO generation simulation at the section 3 and z≈0.3 m is high, while the NO prediction results of the ISAT-DAC algorithm based on the nitrogen-containing skeleton mechanism coupling are better matched with the experimental values, and the NO generation of each section in the furnace can be better predicted. The NO prediction result based on the nitrogen-containing skeleton mechanism coupled ISAT-DAC algorithm has higher accuracy than the post-processing simulation and can be explained from three aspects: (1) compared with the conventional turnkey reaction mechanism, the skeleton reaction mechanism can provide a more detailed description of the pulverized coal fuel combustion process, and can obtain a higher precision temperature and component distribution field; (2) The nitrogen-containing reaction mechanism has been verified in detail and has higher accuracy, and the post-treatment simulation method is only based on the semi-empirical model, and its accuracy is limited; (3) Compared with the conventional EDM vortex dissipation combustion model based on rapid chemical reaction, this simulation adopts the EDC vortex dissipation concept combustion model, which can consider the turbulent and reaction interaction process for finite rate reaction simulation, and can obtain better prediction results for the finite rate reaction of flameless combustion. Therefore, the ISAT-DAC algorithm can improve the accuracy of no no simulation in the furnace compared to the post-processing scheme.
Fig. 5 Prediction results of NO concentration in furnace compared with test data
Finally, comparing the furnace flue gas outlet test data and the prediction results, the relative error between the smoke exhaust temperature, flue gas CO2, O2, CO and NO and the test value is within 5%.
Based on the above results, the prediction results of each monitoring surface and furnace outlet in this simulated furnace are in good agreement with the test data, and the simulation adopts PG2018 skeleton mechanism and coupled ISAT-DAC algorithm suitable for pulverized coal flameless combustion simulation, and the simulation accuracy of fuel nitrogen conversion is improved compared with conventional NO post-processing simulation method.
<h3>2.2 Nitrogen conversion analysis of pulverized coal flameless combustion fuel</h3>
Based on the simulation results that have been verified by the test, the nitrogen conversion mechanism of pulverized coal flameless combustion fuel is further analyzed. The temperature is closely related to the fuel nitrogen conversion process, and the flameless combustion process of pulverized coal has a relatively uniform temperature distribution in the furnace. Further observations showed that there were two main reaction zones in the furnace, one was formed by the fuel being injected downstream and mixed with oxygen, and the other was located at the pulverized coal nozzle and generated by the return of high-temperature flue gas.
Fig. 6 Temperature distribution in the furnace
The nitrogen-containing detailed (skeleton) mechanism simulation method can accurately predict the NO distribution in the furnace, and the NO distribution cloud map based on the nitrogen-containing skeleton mechanism ISAT-DAC simulation and the general package mechanism post-processing simulation shows that the high concentration of NO is distributed in the high temperature region around the powder tube, although there is a high temperature zone downstream of the furnace, the strong reduction reaction between NO and hydrocarbon fuel reduces the NO concentration. The high concentration of NO is distributed at the outlet of the powder pipe, and the estimated no concentration downstream of the furnace is low, which deviates from the actual NO distribution and is not suitable for accurate quantitative analysis.
Fig. 7 No distribution in the furnace
Based on the finite rate simulation of the NITROGEN-containing skeleton mechanism ISAT-DAC algorithm, the distribution of typical nitrogen-containing intermediate components contained in the mechanism in the furnace of the pulverized coal flameless combustion process can be obtained. The HCN content in the furnace is higher, the amount of HCN converted to NH3 is less, the NH3 content is lower and the peak is only about 15×10-6, and the amount of N2O is lower. HCN and NH3 are mainly distributed at the outlet of the powder pipe, which are separated from the pulverized coal, and N2O is distributed around the high temperature zone and participates in the generation and reduction of NO.
Fig. 8 Distribution of intermediate components of HCN, NH3 and N2O in the furnace
The finite rate simulation based on the nitrogen-containing skeleton mechanism ISAT-DAC algorithm can not only predict the generation amount and in-furnace distribution of different nitrogen-containing components, but also analyze the nitrogen conversion path of fuel during the flameless combustion of pulverized coal. The nitrogen conversion path of pulverized coal flameless combustion fuel shows that thermodynamic NO generation is significantly inhibited, and the fuel-based NO generation mainly depends on HCN, NH3 and N2O intermediates, and NCO and HNO are the more critical intermediate components. HCN intermediates mainly generate NO through HNCO/CN and NCO pathways; NH3 intermediates are generated by HNCO and further converted to NH2, HNO, and finally no; N2O pathways mainly participate in NO reduction and contribute less to NO generation. Reaction path analysis showed that CH3CN is also an important intermediate component for no generation, which can be generated through the NCO path.
Figure 9 Nitrogen conversion path of pulverized coal flameless combustion fuel
Due to the coupling of dynamic adaptive mechanism simplification method (DAC) and local adaptive table building method (ISAT) to achieve computational acceleration during simulation, a cloud map of active components in the furnace can be obtained after the dynamic adaptive mechanism is simplified. That is, the ISAT-DAC simplification method can be used to accurately identify the main reaction area in the furnace during the simulation process, and the simplified main reaction area retains only 32 components at most, and the other regions have a score of 0 (no reaction area), thereby saving calculation costs. The main reaction area is located downstream of the powder feed tube, and the pulverized coal meets the high temperature secondary wind and the combustion reaction occurs after the powder feed tube is ejected, and the main reaction area corresponds to the high temperature zone. The active components above the furnace chamber are mainly due to the return of high-temperature flue gases.
Fig. 10 Distribution of active components in the furnace
When the dynamic adaptive mechanism simplification method removes the remaining components that contribute less to the target component, the related reaction containing the component is also removed, so as to obtain a cloud map of the distribution of active reactions in the furnace. The region with more active reactions corresponds to the region with more active components, and the maximum number of reactions in the main reaction area is retained, the weak reaction area is reduced to 80 steps in turn, and the number of flue gas regional reactions is 0 (that is, no reaction occurs).
Fig. 11 Distribution of active reactions in the furnace
Statistical calculation time found that compared with the combustion simulation based on the skeleton reaction mechanism (35 components and 259-step reaction) coupled WITH algorithm, the simulation can be further combined with the DAC algorithm to obtain about 3 times the computational acceleration. Because of this skeleton reaction mechanism (35 components and 259-step reaction), compared with the original PG2018 detailed reaction mechanism (151 components and 1 397-step reaction), the acceleration effect of about 18.6 times can be obtained. That is, the skeleton reaction mechanism coupled ISAT-DAC algorithm can achieve nearly 55 times of computational acceleration compared with the original PG2018 detailed reaction mechanism, and the calculation accuracy is not sacrificed.
3 Conclusion
1) Based on the IFRF 0.58 MW combustion furnace, the finite rate simulation of the nitrogen-containing skeleton mechanism of pulverized coal flameless combustion is carried out, the turbulence-chemical-nitrogen conversion reaction interaction during combustion is considered, and the dynamic adaptive reaction mechanism simplification method is further coupled to achieve computational acceleration. The prediction results of temperature, speed, O2 concentration, CO2 concentration and NO concentration of the furnace monitoring surface and furnace outlet in the furnace were well matched with the test data, which verified the accuracy and applicability of the PG2018 skeleton mechanism and ISAT-DAC algorithm in the flameless combustion simulation of pulverized coal. And the coupling DAC algorithm can achieve nearly 3 times the computational acceleration compared with the skeleton mechanism simulation, and the detailed reaction mechanism can achieve nearly 55 times the acceleration effect.
2) The study of the nitrogen conversion characteristics and nitrogen-containing key intermediate components of pulverized coal flameless combustion fuels shows that NO generation mainly depends on HCN, NH3 and N2O intermediates, and NCO and HNO are the more critical intermediate components. HCN intermediates mainly generate NO through HNCO/CN and NCO pathways; NH3 intermediates are generated by HNCO and further converted to NH2, HNO, and finally no; N2O pathways mainly participate in NO reduction and contribute less to NO generation. Reaction path analysis also showed that CH3CN is also an important intermediate component in the generation of NO, which can be generated through the NCO path.
3) Based on the experimental and verified nitrogen conversion simulation results of pulverized coal flameless combustion fuel, the distribution of key intermediate components (HCN and NH3) and the distribution of active components and active reactions of nitrogen migration and conversion of fuel in the furnace were obtained for the first time, which can provide a reference for the development of NO emission reduction technology, such as improving the effect of NO rekindling and selective non-catalytic reduction (SNCR) in the furnace according to the temperature and component distribution in the furnace.
Reference format
LIU Lu,LI Pengfei,CHENG Pengfei,et al. Study on nitrogen conversion mechanism of flameless combustion fuel based on dynamic adaptive reaction[J].Clean Coal Technology,2021,27(4):123-131.
LIU Lu,LI Pengfei,CHENG Pengfei,et al. Study on fuel nitrogen conversion mechanism in flameless combustion of pulverized coal based on dynamic adaptive chemistry[J]. Clean Coal Technology,2021,27(4):123-131.
Disclaimer: The above content is reproduced from Clean Coal Technology, and the content sent does not represent the position of this platform.
National Energy Information Platform Tel: 010-65367702, Email: [email protected], Address: People's Daily, No. 2 Jintai West Road, Chaoyang District, Beijing