Research On Devolatilization And Nitrogen Conversion Of Lignite Based On Its Chemical Structure

Devolatilization is the first step of coal utilization, which is very important to the conversion of coal. The process of devolatilization has a great impact on the yields of volatile, distribution of fuel nitrogen, softening and swelling of coal particles, formation of soot and the reactivity of char. It is important to fully understand the pyrolysis of coal, which can guide the thermal process of coal utilization, help to improve the effiency of coal combustion, gasification and liquefaction, help to reduce the generation of pollutants. Untill now, kinetics of pyrolysis have been determined by experiments based on empirical methods. The kinetic parameters provided in different literatures are quite different, which make it difficlut to choose the appropriate kinetic parameters in CFD simulations. In addition, the distribution of nitrogen during pyrolysis has a great impact on NOx formation. An appropriate general model of pyrolysis is required.Effects of the pyrolysis temperature(1000°C, 1100°C and 1200°C) and the residence time on the rapid pyrolysis of four lignites were studied in a drop tube furnace. Solid-phase pyrolysis yields and major gas species were measured. The results show that the yields of volatile increase and the yields of char dec rease with the raise of temperature and residence time. The pyrolysis is completed at 250 ms. The yields of volatile at the end of pyrolysis are higher than the volatile measured by proximate analysis. C/N ratio of mature chars under different temperatures is in a narrow range. HCN is the main product of volatile containing nitrogen when the pyrolysis is finished. One step reaction model of pyrolysis has been established. Conversion paths of fuel nitrogen during pyrolysis together with the distribution of nitrogen and the production of nitrogen precursor gases like HCN and NH3 are obtained.The chemical structural parameters of lignites and chars were detectived by 13C-NMR in order to fully understand the evolution of the chemical structure during pyrolysis. The results show that most of the side chains linked to aliphatic carbons are broken during pyrolysis, decreasing the molecular weight of clusters. However, the aliphatic bridges have not been expelled when the pyrolysis is completed. Chemical composition of functional groups on surface of lignites and chars were measured by XPS. C-C of both aromatic and aliphatic carbon is most abundant in four lignites. Carbon bounds to oxygen by a single bond(ethers, hydroxyls) are most abundant in functional groups of carbon bound to oxygen, followed by two oxygen bonds(carbonyl), and carboxyl or ester functional groups are the least. With the increment of residence time and temperature, the C-C bonds increase, the C-H bonds decrease and the total functional groups of carbon bond to oxygen desrease. N-Q has the weakest thermal stability, followed by N-X. With the extension of the residence time, the fraction of N-5 decreases and the fraction of N-6 increases, making N-6 become the main type of nitrogen functional forms in char. N-6 is more stable than N-5. Total yields of HCN can be calculated based on the loss of N-X, N-5 and N-6. Pore structure of lignites and chars were analyzed by absorption isotherm of nitrogen at low temperature. The specific surface area, pore vo lume and pore diameters of them were calculated. A great number of micropores form during pyrolysis with the release of volatile, while the macropores decrease.CPD model was used to simulate the rapid pyrolysis of four lignites at high temperature, which was validated by the good simulation on the release of volatile. The four studied lignites were added to the database of 28 coals. Ash of coals on dry basis considered as an independent variable was introduced into the 13C-NMR correlation. A new correlation of 13C-NMR parameters based on 32 coals is established and verified. YM and BYH lignites were added into the the coalification digram of 12 reference coals for light gases submodel, expanding the interpolation mesh and the range of O/C and H/C. The results of light gases simulated by the expanded CPD are better than that of CPD. Fraction of nitrogen released to volatile can be pridected well by nitrogen release sub-modle, which can track the decay of ring nitrogen.Combustion of four lignites in an entrained flow combuster with multiple reaction segment(EFCM) were simulated by FLUENT. The results show that the temperature simulated by one step reaction model obtained from experiment is lower than that of one step reaction model by default and CPD model. Variation trend of gases during combustion can be predicted qualitatively by all of the three devolatilization models. Quantitatively, the results of one step reaction model obtained from experiment gets the highest accuracy. The conversion of nitrogen in volatile is considerably important to NO generation for lignite. The result of NO formation by using the NOx conversion path and parameters from experment is better. And NO model adopting fuel nitrogen distribution by CPD with the conversion path of nitrogen in volatile by experiment gives the best results of NO formation.