| To satisfy the increasing growth of the global energy demand and environmental protection concerns, a number of gas fields are under development such as scattered onshore and offshore gas fields, associated gas from oil fields, coalbed methane from coal mines, which were vented to atmosphere or flared due to the limitation of the technological and economic situations once upon a time. Stranded gas refers to some gas resources remote from markets or pipelines, which are not currently commercially producible for either physical or economic reasons. The recovery of stranded gas can effectively reduce environmental pollution, as well as increase the production capacity of natural gas. As one of the key options for the transportation of stranded gas, the optimization design of the small-scale liquefied natural gas project has important practical significance to promote the market competitiveness of the natural gas liquefaction industry in China.Concentrated on the optimization of the small-scale natural gas liquefaction process and the plate fin heat exchanger, a comprehensive study by the combination of experiments, fundamental theories and numerical simulations is adopted based on fluid mechanics, engineering thermodynamics and heat transfer, as well as the optimization theory. A series of studies are conducted including the evaluation of prediction models for the physical parameters of natural gas, the design and optimization of the small-scale natural gas liquefaction process, the numerical simulation method of offset strip fins, the correlation of flow and heat transfer characteristics, as well as the optimization of heat exchangers with offset strip fins. This study aims to provide theoretical and technological supports for the economic liquefaction recovery of stranded gas.The main research contents and achievements obtained are as follows:(1) On account that the current comparative study of prediction models for the physical parameters of natural gas is limited to a certain parameter or phase state, a comprehensive analysis is conducted among the SRK property package, the PENG-ROB property package, the LK-PLOCK property package and the GERG-2008 property package for the prediction of gas and saturated liquid densities, specific heat capacities, enthalpies, dew points and phase equilibrium parameters based on the experimental results in a wide range of components, temperatures and pressures. Obvious deviations are observed in the calculation of enthalpy and heat capacity for the SRK property package. The deviations of the two basic parameters of gas density and dew point are relatively large in comparison with the experimental results for the PENG-ROB property package. Large deviations are observed in regions near the critical point for the LK-PLOCK property package, which is not suitable for the calculation of saturated liquid density and vapor-liquid equilibrium. As the GERG-2008 property package achieves good accuracy in predicting all of the thermodynamic properties and phase equilibria over a wide range of working conditions and gas compositions, it is recommended as the basis for predicting the physical parameters in the optimization research of natural gas liquefaction projects.(2) In consideration of the requirements of the small-scale natural gas liquefaction project for compactness, safety and economic benefits, an adaptability study is conducted of natural gas liquefaction processes for the application of skid mounted devices and offshore usage. Based on it, a small-scale liquefaction process adopting single nitrogen expansion with carbon dioxide pre-cooling is put up with. ActiveX is employed to establish a hybrid simulation platform connecting Aspen HYSYS with Matlab for data transmission, where GERG-2008 is used for the calculation of physical parameters. The penalty function is adopted for the conversion of constraint conditions. The single objective function with the minimum unit energy consumption and the multi objective functions with the lowest unit energy consumption and maximum liquefaction rate are constructed. The genetic algorithm is used to optimize the process parameters. Results show that this method overcomes the shortcomings of the mathematical programming in convergence, which can be used for the optimization of other processes. The hot composite curve of the precooling heat exchanger matches well with the cold one, and the cold energy of the refrigerants is well utilized, which indicates the high heat transfer efficiency of the process.(3) Based on the optimization results of the single nitrogen expansion process with carbon dioxide pre-cooling, the economic performance of the process is analyzed. Results indicate that the multiobjective optimizations of the main function method and the non-dominated sorting genetic algorithm with elite strategy are similar to each other. The economic benefit of multiobjective optimizations is much higher than that of the single-objective optimization. The natural gas liquefaction project is sensitive to the price of LNG.The initial investment is the key factor which restricts the economic performance of the project.A comparative analysis is conducted between the single nitrogen expansion process with carbon dioxide pre-cooling and other small-scale liquefaction processes including the N2-CH4 expander cycle, the mixed refrigerant cycle with propane precooling, the mixed refrigerant cycle in skid mounted package. Results show that the CO2-N2 process is suitable for the development of small gas reserves, satellite distribution fields of gas or coalbed methane fields with compact device, safety operation, simple capability and low energy consumption.(4) To study the hydrodynamics and heat transfer characteristics of high performance fins, the geometric model of offset strip fins is established and meshed by ANSYS Workbench software, in which the method to control the quality of near wall grids is studied. Fluent is adopted to conduct the numerical simulation. Based on the calculation results of empirical correlations, the fluid constitutive model under different Reynolds numbers is determined in order to ensure the prediction efficiency and accuracy of offset strip fins. Results indicate that the Laminar model should be used when the Reynolds number is in the range of [300,1000], and the SST k-? model should be used when the Reynolds number is in the range of [1000, 10000]. The average relative deviation of Colburn factor between the numerical results and the calculation results of the Manglik&Bergles correlation is within 13%, and that of the friction coefficient is within 8%. The high prediction accuracy verifies the validity and reliability of the numerical simulation method.(5) Due to the fact that the traditional empirical formula is not able to cover the general specifications of domestic offset strip fins, the correlation of flow and heat transfer characteristics is improved based on the numerical simulation and the industrial standard (Aluminum plate-fin heat exchanger NB/T 47006-2009). By coupling the Manglik&Bergles equation with the ALEX equation, a relational expression containing the full range of domestically produced offset strip fins is proposed. The thermodynamic and transfer properties of natural gas and nitrogen are obtained based on the GERG-2008 equation and the REFPROP model. The numerical simulation results show that the proposed relation accurately describes the flow and heat transfer performance of the natural gas cooling process and the nitrogen heating process in serrated fins.(6) Based on the proposed correlation of domestically produced offset strip fins, the optimization method of heat exchangers with the genetic algorithm is studied by adjusting the interactions between the thermodynamic parameters and the geometric parameters. The penalty function is adopted for the conversion of constraint conditions. Compared with the trial and error method, the genetic algorithm avoids repetitive computation, improves the design quality, and shortens the design cycle. The comparative analysis of single objective optimizations and multi-objective optimizations is conducted. The target with the maximum of the effectiveness is equivalent to the minimum of the entropy number generated by heat transfer. The poor thermodynamic performance of the total annual cost is mainly determined by the large proportion of operating costs in the total cost, the optimization process of which is focus on the minimum of the entropy number generated by pressure drop. The modified entropy generation number coordinates the comprehensive performance of the heat exchanger in both economic and thermodynamic aspects. The effective compromise of the single targets is achieved by the non-dominated sorting genetic algorithm with elite strategy. The multi optimization is flexible in providing multiple solutions to the selection of practical production. |
PDF Full Text Request