Research On Integrated Optimization Of Complex Sweetening Process And Sulfur Network In Refinery

With the beginning of the‘14th Five-Year Plan’,refineries have entered a new stage of quality improvement,transformation and upgrading,and high-quality development.As an important part of the refineries,the optimization and finely mangement of the public resource system have become more and more important with the change of the processing plants.This measure is a key point for the refineries to further reduce energy consumption and improve economic benefits during the energy-saving optimization deep water period.In the public resource system,the desulfurization system uses the N-Methyldiethanolamine（MDEA）solution to absorb the acidic components（mainly H₂S and CO₂）in refinery gas,as an indispensable public system for the refinery to protect the normal operation of other systems and plants.With the deterioration of crude oil and the continuous strengthening of environmental protection regulations,the optimization of the system has become more and more important,but it also faces many challenges.Mainly manifested in the clustering of plants and the research and development of high value-added sulfur chemicals,which diversify the source of refinery gas and the destination of acid gas,and put forward new requirements and higher demands for the design of new processes and the synthesis methodology of the process structure of the desulfurization system.Therefore,this article uses process system engineering methodology,aiming at the amine liquid desulfurization system of refineries,based on the unit level（desulfurization process）and network level（desulfurization system）,to carry out the design and optimization of multi-raw gas processes and multi-product processes,as well as synthesis research on large-scale desulfurization systems.The main research and results are as following:（1）In the multi-raw gas absorption process,through the comparative analysis of the single feed absorption and the multiple feeds absorption,it is found that reasonably adjust the differences in the flow rate,composition and feed stage of each feed gas can optimize the temperature and concentration distribution in the absorber and effectively promote H₂S absorption process.Based on this,the control mechanism of the absorption rate of the multiple feeds was revealed,and the improvement structure of the process of pre-allocation and multi-feeding of raw gas was proposed,and the multi-raw gas sweetening process was designed.The case study shows that the proposed process can achieve the dual effects of increasing absorption selectivity and reducing solvent consumption,ultimately the purity of H₂S in the acid gas was increased from 69.3%to 71.8%,and the operating cose and exergy loss of the entire plant were reduced by 6.53%and 18.75%,respectively.The proposed process can achieve simultaneous optimization of quality and energy,and also provide guidance for the process improvement of the two-conponent chemical absorption process.（2）Aiming at the complex and strong nonlinearity mathematical model,and high-fidelity model requirements of the multi-raw gas sweetening process,this article establishes a simulation-based optimization framework,builds a hybrid modeling platform of Aspen HYSYS and MATLAB through Component Object Model（COM）of Active X,and uses particle swarm optimization（PSO）to optimize the raw gas distribution network.In addition,a stepwise adjustment method was proposed to adjust convergence-difficult column for better convergence.The case study shows that the proposed optimization method can simultaneously ensure the accuracy and efficiency of the solution.After optimization,the multi-raw gas sweetening process can increase the purity of H₂S in the acid gas from 69.3%to 72.3%,while reducing the operating cost by 7.07%.This optimization method also provides guidance for the optimization of the same type of absorption process.（3）In the multi-product desorption process,the existing process can only obtain a single purity acid gas.Through the research on the desorption rate control mechanism,a design method using tower pressure and feed temperature as the main and secondary control means,and using multiple desorbers to obtain multiple streams of acid gas with different purity is proposed and a multi-product sweetening process is designed to meet the raw gas requirements of multiple downstream sulfur recovery units.In addition,the economic evaluation method based on the equipment purchase curve and the environmental evaluation method based on the mixed life cycle inventory model were used,and a comprehensive process evaluation system was established,and the process parameters were optimized and analyzed through single-objective and multi-objective optimization methods.The results show that compared with the conventional chemical absorption process,the economic benefit of the proposed process has increased by 141.95%,but the CO₂ emission has increased by 33.10%,which makes it necessary to evaluate the comprehensive performance of the proposed process in practical applications.In addition,the influence of solvent selectivity in the proposed process is more significant,and the improvement of solvent selectivity can realize the simultaneous optimization of economy and environment,which also provides theoretical guidance for the design and operation of the proposed process in industrial applications.（4）Aiming at the structural optimization of the desulfurization system lacks methodological guidance,and still relies on the experience of engineering technicians.According to the characteristics of the system,a superstructure（sulfur network）covering the sulfur source,sulfur sink,and absorption-desorption intermediate processing process was established,and the Kriging interpolation method was used to establish surrogate models of the absorber and the desorber for optimization.In the process of establishing the surrogate model,the input and output variables of each surrogate model were determined by the trade-off between the model complexity and the optimization complexity,and the adaptive sampling algorithm based on sparsity and nearest neighbor expectation（ASSA-SNN）was used to reduce the sampling time.The established surrogate model has the advantages of high accuracy,wide application range,and can effectively reduce optimization complexity.Taking a petrochemical desulfurization system as an example（9 refinery gas,3 downstream sulfur recovery units）,the proposed method is used to optimize the entire system process structure,and on the basis of the optimized structure,the multi-raw gas sweetening process and multi-product sweetening process are used to strengthen the absorption and desorption twice.The operating cost and annual fixed investment cost of the desulfurization system can be reduced by 16.29%and38.92%,respectively,and the economic benefit will be increased by 15.30%.This confirms the feasibility of the method and also provides methodological support for the process structure design of the desulfurization system.