针对天然气净化厂脱硫单元能耗较高的问题,对天然气净化厂脱硫单元能耗优化方法进行研究。基于ASPEN HYSYS软件,建立天然气净化厂脱硫单元数值模拟模型,利用现场数据验证模型的准确性。通过单因素分析确定了胺液循环量、贫胺液进料温度及原料气温度对能耗的影响最大,将其作为优化参数建立天然气净化厂脱硫单元能耗最优化模型,采用BP神经网络及GA遗传算法相结合的方法进行优化计算。结果表明,经BP神经网络及GA遗传算法相结合的算法进行优化后,胺液循环量、贫胺液进料温度及原料气温度3个运行参数均得到了优化,将优化后的运行参数运用于现场实际生产后,工艺总能耗由10 614.99 kW下降至8 297.59 kW,能耗降低了21.83%。

Aiming at the high energy consumption problem at desulfurization unit in natural gas purification plant,the corresponding optimization method is studied.Based on ASPEN HYSYS software,the numerical simulation model is established for the desulfurization unit in natural gas purification plant,and the accuracy of the model is verified according to field data.Through single factor analysis,it is determined that amine circulation volume,lean amine feed temperature and feed gas temperature have the greatest impact on energy consumption.Taking these three factors as optimization parameters,the optimization model is established for energy consumption of desulfurization unit in natural gas purification plant,and the method combining BP neural network with GA genetic algorithm is utilized for optimization calculation.Results show that after optimizing by the method,three operation parameters including lean amine circulation volume,lean amine feed temperature and feed gas temperature are all optimized.After the optimized operation parameters are applied to the actual operation,total energy consumption of the process is reduced from 10 614.99 kW to 8 297.59 kW,a reduction of 21.83%.It can be seen that the parameter optimization method based on BP and GA algorithm can effectively help to reduce the energy consumption of desulfurization unit.

尹晓云(1995-),女,博士,在站博士后,主要从事气田节能低碳、天然气净化、油气集输多相流等方面的研究工作,通讯联系人,13541088440,yinxy0122@163.com。