For the digital twin completes the mapping and interaction between physical space and digital space which shows great potential for development in the industrial field. In view of the low detection efficiency of natural gas dehydration performance parameters and the inability of online optimization of gas station process parameters, this paper applies digital twin in the chemical industry, and establish an overall framework of the digital twin system of the (triethylene glycol, TEG) dehydration. On the one hand, the geometric model of twin system is constructed by the combination of physical device. On the other hand, the flow model dehydration system technology is established based on the real-time driver of physical data. Finally twin model of dehydration is established by designing virtual-real mapping model and complete the mapping of physical space and digital space which can realize the parallel operation of the physical device and the virtual device. By the proposed digital twin system, real-time prediction of natural gas water dew point and other dehydration performance parameters can be realized. To achieve the aim of low power, combining the optimization algorithm and twin model, the optimization of dehydration process parameters is realized, and the economic efficiency is improved. Keywords：dehydration of TEG; digital twin; prediction of dehydration performance; process optimization.