Study Of Liquid Loading Characteristics In The Lower Elbow Of Hilly-terrain Wet Gas Pipeline

With the change of the temperature and pressure,the natural gas always drop out liquid in the hilly-terrain wet gas pipeline.If the gas speed is relatively low,the formed liquid phase tends to accumulate in the lower elbow of the pipeline and lead to many problems affecting the normal operation of the pipeline.In addition,due to the presence of the accumulated liquid pool,the originally single gas flow will become gas-liquid two-phase flow.Whether the accumulated liquid can be removed from the elbow by the gas flow depends entirely on the operating conditions of the pipeline.Therefore,research the removal of accumulated liquid in the lower elbows of gas-liquid pipelines by the gas flowing over the liquid pool,and accurately predict the accumulated liquid flow characteristics in the lower elbow of wet gas pipeline is very important for the design and safety operation of the gas pipeline located in undulation terrain.Based on literature study,this thesis analyzed the formation mechanism and flow conditions of accumulated liquid in the wet gas pipeline and indicated that the accumulated liquid in the elbow will result in a special flow in the upward inclined legs with the gas flow rate increases,i.e.zero net liquid flow(ZNLF).Then the ZNLF was investigated as the steady-state two-phase stratified flow and the governing equations was established based on gas and liquid phase momentum equations.The closure correlations of the calculation parameters were determined according to the different gas-liquid interface shape model.Finally,the predicted critical gas velocities were compared with the experimental data.On the basis of theory study and data analysis,this thesis has improved the ARS mechanistic model from two aspects:Firstly,the liquid wetted perimeter correlation was corrected;Secondly,the interfacial friction coefficient was improved.Through the comparison of experimental data and results of the improved model,the results of the improved model in this thesis shows small deviation with experiment data.