Research On Intelligent Control Method Of Multi-States Pumping In Low Permeability Wells

At present,most of China’s oil fields have entered the middle and late stages of exploitation,and most of the oil wells are in a low permeability or even ultra-low permeability state,resulting in a mismatch between the working state of the pumping units and the fluid supply capacity of the oil wells,which causes a great waste of electrical energy.According to statistics,the ineffective power consumption of pumping units can be up to 30% of the total power consumption of the pumping system.Therefore,improving pumping efficiency is extremely important for oilfield production.For this reason,this dissertation classifies the well state and type according to the well submergence depth change condition,and then researches the pumping control method.The main research is as follows:(1)Analysis on oil well state.By analyzing the comprehensive model of well submergence depth,the wells are classified into four types: full pumping type,transition type,non-full pumping type and stopping type,and two states: full pumping state and non-full pumping state,and the well types and state transitions are analyzed in detail.The pumping control is analyzed for the state of the oil well,including the control of pumping speed during continuous oil recovery and the adjustment of stopping time during intermittent pumping oil recovery.(2)Research on pumping speed control method.By analyzing the revenue components of a single well during production,combining the relationship between motor power and fluid production when the pumping unit is in operation,and based on the revenue model on pumping unit stroke,the fitted relationship between pumping efficiency and submergence depth is analyzed,and then a relationship model between submergence depth and pumping unit stroke is established for the purpose of pumping rate control.The accuracy of the model is verified based on solving and fitting the unknown parameters in the model.The accuracy of the model is verified based on solving and fitting the unknown parameters in the model.The results of the model tests show that the model can be used to obtain the number of pumping strokes required to obtain the maximum revenue from the well based on the current submergence depth of the well.(3)Research on the method of adjusting the stopping time for intermittent pumping.Firstly,the method of adjusting the stopping time of inter-pumping by using the ratio of the predicted fluid production to the standard fluid production was proposed.Then,on the basis of the theoretical knowledge of the influencing factors of fluid production,PSO optimization algorithm and BP neural network,a PSO-BP fluid production prediction model is established.Finally,the predicted liquid yield of the model is applied to adjust the stopping time of intermittent pumping,and the electricity consumption after the adjustment is analyzed.The analysis results show that the intermittent pumping operation process is optimized by adjusting the stopping time of intermittent pumping.Test and experimental results show that the model of the relationship between submergence depth and pumping stroke and the method of adjusting the stopping time of intermittent pumping proposed in this dissertation can effectively control the pumping speed and determine a reasonable stopping time of intermittent pumping to achieve the purpose of maximizing the revenue.