Research On Grey Modeling Based On Efficiency Improvement Of Low Permeability Oil Well

There are abundant permeability oil gas in china.Recently,up to 60% proven oil gas reserves is low permeability oil gas.Low permeability oil gas becomes the main body of oil gas reserves.It is of great significance to improve the mining efficiency and reduce the mining cost of the low permeability oil.The yield and mining efficiency of the low permeability oil well is closely related to the geological structure and pumping equipment.Different from middle-high permeability reservoir,the low permeability oil reservoirs has the characteristics of nonlinear seepage law,non-constant permeability and threshold pressure gradient,so,the productivity prediction of low permeability oil well needs to consider the influence of these above factors.The low permeability of oil reservoir makes the low permeability oil wells are prone to fault and difficult to be identified.With the continuous production of oil wells,the oil pump will be inadequate,so intermittent mining is a good choice which needs a reasonable intermittent pumping cycle.In order to improve the productivity and efficiency of the low permeability oil well,this dissertation studied the low permeability oil well from three aspects: productivity prediction,fault recognition and intermittent pumping cycle control.The productivity prediction is the foundation of improving the production of the low permeability wells.When the production capacity decreases significantly,fault recognition need to be done for the low permeability oil well.If the well is recognized as deficient-liquid supply,intermittent production mode will be used for its mining,that through optimizing the pumping cycle to improve productivity and save energy.The main contribution of this dissertation includes the following aspects:Focusing on the capacity prediction of the low permeability oil wells,through analyzing the seepage flow mechanics mechanism of low permeability oil wells,considering the factors of starting pressure gradient,stress sensitivity and fracturing stimulation,we established the oil production grey model(OPGM(1,1)),derived the parameters solution formula of the model,and displayed its solution algorithm.We carried out multiple transformation on OPGM(1,1),and proved that multiple transformation is not affected by the accuracy of OPGM(1,1)model,then showed the transformation formula between transformed parameters and the original parameters of the model.The influence of function transformation on the smoothness and precision of the grey model was studied,and the necessary and sufficient condition for improving the reverting precision of the grey model by function transformation was given and proved.We improved the accuracy of OPGM(1,1)model through function transformation,and used OPGM(1,1)model to predict the yield of low permeability oil well.Aiming at the problem of fault recognition on low permeability oil well,this dissertation determines the fault types by the correlation degree between the measured indicator diagram and the standard fault indicator diagram.Firstly,extract feature vector of the indicator diagram from the three aspects of two dimensional invariant moments,grey statistical characteristics and Fourier descriptors.Then,establish grey relational degree model and grey clustering model under 2-norm,define the grey number synthesis operation rules which follow certain information criteria,and construct the grey relation clustering recognition model.Finally,pre-process the digital feature vectors,which embody the differences between vectors,use the grey relation clustering recognition model to distinguish the indicator diagrams,and show the comprehensive identification results.Aiming at the problem of intermittent pumping cycle control in low permeability oil wells,the intermittent pumping cycle was controlled through determining the well opening time and shutting time.This dissertation analyzed the seepage flow mechanics properties and grey attributes of the pumping oil well.In well opening period,established the well opening seepage flow grey model OWSGM(1,1),and displayed the parameters solution formula and model solution steps.In well shutting period,solved the problem of GM(1,1)power model stability by using the vector transformation method.For low permeability intermittent pumping oil wells,GM(1,1)power model is used to predict well opening time in well shutting period,and the OWSGM(1,1)model is used to predict well shutting time in well opening period.