Research On Optimizing Wells Or Layers And Productivity Predication For Fracturing In ChaoYanggou Oilfield

There are some problems in the process of selecting fracture oil wells and layers andproductivity predication, such as subjectivity, inefficiency etc. Therefore starting from theactual situation of oilfield, this paper deeply studied on methods of select fracture wells andlayers and productivity predication by using fuzzy theory mathematics and neural networks inChaoYanggou Oilfield.On the basis of scientific mathematical statistical method, data collected fracturingWells from the scene, and through the finishing operation got nine possible factorsinfluencing the fracturing effect. Using single factor analysis method and gray correlationanalysis method, determine the main factors influencing the fracturing effect, including theeffective thickness, the moisture content before pressure, connecting direction number, initialproduction pressure and recovery degree. Using the least square method, determine thestandard selecting well of the major factors which Impact on the increased amount of oil afterfracturing, then, using the AHP, determine the weight of each main factor, finally, using thefuzzy mathematics method, determine Comprehensive standards of fracturing well chosenwell. Application of fuzzy evaluation method to establish the optimum fracturing Wells layermodel, firstly make the potential evaluation of fractured Wells and the determination ofpotential layers, once moer make the actual analysis fracturing oil well of three blocks, finallyfracturing selection results are obtained, the example’s results proved that this model werepractical. Application of multivariate linear regression analysis and BP neural network methodproductivity, prediction model is respectively established, it can predicate, Compared with theproductivity of actual fracture, make the error analysis, the results proved that the BP neuralnetwork method is more effective in predicting the productivity after fracturing. Acdroding tothe field data, using mathematical software MATLAB, establish appropriate production afterfracturing forecasting software of neural network structure, and analyze production afterfracturing in three blocks for fracturing, then good results have been achieved.Through the analysis and research in this paper, we can enhance the understanding of thegeological characteristics of the oilfield, achieving the goals of improving the level of theoilfield reservoir subdivision; Improve the design by reasonable fracturing, then achieve theoptimizing fracturing construction, reduce the proportion of inefficient Wells, it has guidingsignificance to increase the degree of use of reservoir remaining reserves and the effect afterfracturing.