Study On The Main Control Factors Of Volume Fracturing Productivity Of Horizontal Wells In Block Ma X

Xinjiang oilfield is the key area of oil and gas exploration and development in China.Mahu oil region belongs to conglomerate oil reservoir.After several years exploration and development practice,a set of high-efficiency development technology has been preliminarily formed and good development results have been achieved.But the Heterogeneity of conglomerate formation is strong,and the difference between blocks and wells is great.There are many factors that affect the effect of volume fracturing,the main controlling factors are not clear,and the evaluation method of fracturing effect has not been established.Based on the geological,engineering and production data of block Max in Mahu Oilfield,this paper firstly applies the statistical method to analyze qualitatively the influence of each factor on production,the main controlling factors of productivity are analyzed by grey correlation method and random forest method,and the results show that the main controlling factors of geology are average oil saturation,porosity and the thickness of oil layer encountered by drilling,the main control factors of the project are length of horizontal section,sand ratio,total sand quantity,ratio of preliquid and total liquid quantity.On this basis,the daily production of oil wells in the target area of Mahu oilfield is predicted by using random forest method,BP neural network method and Support vector machine method respectively,the prediction accuracy of the Support vector machine method is 94.04%,that of the random forest method is 86.37%,and that of the BP neural network is 85.64%.The cumulative production of 60 days,90 days and 120 days can be predicted by using the Support vector machine method,and the prediction accuracy also meets the requirements.Therefore,with less data,the Support vector machine method is suitable for the prediction of production and the optimization of fracture design in the studied block.