Prediction of physical properties of deep shale reservoirs and optimization of favorable exploration areas based on machine learning algorithms
To achieve the prediction of the physical properties of deep shale reservoirs and the selection of favorable exploration areas,Longmaxi shale in Luzhou Block of Sichuan Basin was taken as an example.Based on the physical property test of core samples,predicted physical properties of the reservoirs were compared by three machine learning algorithms:Support Vector Regression(SVR),Gradient Boosting Decision Tree(GDBT),and Extreme Gradient Boosting(XGBoost),and the appropriate evaluation model was selected.Then,the physical properties of deep shale reservoirs were predicted,and the favorable exploration areas were selected.The results showed that:1)Among the three popular machine learning algorithms,GDBT was the most suitable algorithm to predict the porosity of deep shale reservoirs in the study area,and XGBoost was the most suitable algorithm to predict the permeability of deep shale reservoirs in the study area;2)Based on the above-selected models,the porosity of Longmaxi shale in Luzhou Block ranges from 2.67％to 9.67％,with an average of 4.88％,and the permeability ranges from 3.22 μD to 28.63 μD,with an average of 11.34 μD;3)Based on the physical properties and gas content of Longmaxi shale reservoir,7 Class Ⅰ favorable areas,3 Class Ⅱ favorable areas,and 4 Class Ⅲ favorable areas were identified.Based on the actual situation of the study area,this achievement can provide a reference for the evaluation of shale reservoirs and the prediction of favorable areas in the study area or similar areas.
万方数据
维普
