Prediction Model of Rock Mechanics Parameters in Ultra-Deep Fractured Formations Based on Big Data
The rich oil and gas resources in ultra-deep reservoirs are the focus of oil and gas development at present.However,due to the high fragmentation and strong heterogeneity of rock mass in reservoirs,the traditional method of predicting reservoir mechanics parameters has a large error,which causes great difficulties in engineering design and operation.Based on a large number of experimental and field operation data,the interrelationship among reservoir logging data,reservoir fracture data,and rock mechanics data was analyzed.A multi-parameter constraint was established based on the physical correlation between rock mechanical properties,geological reservoir characteristics and logging interpretation,and a multivariate nonlinear regression fitting model was developed to predict rock mechanics parameters in the whole reservoir section of ultra-deep fractured reservoirs.The prediction model overcame the problem of large calculation errors caused by small amount of fractured formation data and could determine the rock mechanics parameters in the whole reservoir section.Compared with the actual operation parameters,the prediction accuracy is more than 90% .The research results provide technical support for the safe operation of drilling and completion in ultra-deep fractured formations.
rock mechanicsfractured formationsbig data analysisnonlinear fittingdata augmentationprediction model
万方数据
维普
