三肇凹陷A区块葡萄花油层缝网压裂参数优化实践OA

Putaohua reservoir in Sanzhao Sag of Songliao Basin is a typical low-porosity and low-permeability reser-voir.With the increase of repeated fracturing,fracturing effect has been declining annually.In order to explore the main controlling factors influencing the fracturing effect of fracture network stimulation in Putaohua reservoir of block A,various data parameters of the pilot block are preprocessed by clustering analysis method.Stacking ensem-ble algorithm is selected to evaluate the fracturing effect,developing an optimization chart for fracturing parameters.The results show that,the application of clustering analysis for converting discrete data into 2-4 types of categorical variables ensures the correlation coefficient of regression algorithm test set up to>83%.Comprehensively consider-ing the prediction results of different algorithms,the prediction accuracy can be increased by 5 percentage points based on ensemble algorithm.The optimal construction parameters for pilot wells in block A of Sanzhao Sag vary sig-nificantly depending on different reservoir characteristics.Based on different reservoir characteristics,the weights of influencing factors are determined to select 9 main controlling factors with greater weight including effective thickness and sand fracturing intensity,and charts for sand and fluid fracturing parameters optimization are estab-lished.The practical application shows that the initial daily oil production of 20 wells in the pilot block increased by 30%year-on-year.The research provides theoretical basis and scheme for well selection,layer selection and fractur-ing scale design in similar reservoirs.