Daily Production Intelligent Prediction Model of Single Well in SHB Oilfield Based on Support Vector Machine

Aiming at the reservoir with natural fractures in SHB oil and gas field, the support vector machine (SVM) and support vector regression method were used to predict the daily production of 18 blow wells in SHB main fault zone. By inputting variables include drilling and completion data, production dynamics, bottomhole pressure, and off-well production time, and taking the predicted output value as output variables, the daily production forecast data of a single well ranging from 120 days to 1200 days was obtained. Compared with the traditional Arps decline prediction method for 100 days of actual production data, SVM is not only more efficient than the traditional DCA analysis method, but also avoids geological modeling and a lot of historical fitting work. It has certain reference value to the formulation of reasonable production system in SHB block.