Abstract
In initial fracturing of tight oil and gas reservoirs, due to the influence of geological and technological factors, the fracture conductivity has decreased, and the single-well productivity has been reduced. It is urgent to repeat transformation to restore or increase productivity. Well selection and layer selection is one of the key factors that affect the design of re-fracturing and the effect of stimulation.
Based on a big database of well-sites, establishing machine intelligence theory determines the elasto-plasticity, permeability, porosity, completion parameters, production decline parameters and skin coefficient that affect the effect of re-fracturing stimulation by dimensionless parameter method of well and layer selection and its stimulation evaluation model.
Combined with artificial neural network and BP algorithm, the index weights of strata with different reservoir physical properties are calculated to analyze the final evaluation value of fracturing effect. On the basis of remaining oil distribution research, scale extended fracture repeated fracturing is increased, injection-production well pattern is improved, scale repeated fracturing effect is increased, well pattern is improved, target layer is repeatedly fractured, and oil increase effect is obvious after fracturing.