ESTABLISHMENT OF WELL LOGGING INTERPRETATION MODEL IN QINGXI OILFIELD

Qingxi oilfield is an important block of exploration and development in Yumen oilfield. As the reservoirs are complex in lithology, their shale content is high and the reservoir spaces are dominated by fractures, accurate identification of lithology and layer division are difficult with conventional well logging data and reservoir parameters cannot be accurately determined, leading to erroneous logging interpretation results. Moreover, the previous well logging interpretation methods are also not applicable. Along with the application of new well logging technologies, such as image logging and array sonic logging, various fracture identification methods, which are mainly based on image logging data and assisted by conventional logging data, have been developed in Qingxi oilfield in recent years. Array sonic energy attenuation data are used to identify the effectiveness and development of fractures. Neural network technique is used to process various reservoir parameters. In view of the two main reservoir lithologies in Qingxi oilfield, a logging interpretation model is established to discriminate the properties of reservoir fluids by using different cross plots and through integration of the responses of mud resistivity, well temperature logging and other conventional logging data with composite geologic logging and regional OWC. Application of these technologies has significantly improved the coincidence rate of well logging interpretation in Qingxi oilfield, and made great contributions to geologic engineering and study of well logging adaptiveness.