A New Productivity Prediction Hybrid Model for Multi-Fractured Horizontal Wells in Tight Oil Reservoirs

Productivity analysis relates to the quantitative evaluation of factors for individual members as well as to their comprehensive influence. It is difficult to describe the nonlinear relationship of various factors. A novel and original combination of grey relation analysis (GRA) and fuzzy logic productivity prediction hybrid model is proposed in this study to solve the problem. First, based on the parameters of the constructed wells, several elements were chosen as appraisal factors. The multi-level appraisal system was established to describe the nonlinear relationship of various factors. Then, GRA was used to calculate the weight factor and determine the main factors that influence the result of multi-fractured horizontal wells (MFHWs). Finally, coupling GRA with fuzzy logic to calculate the comprehensive evaluation score (CES), which is used to predict productivity and determine the classification value of the reservoir to evaluate the reservoir quality. The hybrid prediction model has been successfully applied to 18 wells of the tight oil field in Northeast China. Practical application results demonstrated a good agreement between the measured initial production and the output of the hybrid model. The new model can be used to predict production for MFHWs quickly and economically.