OPTIMIZING HYDRAULIC FRACTURING PARAMETERS USING GENETIC PROGRAMMING

Abstract

The researcher studied the development of new three simple models to predict the hydraulic fracturing parameters using artificial intelligence technique (Genetic Programming). The main three parameters are Fracture Pressure, Fracture Half Length and the Fracture Width. Where the inputs of the first model for calculating the fracture pressure are: current pressure, porosity, permeability, depth, Young’s modulus, Poisson ratio. The inputs for the fracture half-length model are: Fracture Pressure, young’s modulus, permeability, reservoir thickness. The inputs for fracture width model are: Fracture Pressure, Poisson ratio, permeability, reservoir thickness. The models were built using 180 points and verified using Monte Carlo technique. In addition, it was verified against three actual case studies.