Abstract

Prediction of permeability from well logs in heterogeneous formation is a difficult and complex problem to solve by conventional statistical methods. The correlation between parameters like porosity with permeability for different we lls alone may be a crude approximation for permeability estimation, even for homogeneous formations. Recently artificial neural networks have been successfully used for solving many complex problems in reservoir permeability estimation. In this work, the neural network technique is utilized for the permeability estimation of coal formations. The back propagation neural network (BPNN) permeability prediction model has been developed from a data set consisting of well test permeability and well log data from