A comprehensive approach to select completion and fracturing fluid in shale gas reservoirs using the artificial neural network

Abstract

For the development of shale gas reservoirs, hydraulic fracturing is essential to create fractures and keep the fracture network open to help gas escape from rock matrix with low permeability. However, the extremely low permeability of shale gas reservoirs requires complex and time-consuming analysis to select optimum completion and fracturing fluids. This paper presents a comprehensive approach based on the artificial neural network (ANN) to determine completion methods and fracturing fluids for shale gas reservoirs. To develop the ANN system, the relationship between reservoir parameters and hydraulic fracturing was investigated, and the results were used to obtain ranges of key reservoir properties. Based on the learning data set generated from the categorized ranges, the optimum ANN architecture was designed by adjusting the ANN design parameters, such as training algorithms, and the number of hidden layers and hidden layer neurons. The developed system was also converted to a graphical user interface, to make it more practical for users to access the system. The system was validated by comparing the result values of the system with the desired values, and this revealed that the system showed a high accuracy of correlation factor of over 0.9. Field application for the system was also conducted by using field data and showed that the result values significantly matched with the targeted values. Therefore, the selection system can be an effective tool to determine the optimum completion methods and fracturing fluids in accordance with the reservoir characteristics.