Determination of in situ stresses and elastic parameters from hydraulic fracturing tests by geomechanics modeling and soft computing

Hydraulic fracturing is the most effective technology for determination of the minimum horizontal in situ stress, σh, in a rock formation. The maximum horizontal in situ stress, σH, is often determined by the minimum horizontal in situ stress, breakdown pressure, Young׳s modulus E and Poisson׳s ratio v with elastic rock behavior assumed. In this paper, a pressure back–analysis method is proposed for determination of these parameters (e.g., σH, σh, E, v) based on borehole pressures monitored in a hydraulic fracturing test. In the proposed method, an artificial neural network (ANN) is used to represent the relationship between maximum and minimum horizontal in situ stresses, elastic parameters and borehole pressure values; a forward model is applied to perform 2-D numerical simulation of a hydraulic fracturing process to create necessary training and testing samples for the ANN model; the genetic algorithm (GA) is employed to search the set of unknown in situ stresses and elastic parameters in a global space based on appropriate fitness function. A hypothetical numerical experiment is conducted in detail to validate the new method. Results show that the proposed pressure back-analysis method using ANN-GA can effectively determine maximum and minimum horizontal in situ stresses and elastic parameters from borehole pressure values in hydraulic fracturing tests.