Author : Shike Zhang
Publisher :
ISBN 13 : 9781303634079
Total Pages : 163 pages
Book Rating : 4.6/5 (34 download)
Book Synopsis Petroleum Geomechanics Characterization Using Coupled Numerical Modeling and Soft Computing by : Shike Zhang
Download or read book Petroleum Geomechanics Characterization Using Coupled Numerical Modeling and Soft Computing written by Shike Zhang and published by . This book was released on 2013 with total page 163 pages. Available in PDF, EPUB and Kindle. Book excerpt: Petroleum geomechanics characterization is very important for conducting successful geomechanics operations such as reservoir compaction, associated surface subsidence, wellbore stability, hydraulic fracturing and coupled geomechanics-reservoir simulation in petroleum engineering. However, there are two big challenges which restrict accurate petroleum geomechanics characterization. The first challenge is deep subsurface formation. Restricted access to subsurface formation makes this characterization challenging. Consider oil, gas and geothermal reservoirs both conventional and unconventional which are more than the average depth of 200 m, making it very difficult for engineers and geoscientists to accurately obtain the geomechanical parameters with the traditional methods such as core analysis, field test methods and observation of discontinuity states. The other challenge is that there is usually a nonlinear relationship between the geomechanical parameters and the geomechanical behavior of rock mass. Because the traditional methods are mainly based on the linear theory, this remains a challenge and desires that we seek to develop an effective and reliable linear and/or nonlinear regression procedure to overcome the complex spatial relationship for petroleum geomechanics characterization. Inverse analysis can help overcome these challenges by taking advantage of field-observed and recorded information. Thus, inverse analysis has been receiving increasing attention in various fields of science and engineering, and can be defined as the one in which model parameters are obtained from the given observed data. In this dissertation, a displacement and pressure based inverse analysis method using coupled numerical modeling and soft computing is presented in more detail for petroleum geomechanics characterization in oil, gas and geothermal reservoirs. It can effectively overcome several disadvantages of traditional methods in estimating the geomechanical parameters, for example sample disturbance problem encountered in laboratory testing of geomechanical parameters, high cost to obtain core from the in-situ formation in the deep subsurface, and lack of effective tools to measure the fracture properties directly. In this integrated inverse analysis method, a forward modeling is used to (1) create the training and testing samples for artificial neural network (ANN) model, and (2) verify if the identification results are accurate by using back substitution calculation. The ANN model is used as an alternative to the numerical modeling to establish the objective function for optimization search of genetic algorithm (GA). And GA is used as an optimization and search tool to characterize the geomechanical parameters based on the objective function established by a combination of the ANN-predicted values and the field-observed information. In addition, some applications are presented in detail in this dissertation to illustrate the proposed methodology can effectively conduct petroleum geomechanics characterization by utilizing the available field-observed and recorded information in petroleum engineering. For instance, with measured ground surface movements, petroleum goemechanics properties such as Young's modulus, Poisson's ratio, internal friction angle and cohesion, can be characterized; with measured wellbore deformation during drilling, petroleum geomechanics properties such as the maximum and minimum horizontal earth stresses and the spacing, aperture, orientation stiffness of natural fracture, can be characterized in the oil, gas and geothermal reservoirs; with wellbore pumping pressures during hydraulic fracturing tests, petroleum geomechanics such as horizontal earth stress state, elastic parameters and fracture properties can be characterized in both conventional and unconventional reservoirs. This work also demonstrates that an inverse analysis technique based on the integration of coupled numerical modeling, artificial neural network and genetic algorithm combination is an effective and robust method to characterize natural fracture stiffness and spacing that there is currently no effective method of estimating in petroleum engineering.
