Developments in modeling and optimization of production in unconventional oil and gas reservoirs
MetadataShow full item record
The development of unconventional resources such as shale gas and tight oil exploded in recent years due to two key enabling technologies of horizontal drilling and multi-stage fracturing. In reality, complex hydraulic fracture geometry is often generated. However, an efficient model to simulate shale gas or tight oil production from complex non-planar fractures with varying fracture width along fracture length is still lacking in the petroleum industry. In addition, the pore size distributions for shale gas reservoirs and conventional gas reservoirs are quite different. The diffusivity equation of conventional gas reservoirs is not adequate to describe gas flow in shale reservoirs. Hence, a new diffusivity equation including the important transport mechanisms such as gas slippage, gas diffusion, and gas desorption is required to model gas flow in shale reservoirs. Furthermore, there are high cost and large uncertainty in the development of shale gas and tight oil reservoirs because of many uncertain reservoir properties and fracture parameters. Therefore, an efficient and practical approach to perform sensitivity studies, history matching, and economic optimization for the development of shale gas and tight oil reservoirs is clearly desirable. For tight oil reservoirs, the primary oil recovery factor is very low and substantial volumes of oil still remain in place. Hence, it is important to investigate the potential of CO₂ injection for enhanced oil recovery, which is a new subject and not well understood in tight oil reservoirs. In this research, an efficient semi-analytical model was developed by dividing fractures into several segments to approximately represent the complex non-planar fractures. It combines an analytical solution for the diffusivity equation about fluid flow in shale and a numerical solution for fluid flow in fractures. For shale gas reservoirs, the diffusivity equation of conventional gas reservoirs was modified to consider the important flow mechanisms such as gas slippage, gas diffusion, and gas desorption. The key effects of non-Darcy flow and stress-dependent fracture conductivity were included in the model. We verified this model against a numerical reservoir simulator for both rectangular fractures and planar fracture with varying width. The well performance and transient flow regime analysis between single rectangular fracture, single planar fracture with varying width, and single curving non-planar fracture were compared and investigated. A well from Marcellus shale was analyzed by combining non-planar fractures, which were generated from a three-dimensional fracture propagation model developed by Wu and Olson (2014a), and the semi-analytical model. Contributions to gas recovery from each gas flow mechanism were analyzed. The key finding is that modeling gas flow from non-planar fractures as well as modeling the important flow mechanisms in shale gas reservoirs is significant. This work, for the first time, combines the complex non-planar fracture geometry with varying width and all the important gas flow mechanisms to efficiently analyze field production data from Marcellus shale. We analyzed several core measurements for methane adsorption from some area in Marcellus shale and found that the gas desorption behaviors of this case study deviate from the Langmuir isotherm, but obey the BET (Brunauer, Emmett and Teller) isotherm. To the best of our knowledge, such behavior has not been presented in the literature for shale gas reservoirs to behave like multilayer adsorption. The effect of different gas desorption models on calculation of original gas in place and gas recovery prediction was compared and analyzed. We developed an integrated reservoir simulation framework to perform sensitivity analysis, history matching, and economic optimization for shale gas and tight oil reservoirs by integrating several numerical reservoir simulators, the semi-analytical model, an economic model, two statistical methods, namely, Design of Experiment and Response Surface Methodology. Furthermore, an integrated simulation platform for unconventional reservoirs (ISPUR) was developed to generate multiple input files and choose a simulator to run the files more easily and more efficiently. The fracture cost was analyzed based on four different fracture designs in Marcellus shale. The applications of this framework to optimize fracture treatment design in Marcellus shale and optimize multiple well placement in Bakken tight oil reservoir were performed. This framework is effective and efficient for hydraulic fracture treatment design and production scheme optimization for single well and multiple wells in shale gas and tight oil reservoirs. We built a numerical reservoir model to simulate CO₂ injection using a huff-n-puff process with typical reservoir and fluid properties from the Bakken formation by considering the effect of CO₂ molecular diffusion. The simulation results show that the CO₂ molecular diffusion is an important physical mechanism for improving oil recovery in tight oil reservoirs. In addition, the tight oil reservoirs with lower permeability, longer fracture half-length, and more heterogeneity are more favorable for the CO₂ huff-n-puff process. This work can provide a better understanding of the key parameters affecting the effectiveness of CO₂ huff-n-puff in the tight oil reservoirs.