A novel approach to model enhanced coal bed methane recovery with discrete fracture characterizations in a geochemical simulator

• A reservoir simulator has been extended to model CO2-ECBM by ‘chemical reaction’ approach.
• A realistic sample showed our shape factor calculation with discrete-fracture model.
• The discrete fracture approach has also been applied for CO2-ECBM at field scale.
• Results indicated mass transfer between fluid phases can affect methane recovery.
• Application to a complex fractured reservoir demonstrated our CO2-ECBM model capability.

Coal bed methane (CBM) has received increasing attentions as a significant energy resource. Numerical modeling of the CBM recovery processes entails simulation of the complex coupled mechanisms, including desorption of methane from the coal matrix surface, diffusion of gas to coal cleats, and gas flow from coal cleats to the wellbore. Beyond these complex flowing mechanisms, it is crucially important to represent coal cleats, structural fractures, and/or hydraulic fractures realistically in CBM simulators as they provide flow channels and dominate CBM flow behaviors. Existing CBM simulators are typically extended from oil and gas reservoir simulators with either black-oil or compositional formulation, and sorption and desorption are usually modeled by the Langmuir isotherms. The concept of shape factor is commonly used to characterize the flow between matrix and cleats (or fractures). When the shape factor is treated as only a function of cleat spacing, the detailed characteristics of actual cleats (or fractures) are missing, including the spatial distribution of cleats and interconnectivity of cleats. In this study, we propose a new workflow to perform a 2-D coal bed methane recovery simulation with discrete fracture model (DFM) in consideration of both structural fractures (large-scale fractures) and cleats (small-scale fractures). There are two key steps in our approach. First, we use a detailed network of discrete fractures characterized from core samples to represent the actual distribution of identified cleats, and calculate the shape factor of the realistic cleated coal sample by running a flow simulation to pseudo-steady state. Second, we apply the shape factor to field-scale simulations in which large-scale fractures are modeled as DFM. For this purpose, we treat gas sorption and desorption as a “chemical reaction”, and we developed an extension to an existing geochemical-reservoir simulator. We implemented both a “pseudo-compositional” and a “full-compositional” module to study the effect of mass exchange between gas and aqueous phases. We validated our new formulation and simulator development from a benchmark case in which our simulation results show close agreement with commercial simulators. We also demonstrated the significance of modeling mass exchange between fluid phases on CBM recovery in some cases, which is commonly missing in most commercial simulators. Finally, we presented our workflow in modeling an enhanced CO2-ECBM recovery process to a complex fractured coal bed with both large-scale tectonic fractures and small-scale cleats.