A guideline on optimizing fracture modeling for fractured reservoir simulation

Natural fractures are widely observed in unconventional reservoirs and their presence could lead to complex fracture network and a higher production rate by intersection with hydraulic fracture treatments. Focusing on various fractures categories, the objective of this paper is to build and verify various fracture modeling corresponding to different fracture categories depending on their effects on production and pressure, which in return provides the most effective fracture modeling approach for unconventional fractured reservoirs. A micro-scale model was firstly built to simulate fluid flow in fractured porous mediums. Equivalent permeability is computed and compared on some synthetic numerical cases, where different sensitivity parameters including isolated fracture and connected fracture are considered. Next, a reservoir-scale model including stimulated reservoir volume (SRV) and the un-fractured matrix was established. The quantitative analysis focusing on fracture aperture, matrix permeability, and the connectivity of fractures in & outside of the SRV, are conducted to evaluate the effects of different fracture categories on the production rate of fracture models. The micro model reveals that equivalent permeability is highly sensitive to whether the fracture is directly or indirectly connected to the outlet boundary. Based on the reservoir model, a production comparison chart including three categories is given, where fracture aperture varies from 100nm to 1000μm and matrix permeability varies from 100nD to 0.1mD. If the fracture aperture is narrower than 1μm and the matrix permeability is between 0.01mD and 0.1mD, or the fracture aperture is narrower than 10μm and the matrix permeability is larger than 0.1mD, the effect of isolated and connected fractures could be described simply by analytical apparent permeability model instead of the numerical models which characterize fractures explicitly, because the production error is lower than 5.00%. However, if the fractures aperture is larger than 100μm, both the connected fractures and isolated fractures must be explicitly considered to analysis the contributions. As for the reservoirs with SRV, when the fracture aperture is between 100nm and 1000μm and matrix permeability varies from 100nD to 0.1mD, we only need to explicitly handle the fractures directly or indirectly connected to the wellbore, because the disparity caused by fractures isolated from wellbore inside & outside of the SRV is less than 5.00%. This paper quantitatively classifies fractures into several categories and it evaluates the influence of isolated and connected fractures on production and pressure of fracture models, where isolated natural fractures are particularly analyzed. The results provide valuable guidance to optimize fracture modeling for unconventional fractured reservoirs.