Optimum Fracture Conductivity for Naturally Fractured Shale and Tight Reservoirs

Abstract

Summary With the development of unconventional shale and tight reservoirs, stimulation treatments that place multiple transverse fractures have received greater attention in recent years. The post-fracture productivity of such low-permeability reservoirs is largely determined by the matrix/fracture contact area with appropriate fracture conductivity. Although it is often anticipated that the fractures are infinitely conductive, the general belief is that production increases with the proppant amount injected. This paper presents an approach to assess the proppant amount injected by determining the optimum post-fracture conductivity. First, through use of 3D finite-difference reservoir simulations in a naturally fractured reservoir, which has both the hydraulic fracture and natural fractures modeled explicitly as discrete gridblocks, we find cumulative production as a function of fracture conductivity. For a fixed propped length and production time, we observe a critical conductivity beyond which the production is insensitive to the conductivity. The critical conductivity is then obtained as a function of the propped length and production time. The numerical results show that the critical conductivity increases with propped length and decreases with production time. The effect of stimulated natural-fracture properties (spacing and permeability) on the critical conductivity is then investigated. For reservoirs with matrix permeability in the range of 20 to 1,000 nd, natural fractures increase the short-term critical conductivity, but decrease the medium- to long-term conductivity. The paper also evaluates the influence of water production, cluster spacing, and flowing bottomhole pressure (BHP) on the critical conductivity. This study demonstrates that fracture designs that are based on pseudosteady-state solutions are not appropriate for naturally fractured shale reservoirs and can lead to significantly lower initial production. Considering conductivity degradation over time, fracture designs that target achieving 1-year critical conductivity are recommended. A simple, yet robust workflow that is based on knowledge of 1-year critical conductivity is also presented for systematically selecting the type and amount of proppant for stimulation treatment. Such a workflow can mitigate trial-and-error-based and data-driven approaches in the industry. An example is demonstrated for the Marcellus play.