Modeling the interaction between hydraulic and natural fractures using three dimensional finite element analysis

Abstract

Natural fractures are present in almost every formation and their size and density definitely affect the hydraulic fracturing job. Some of the analysis done in the past shed light on hydraulic fracture (HF) and natural fracture (NF) geometries. The interaction of the HF with existing NF in a formation results in a denser fracture network. The volume of rock covering this fracture network is called the stimulated reservoir volume (SRV). This SRV governs the hydrocarbon production and the ultimate revenue generation. Moreover, past studies show that a microseismic interpreted SRV can be different than the actual SRV. Additionally, there is always limited subsurface access, which makes it imperative to understand the HF – NF interaction to plan and execute a successful hydraulic fracturing job. A three layered, three dimensional complex geomechanical model is built using commercially available finite element analysis (FEA) software. A propagating HF approaching mainly orthogonal NF is studied and analyzed. Cohesive pore pressure elements in FEA software capable of modeling fluid continuity at HF – NF intersection are used to model the HF – NF interaction. Furthermore, a detailed sensitivity analysis considering the effect of stress contrast, job design parameters, NF properties, and properties of the formation is conducted. The sensitivity analysis of properties such as principal horizontal stress contrast, job design parameters, NF properties and properties of target formation reveals a broad variation in the impact of the sensitivity parameters on the HF, NF, and HF-NF geometry and interaction. The observations and the corresponding conclusions were based on broadly classified sensitivity parameters. The most important parameters solely for HF resultant geometry are observed to be a high stress contrast with stress reversal, highest injection rate, and farther NF distance from the injection point. The least important parameter is observed to be the scenario with almost equal horizontal stresses. However, the most important parameter solely for resulting NF geometry is only the high stress contrast with stress reversal. Conversely, for the considered sensitivity cases, the least important parameters are the injection rate, lower injection viscosity (10 cP), higher NF leak-off coefficient, target formation thickness, Young’s modulus, and lowest value of target formation Poisson’s ratio. Collective conclusions for considering HF-NF are also obtained.