Modeling tracer flowback in tight oil reservoirs with complex fracture networks

Abstract Tight oil reservoirs are commonly developed by a combination of horizontal wells and multistage hydraulic fracturing techniques. The characterization of complex fracture networks after fracturing treatments is a common challenge for unconventional reservoirs. Tracer flowback profiles can reveal a fracture network at an individual stage. However, the numerical simulation of tracer transport in fractured reservoirs usually employs a dual-porosity continuum medium model or describes fractures by an enlarged fracture aperture in a regular grid system that is not capable of modeling a complex fracture network. Especially, the influences of the properties of fracturing fluid and variations in fracture permeability caused by pressure changes on tracer flowback profiles are not investigated. In this paper, a two-phase (water and oil), four-component (oil, water, fracturing fluid and tracer) numerical simulator is developed based on a discrete fracture model for modeling tracer injection and flowback in tight oil reservoirs. With respect to fracture network modeling, the simulator is capable of explicitly modeling an arbitrary distribution of fracture networks with a real fracture aperture by triangular grids. With respect to mechanism consideration, the simulator considers the influences of fracturing fluid and pressure-dependent fracture permeability on tracer flowback profiles. The properties of tracer dispersion and adsorption are also considered in both fractures and matrix. To validate the proposed model, a comparison is performed with the commercial software Eclipse. Sensitivity studies are performed to investigate the influences of parameters on tracer flowback profiles. The presented simulator is applied in the Changji oilfield to characterize individual stages using a stochastic method. Results show that the developed simulator is accurate as with Eclipse. The filed application indicates that the presented simulator can model tracer flowback and production performance. Sensitivity studies show that an increase in permeability modulus and the viscosity of fracturing fluid delays the peak arrival time. An increase in permeability modulus decreases the peak amplitude while an increase in the viscosity of the fracturing fluid increases the peak amplitude. An increase in a dead pore volume and a residual resistance factor has little impact on the peak position and lowers the tails of tracer flowback profiles. The sensitivity studies provide valuable information in characterization of a fracture network using tracer flowback data.