Abstract
Complex fracture networks are usually created using hydraulic fracturing in tight oil reservoirs. The fracture networks between individual stages are considerably different. It is difficult to characterize individual-stage fractures using rate analysis. The production performance using SRV derived from microseismic events is usually inaccurate. Tracer injection-flowback can be employed to estimate individual-stage fractures. In this paper, a workflow for the fracture network characterization combing hydraulic flowback rate, tracer flowback profiles, and microseismic events is presented. Lorentz derivative curve is established to recognize large-scale fractures based on the time moment theory. The effective pore volume of the fracture network is estimated using the material balance method based on the flowback rate. The stochastic fracture network constrained by fracture volume and microseismic data is generated and the fracture-network properties are adjusted to match tracer breakthrough profiles and flowback rate using an in-house tracer transport simulator. The fracture networks can be classified into three patterns: small fracture-dominated network; several large fractures-dominated network; small and large fractures coexisting network. The number of the turning point in the Lorentz derivative curve can infer the number of large fractures. The number of microseismic events and orientation controls the number of fractures and connectivity. The estimated fracture volume controls total fracture length and width. All the information provides a rough estimation for the fracture network in the random fracture network generation. The properties of fractures are inversed by matching the flowback rate and tracer flowback profiles and the fracture conductivity is more sensitive to tracer flowback profiles than flowback rate. The presented workflow provides an efficient way to characterize the individual-stage fracture network using different tracers in the process of fracturing-flowback. Different fracture patterns are summarized according to the tracer flowback profiles, which is vital to production performance prediction and fracturing optimization.
