Interpretation of interwell tracer tests using discrete fracture model

Abstract

Well-to-well tracer tests are used for getting knowledge about the flow behavior in the reservoir. The results of interpretation of these tests are especially important when enhanced oil recovery techniques that use expensive fluids such as surfactants, micellar fluids, or polymers are modeled. The classical method of quantitative interpretation assumes that the tracer filtration channels are completely isolated from the formation. In this paper, an alternative method for interpreting the results of well-to-well tracer tests is proposed. High-permeability channels are represented as discrete fractures that connect producing and injection wells. A numerical solution of the convective transport equation is used for modeling the tracer flow. The flows between the fracture and the reservoir are fully resolved, so the above-described drawback of the classical technique is overcome. The interpretation is carried out by constructing a reservoir model and determining the model parameters by solving the inverse problem. The model is verified, the well-posedness of the inverse problem is analyzed, and the comparison of the method with the classical interpretation method is performed. Using two fractures as an example, it is shown that the quasi-solution of the inverse problem exists and continuously depends on the input data. The interpretation results with this method are shown to be similar to the results of the classical interpretation in the absence of crossflows between the fracture and the reservoir. It was found that the presence of crossflows can increase the estimated fracture volume by 200 or more times. It is also shown that the crossflows between the fracture and the formation can completely prevent water breakthrough along the fracture into the production well.