Predicting proppant flowback from fracture-stimulated wells

Abstract

In this thesis, the usefulness and benefits of predicting proppant flowback in the design stage of hydraulic fracturing treatments are evidenced. A new prediction model, as well as a methodology, is proposed in this work. These tools will help companies handle this phenomenon and consequently conduct more efficient fracturing treatments. Currently, proppant flowback is responsible for creating operational complications, increasing expenses and decreasing the productivity of fracture stimulated wells. So far, there have been some empirical approaches that have tried to explain this phenomenon and have helped identify the most important factors that influence it. However, all previous models have drawbacks and do not extend to all practical applications. In this work, the most relevant studies in the area were analyzed in order to clarify the mechanisms that govern the proppant flowback phenomenon. After doing this, the most consistent available features were included in a proposed semi-mechanistic model. This model is considered to be the most rigorous available approach to predict proppant flowback in future treatments. As part of this study, field cases that reported back-production of proppant were analyzed. From this analysis, it was demonstrated that proppant flowback was possible to anticipate. In addition, it was suggested the inclusion of a "Stability Criterion" in the design of future optimum fracturing treatments.