Prediction of the continuous probability of sand screen-out based on a deep learning workflow

Abstract

Sand screen-out is one of the most serious and frequent challenges that threaten the efficiency and safety of hydraulic fracturing. Current low prices of oil/gas drive operators to control costs by using lower viscosity and lesser volumes of fluid for proppant injection - thus reducing the sand-carrying capacity in the treatment and increasing the risk of screen-out. Current analyses predict screen-out as isolated incidents based on the interpretation of pressure or proppant accumulation. We propose a method for continuous evaluation and prediction of screen-out by combining data-driven methods with field measurements recovered during shale gas fracturing. The screen-out probability is updated, redefined and used to label the original data. Three determining elements of screen-out are proposed, based on which four indicators are generated for training a deep learning model (GRU – Gated Recurrent Units, tuned by the Grid search and Walk-forward validation). Training field records following screen-out are manually trimmed to force the machine learning algorithm to focus on the pre-screen-out data, which then improves the prediction of the continuous probability of screen-out. The Pearson coefficients are analyzed in the STATA software to remove obfuscating parameters from the model inputs. The extracted indicators are optimized, via a forward selection strategy, by their contributions to the prediction according to the confusion matrix and root mean squared error (RMSE). By optimizing the inputs, the probability of screen-out is accurately predicted in the testing cases, as well as the precursory predictors, recovered from the probability evolution prior to screen-out. The effect of pump rate on screen-out probability is analyzed, defining a U-shaped correlation and suggesting a safest-fracturing pump rate (SFPR) under both low- and high-stress conditions. The probability of screen-out and the SFPR, together, allow continuous monitoring in real-time during fracturing operations and the provision of appropriate screen-out mitigation strategies.