Abstract
This article, written by JPT Technology Editor Chris Carpenter, contains highlights of paper SPE 199707, “A Data-Driven Approach to Screenout Detection for Horizontal Wells,” by Xiaodan Yu, Whitney Trainor-Guitton, and Jennifer Miskimins, SPE, Colorado School of Mines. The paper has not been peer reviewed. Multistage hydraulic fracturing has gained global popularity as more tight geologic formations are developed economically for hydrocarbon resources. However, screenout is a major issue caused by the blockage of proppant inside the fractures. The complete paper presents a screenout-classification system based on Gaussian hidden Markov models (GHMMs) trained on simulated data that predicts screenouts and provides early warning by learning prescreenout patterns in surface-pressure signals. The methodology is a useful tool for early screenout detection and shows the promise of other fracturing time-series data analysis. Materials and Methods In the complete paper, fracturing treatment data are generated using a hydraulic fracturing simulation software. A well-logging profile acquired from a vertical well located in the Denver-Julesburg (DJ) Basin is used to generate the reservoir rock properties in the fracturing simulations.
Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
