Advances in Prediction of Stuck Pipe Using Multivariate Statistical Analysis

Abstract

ABSTRACT Stuck pipe is known to occur as a result of several major mechanisms including wellbore instability, improper hole cleaning, and differential sticking forces. Using an evolution of the discriminant analysis methodology first published by Chevron1,2, a second generation engineering statistical model has been developed that incorporates the use of physical stuck pipe parameters in combination with a large body of drilling data to quantify the probability of stuck drill pipe for a given day of a well. Further, the tool predicts which stuck pipe mechanism is driving sticking risk and can thus be used to modify drilling variables to minimize the likelihood of sticking. An example application of the model in a recent Gulf of Mexico well is given.