A decision support model for differential sticking avoidance

Abstract

An innovative theoretical model to quantify the risk of differential sticking is presented. The proposed risk assessment is based on the concept of likelihood versus consequence. The likelihood of the problem’s occurrence in a given wellbore segment (case) is evaluated from a knowledge-based model and translated by a similarity measure of relevant operational conditions between the target case and historical cases with known outcomes. The stand alone module performed satisfactorily and predicts the likelihood of occurrence by more than a chance probability, demonstrated by a rate of sixty eight percent (68%) correct predictions against field data from forty four wells drilled by different operators in several fields.The consequence assessment is performed through an unidimensional mechanistic model that predicts the downhole overpull (differential sticking force) and performed well while estimating reported overpulls from known field instances of the problem. Together, the models serve as a risk assessment tool able to correctly describe risk operational trends while designing or drilling wells, with critical situations being defined as high likelihood plus high potential overpulls. Both models utilizes unique experimental data about mechanical properties of drilling fluids filtercakes (hardness, torque resistance and adhesion-cohesion strength) under simulated downhole conditions, raised through the HTHP Mudcake Characterization Equipment developed during the course of this research work. Moreover, the study contributes towards the development of modern predictive models aiming at combining large amount of available operational drilling data (LWD, PWD, mudlogging, survey, drilling reports, etc), expert’s knowledge, laboratory data and phenomenological models in order to optimize drilling operations.