Design Methodology and Operational Practices Eliminate Differential Sticking

Abstract

Summary This paper discusses a successful initiative begun 6 years ago to eliminate differential sticking across global operations. In the 5-year period from 2004 through 2008, there were only three differential-sticking events in 3,476 wells drilled with the recommended practices. There were an additional 17 sticking events with designs that did not conform to recommended practices, and 14 of these were freed. The drilling environment was diverse. Overbalances in excess of 1,000 to 2,000 psi were common in multidarcy rock and at high angle, and depleted reservoirs have been drilled with overbalance as high as 7,800 psi in vertical wells. The early focus of the stuck-pipe-avoidance practices was the elimination of differential sticking. However, some level of sticking occurs routinely in drilling operations, and these events become problematic only if the force required to initiate pipe movement exceeds what can be delivered to the stuck point. It is now accepted that sticking cannot be prevented and that elimination of sticking is not a proper design objective. The philosophical objective has now shifted from elimination of sticking to "maintaining conditions that allow the pipe to be pulled free," assuming that it will become differentially stuck. The desire to maintain this ability to move the pipe has required the implementation of a range of practices, some of which were not common in the industry. Changes were made in bottomhole assembly (BHA) design, fluid design, real-time cake-shear-strength recognition, and realtime cake-remediation practices. A finite-element (FE) model was also applied to redesign new systems or applications that lie outside the operator's previous experience. The stochastic model predicts cake growth and sticking force and the probability that it will be possible to deliver a force that can free the pipe for any given still-pipe time. The model inputs were calibrated through pullout tests with a variety of fluids to determine mechanical cake-strength properties, the rate at which those properties develop, changes in the pressure transient through the cake as it matures, and the cake contact areas and geometry at any point in time. Engineering and operations training also contributed greatly and allowed relatively uniform implementation to be achieved across a large, globally diverse operation in less than 1 year. A small number of noncompliant designs continued to be used, and these contributed greatly to the incidence of stuck pipe in the first 3 years. Last year, there was only one incident of stuck pipe with a noncompliant design. The paper describes the underlying sticking concepts, the engineering design and field practices used, the modeling capability, and the field results.