The severe loading conditions experienced by the casing strings in thermal wells used in Cyclic Steam Stimulation (CSS) and Steam Assisted Gravity Drainage (SAGD) operations pose a significant challenge in terms of the casing connection sealability performance and integrity. The industry has recently developed standards and protocols for assessing and qualifying connection sealing performance for such applications. The comprehensive programs include both Finite Element Analysis (FEA) evaluations and a series of full-scale physical tests. The FEA evaluations are required to determine the worst-case connection design scenarios, in terms of geometry, material properties and make-up torque, for fabrication of the specimens to be assessed in the physical tests. The scope of the FEA work includes analyzing the connection sealability performance under selected load cases representative of the intended field service conditions. A suitable minimum sealability criterion is required to assess the adequacy of the predicted connection sealing performance for the specific application.

This paper presents the results from an experimental metal-to-metal seal investigation and describes their use to establish a simplified sealability criterion for premium casing connections. The leakage response of several different seal samples was determined as a function of contact stress level for differential gas pressures of 3.3, 7.4, 12.1 and 16.5 MPa, corresponding to Application Severity Levels (ASL) of 240, 290, 325 and 350 per ISO/PAS 12835. The test results were used to characterize the relationship that existed between the leakage rates recorded in the tests and the selected control variables: differential pressure; seal length; tubular size; and seal contact stress level. The paper also demonstrates the use of the proposed sealability evaluation criterion in conjunction with the FEA results established for a 244.5 mm, 59.5 kg/m L80 premium connection subjected to thermal cycle loading.

You can access this article if you purchase or spend a download.