In the last two years, many of the available tubing and casing connections have been evaluated at the Exxon test facility. While we have not tested every available connection on the market, these tests have covered a range of connection types which incorporate the most commonly utilized connection design features. These features include various thread forms and configurations and also independent seal designs and seal combinations. In all cases, testing showed behavior that either differed from the manufacturer's published claims or varied from the behavior predicted by generally accepted mathematical models. The behaviors this paper will discuss include stress magnitudes and distributions, leak resistance under combined loads, and individual pressure and tension capabilities. Conclusions based on this test data will be presented for each design feature tested. Also included in presented for each design feature tested. Also included in the discussion are overviews of the new technology in connection designs which have evolved, in the most part, from attempts to solve the problems brought out in the existing test data. While most of these problems have not led to catastrophic failures in the field or laboratory, they have pointed out the need for better connection performance models and laboratory verification of these performance models and laboratory verification of these models.
Humble Oil and Refining Company (now Exxon Company, U.S.A.) began testing tubular connections in the 1950s in an effort to improve the reliability of the standard API connections. Out of the results of this early testing, torque-turn was invented. Torque-turn is a procedure for making up API tubing and casing connections procedure for making up API tubing and casing connections which ensures full thread engagement and maximum sealing reliability. With these encouraging results, we have continued to test both API and proprietary connections to verify the load carrying capabilities and field reliability of the connections to be run in our wells.
As the well environments have become more corrosive and downhole loadings have increased with more complicated wellbore geometries, the type of tests performed and data collected have become more numerous performed and data collected have become more numerous and extensive. We have progressed from hydrostatic burst tests through long-term internal gas pressure tests with tension and temperature cycling, to internal and external strain gage stress analysis under makeup and combined loading to failure.
The design and manufacture of a threaded connection is no simple task. These connections are subjected to complicated loadings, often including internal and external pressure, tension or compression, bending, torsion, extreme pressure, tension or compression, bending, torsion, extreme temperatures, and corrosive fluids. These loads can go from one extreme to the other during normal well operations such as stimulation versus flowing conditions. Often these loads are applied as shock loading rather than static loading. The connection must be designed so that it can be machined within reasonable tolerances and repeatability and made up and run by people who are usually more concerned with getting it in the hole fast rather than carefully. In addition, axial loads must be carried by the connection threads so that the failure mode and load value are predictable and so that axial loads do not harmfully affect the sealing capability of the connection. Probably the most contrary design conditions are the need Probably the most contrary design conditions are the need to seal high pressures (which usually requires high interferences) with the need to keep the stress level low to avoid sulfide stress cracking in sour (H2S) environments.
Our discussion will be separated into four categories.
1. Axial Loads
2. Pressure Loads
3. Combined Loads
4. Corrosion Resistant Alloys