The procedures developed by Exxon Production Research Company (EPR) for evaluating the performance of premium tubing and casing threaded connection designs using a combination of nonlinear finite element analysis (FEA) and full-scale physical tests are described in this paper. Over the past several years, EPR has developed advanced techniques for nonlinear FEA to accurately and efficiently calculate the stresses and deformations in threaded connections. Concurrently, EPR's Tubular Goods Test Facility (TGTF) was constructed to assist in verification of FEA models and to conduct full-scale physical tests. Evaluation procedures have been developed that are based on using the combined results of state-of-the-art nonlinear FEA and rigorous full-scale physical tests. Finite element models are "linked" to full-scale tests to form the tools for evaluating threaded connections. Several years of experience in implementing these two evaluation tools together have revealed a unique and valuable synergism between analytical models and physical tests. EPR's evaluation philosophy is to use knowledge gained from advanced accurate FEA to test fewer connections with greater confidence in the results.

For burst design, engineers routinely assume the casing annular space is filled by a fluid equivalent. This assumption ignores mechanical resistance provided by solid cement. Some studies addressed this shortcoming by modeling the cement sheath as a solid with elastic failure criteria. Prior work used cement elastic modulus and Poisson ratio to classify cement as 'ductile' (soft) or 'brittle' (hard).In the current study, numerical results from finite element analysis (FEA) indicate that casing burst resistance is increased by the presence of the cement sheath. This study focuses solely on improvement offered by the cement sheath to casing burst resistance and ignores consequences of cement failure on overall well integrity.Comparisons are provided for casing burst resistance assuming various backup profiles. These include fluid hydrostatics, solid cement matrix (both elastic and plastic response) and cement as 'loose' particles. The fluid hydrostatics include: a) mud weight in hole; b) cement slurry density; c) mixed-water density; d) normal pressure (salt-water column); and e) actual pore pressure. Calculations show that these fluid profiles are conservative when used as burst-resistance backup. Original cement slurry density is least conservative.Since well designers are familiar with fluid profile backup assumptions in casing burst design, recommendations are provided to approximate cement behavior as particles with a fluid profile. This allows ease of calculation and is consistent with current practice. Guidelines are provided to explicitly calculate the enhanced casing burst resistance due to the particulate cement.

Summary For burst design, engineers routinely assume that the casing annular space is filled by a fluid equivalent. This assumption ignores mechanical resistance provided by solid cement. Some studies addressed this shortcoming by modeling the cement sheath as a solid with elastic failure criteria. Prior work used cement elastic modulus and Poisson's ratio to classify cement as "ductile" (soft) or "brittle" (hard). In the current study, numerical results from finite-element analysis (FEA) indicate that casing burst resistance is increased by the presence of the cement sheath. This study focuses solely on improvement offered by the cement sheath to casing burst resistance and ignores consequences of cement failure on overall well integrity. Comparisons are provided for casing burst resistance, assuming various backup profiles. These include fluid hydrostatics, solid cement matrix (both elastic and plastic response), and cement as "loose" particles. The fluid hydrostatics include mud weight in hole, cement-slurry density, mixed-water density; normal pressure (saltwater column), and actual pore pressure. Calculations show that these fluid profiles are conservative when used as burst-resistance backup. Original cement-slurry density is least conservative. Because well designers are familiar with fluid profile backup assumptions in casing burst design, recommendations are provided to approximate cement behavior as particles with a fluid profile. This allows ease of calculation and is consistent with current practice. Guidelines are provided to explicitly calculate the enhanced casing burst resistance caused by the particulate cement.

Summary The existing American Petroleum Institute (API) equation for internal leak predicts the internal pressure to overcome the pin-box contact pressure generated from the makeup interference plus the energizing effect of internal pressure, which enhances the seal. For threaded connections, internal and external pressures close the connection and increase the leak resistance, whereas axial loads open the connection and decrease the leak resistance. These competing effects must be included to accurately assess the connection leak resistance under any combination of loads at any point in any string. Following the same approach used by the API for internal leak, this paper obtains similar results for external leak. For API connections, the effects of combined axial force and backup pressure are then incorporated into the internal/external leak equations using results from a “toy connector” elastic model. Sensitivities of leak ratings to combined loads for API connections are presented for both tubing and casing sizes. An example design case shows the importance of considering combined loads.

Because pin and box are offset, applied axial tension (or compression) creates axial shear across engaged threads that opens a connection and reduces leak resistance. Less understood is the pin-end effect which does the same thing. Internal pressure acting on the pin nose pushes the pin out of the box, creating an additive axial force that contributes to the thread shear and decreases leak resistance. This paper addresses these axial loads for threads with non-symmetric flank angles. Multiple new concepts are presented. First, the thread shear equation, which depends on the pin-end effect, is developed for Buttress type threads in terms of the two flank angles. Second, the leak ratings for specific Buttress connections are presented in terms of a new leak criterion with dependence on backup pressure and axial force. Third, the leak dependence of Buttress type threads on backup pressure is shown to be significant and suggests that qualification testing of any connection for leak, without backup pressure, is not adequate.