The purpose of the study was to identify and evaluate the variables that detennine the leak resistance limit state functions of API 8-round and buttress [1] connections. This work will be incorporated into Load and Resistance Factor Design (LRFD) equations. Connectors are an integral part any well design program. Therefore, it is vital that they be included in the LRFD design approach. Makeup, tension, internal pressure and dimensional data were among the variables in the evaluation, which was based on finite element analysis, testing and structural mechanics. The leak resistance limit state for round thread connections is defined by contact pressure, stab flank engaged length and coupling yield, while for buttress is defined by contact pressure, stab flank clearance and coupling yield. Leak pressure, as defined by API Bul. 5C3 [2], is a function of makeup and dimensional data independent of thread type, tension, and pipe inside diameter and valid only in the elastic regime. Tension is detrimental in the leak resistance of 8-round connections, but does not compromise buttress leak resistance. Regression analysis was perfonned on structural mechanics results to produce the highest correlation to finite element results to account for end effects on round thread connections. It was detennined by testing that stab flank contact over a minimum length of engagement is not sufficient to prevent 8-round leak, despite sufficient contact pressure level. Teflon impregnated thread compound should be the choice for API buttress. The indications are that optimized makeup should be considered in the leak resistance capacity of API connections. Excessive makeup and/or tension or pressure could result in coupling yield causing leak upon re-pressurization. Further analytical and References and figures at end of paper 653 --------------experimental work would improve the degree of accuracy with which the leak resistance capacity can be detennined.

The original American Petroleum Institute (API) tubular specifications were established in the early 20th century with the intent to standardize pipe sizes and connections so that material from one mill/user could be assembled with material from other sources. The original API specifications for threaded connections did not focus specifically on ensuring leak resistance-only interchangeability. Today, the importance of connection integrity and reliable leak resistance is more widely recognized. Casing and tubing strings represent the primary environmental and safety barrier(s) for oil and gas containment. The threaded connections for these strings are one of the performance-limiting and most critical features for the tubular product. To address the leak resistance issue, an API work group consisting of user and manufacturer representatives developed a performance-based connection specification for LTC (Long, Threaded and Coupled)1 connections for casing and tubing. These new specifications are currently being finalized as a Supplemental Requirement to API Specification 5CT, SR22. Development of these supplemental specifications employed Finite Element Analysis (FEA), full scale physical testing, and evaluations of product specifications, inspection and gaging methods, process control, makeup and thread compound. The objective of this API initiative is to minimize costly failures, increase performance, improve delivery and decrease inventory, and minimize supplier and user life cycle cost by developing an engineered connection as an industry standard. API Supplemental Requirement (SR) 22 specifications enable manufacturers to supply a reliable fieldready product with the performance users require. The Incentive to Develop API SR22 Connections Tubulars are a fundamental component of well design. Tubular connections are one of the performance-limiting aspects of tubular products. The critical nature of connection performance is reflected in the Industry's failure experience. Typically, connection failures account for up to 90% of all tubular failures. One user's failure database, for example, cites twice as many connection failures as all other modes combined (Fig. 1). Remarkably, typical well design methodologies do not address threaded connection performance in any detailed fashion. Rather a safety factor is applied to pipe body ratings. In contrast, SR22 methods specifically address connection performance and reliability. Performance limits have been defined and specifications have been developed to ensure leak resistance of 100% API internal pressure rating with a tension design factor of 1.6. The objectives of developing an engineered performancebased LTC connection as an industry standard are:Identify and eliminate the sources of connection failures,Optimize performance and cost,Save suppliers and users time and labor, andImprove user-supplier relationships. These objectives have been pursued through an API work group representing joint industry input from both the manufacturer and operator communities. The LTC is an 8- Round thread form connection. 8-round connections are the workhorse of the industry, representing over 50% of OCTG production. 8-Round connections are supplemented with API Buttress and proprietary connections in deep or critical wells.

This paper covers research funded by the API Production Research Advisory Committee (PRAC) on leak resistance of ~PI 8-~ou~d connectors. It details the sensitivity of leak resistance to variations in makeup turns, pipe diameter, grade, and applied tep-Slon. FIndIngs show that the leak resistance of the connector relative to pipe-body ratings increases with the number of makeup turns ~nd decreases as diameter and yield strength increase. Finally, tension is found to lower leak resistance in a manner that renders hydrotesting Insufficient for defining leak resistance in typical service conditions.

Summary This paper presents stress and leak-resistance equations based on the theoryof elasticity for API 8-round connectors in tension and compares results ofthese equations to those of the finite-element method (FEM) and full-scalephysical testing. The new equations identify significant nonconservativeaspects inherent in current API methodologies. Introduction Interest concerning leak resistance of API 8-round connectors prompted theAPI to fund research that identified and assessed many parameters affectingleakage. This research, based on the FEM, provided details that mark a turningpoint in the definition of leak resistance. Although trends in leak resistanceare clearly identified, quantitative application of these data to 8-roundconnectors other than those analyzed is not possible. Consequently, equations were developed that evaluate API 8-round connectorstress and leak resistance on the basis of the theory of elasticity. Theseequations consider such parameters as coupling OD, pitch diameter, pipe wallthickness, engaged thread length, taper mismatch, number of turns duringmakeup, tension, and internal pressure in determining connector stress stateand leak resistance. The equations provide insight into the interaction ofthese parameters as well as a basis for specifying load capacity. API 8-round Connections Fig. 1 shows a cross section of an API 8-round connector. The threadedregion of the connector performs a dual function: it transfers axial loadbetween lengths of pipe and seals formation and internal fluid pressures. Roundthreads form sealing surfaces if the threads are free of foreign substancesthat prevent surface contact and if root and crest voids are sufficiently"plugged" with thread compound. Surface contact is established duringmakeup. Relative axial advancement in the tapered thread geometry results in aninterference fit between pin and coupling. The amount of interference willgenerally vary along the threaded region and depends on pin and box taper andthe number of turns during makeup. During makeup, root and crest voids arefilled and thread flanks are plated by the metal particles suspended in thethread compound. Fig. 2a shows a thread with the contact pressures resulting from makeup, indicated by the resultant loads on the stab and load flanks, Fcsf and Fclf, respectively. Because of Poisson's effect (sit), these individual thread forceswill generally not be equal; however, force equilibrium is maintained throughthe threaded region. A greater number of makeup turns causes greaterinterference and a higher thread-flank contact pressure. e.g. 2b shows the samethread after tension was applied. Stabflank contact pressure is reduced by anamount proportional to the applied tension, delta y, and Poisson's effect. Load-flank contact in-creases because of the axial load and is slightly reducedbecause of Poisson's effect. Fig. 2c shows the same thread with sufficient tensile load applied toeliminate stab-flank contact. The axial component of load-flank force is nowequal to the applied tension load. The thread has no leak resistance because ofstab-flank separation. The thread moves closer to a jumpout failure as tensionincreases. Fig. 2d shows a thread with makeup and pressure loading and no axial load. Internal pressure loading increases contact on thread flanks, as shown. Fig. 2 also shows that the thread-flank loads necessary to maintain leakresistance vary with makeup, tension, and internal pressure. Increasing thenumber of makeup turns causes higher thread-flank loads, which increase leakresistance. Stab-flank loads decrease with tension, which reduces leakresistance. Internal pressure increases flank loads and leak resistance. Thefollowing analysis concerns leak resistance in tension and examines stab-flankcontact pressures. The value of contact pressure required to seat a specific fluid pressure isan empirically based number and the subject of ongoing re search. Because somesolid particles in the thread compound plate thread flanks while the vehicle(silicon- or petroleum-based) is extruded with time and load, the sealingmechanism of the thread compound can be considered analogous to that of agasket. Research on the leak resistance of gaskets uses a sealing factor, A, toevaluate the effectiveness of a gasket: A = Pcr/Pi. A lower value of A relates to a more effective seal. Yield strength ofsurface material, contact pressure, thread compound filler material. and surface roughness all affect the A values. Because contact pressure is one parameter that affects the value of A andbecause contact pressure varies with the number of turns during makeup, connector diameter, weight, and grade, the sealing factor for 8-roundconnectors is a variable. An assessment of the sealing factor's effect onequation accuracy is made when it is compared with test data.

Evaluation of the structural capacity of perforated casing was perfiied for Mobil North Sea Ltd. applications.