1.1 Project Background
The work presented in this report is the third in a series of annual projects conducted by the Production Research Advisory Committee (PRAC) of the American Petroleum Institute to investigate the factors affecting leak resistance in API connectors. PRAC projects 84-53 and 85-53 investigated leak resistance in 8-Round connectors in 1984 and 1985, and this project 86-53 analyzes the factors affecting leak resistance of Buttress connectors.
The Technical Advisory Committee established by PRAC to direct the work consists of the following:
The contract for the work was executed March 19, 1986. Status meetings of the Technical Advisory Committee were held on April 2, May 13 and November 19, 1986. Also, presentations were made at the API Standardization Conference in New Orleans June 23-26, 1986, and API task group work week in San Antonio December 8-11, 1986.
1.2 Buttress Connector Background
The Buttress connector was patented November 27, 1956 by Mr. Samuel Webb and assigned to United States Steel Corporation. Statements in the patent indicate that the goal of the Buttress connector is to provide joint strength equal to pipe body strength, and that high leak resistance should not be expected unless the thread clearance on stab flanks is closed. The patent recommends mismatched leads to accomplish leak resistance.
API design equations for Buttress connectors are based on tests conducted at U.S. Steel in 1954-5. The tests were conducted by Mr. Webb, and were interpreted by Mr. Bill Clinedinst to establish API performance properties at pipe yield for both tension and internal pressure.
Despite relatively large thread clearances with nominal dimensions, early experience with Buttress did not indicate difficulties with leak resistance. One explanation is that tolerances were not held as close with early manufacturing techniques and variations in lead caused clearances to close, resulting in higher leak resistance. However, more recent manufacturing techniques hold tighter tolerances on lead, introducing clearances and lowering leak resistance.
One popular approach to achieve leak resistance in Buttress connectors is to use tin plating. During make-up the tin plating fills the thread clearances to seal the leak path. However, API performance properties do not account for the presence or absence of tin plating on threads.
The overall objective of this project is to develop information for the API Task Group on performance Properties for API Buttress connectors. Specific objectives are as follows:
1. Verify computer model accuracy by comparison of model predictions to laboratory test data.
2. Define the contact pressures, thread clearances, and stresses in Buttress connectors for nominal dimensions in 5-1/2″, 9-5/8″, and 13-3/8″ casing.
3. Establish the changes in contact pressures, thread clearances, and stresses in Buttress connectors as a result of changes in diameter, make-up, thread lead, and tension.
1.4 Work Description
Work conducted in this project consists of the following tasks:
1. Simulate conditions for tests conducted by Nippon Steel Corporation, and compare calculated and measured strains.
2. Generate computer models for 5-1/2″, 9-5/8″, and 13-3/8″ nominal Buttress connectors, and modify the models for mismatched leads in 9-5/8″ and 13-3/8″.
3. Execute make-up and tension load cases for the connector models created in step 2.
4. Plot and interpret stresses, strains, crest/root contact pressures, thread flank clearances, and displacements.
5. Communicate results of project through TAC meetings, presentations at API conferences, discussions, progress reports, and this final report.
Conclusions apply to buttress connector performance only, but may be extended to other connectors in some cases. In view of the lack of a reliable leak criteria for leak through a thread clearances, results and conclusions in this project address the factors affecting leak, but do not predict specific leak pressures.
Some conclusions presented here may be obvious to some readers, but are included for completeness. Following are conclusions based on the calculations in this project and interpretations that draw on related experience:
1. Clearances and displacements due to make-up and tension loads are significantly greater for API BTC connectors when compared to 8-Round. This requires thread compound to serve the major role in leak resistance by sealing the larger clearances.
2. Due to relatively large thread clearances in Buttress connectors, the most likely leak path is through the thread stab flank clearance spiral.
3. An important factor influencing leak resistance in Buttress connectors is the change in thread clearances during make-up and tension loads. During make-up, radial compression on the pin causes it to lengthen axially due to Poisson's effect and radial expansion of the coupling causes it to shorten. The lengthening of the pin and shortening of the box cause load flanks near the end of pin to open. Load flank stand-off near the end of pin increases with make-up turns. Subsequent tension closes load flanks, thereby opening stab flanks and introducing a leak path.
4. During hand tight make-up, load is transferred from the thread stab flanks to the load flanks. The initial stab flank contact may cause thread compound to be wiped off and may affect leak resistance.
5. Mismatched leads, properly applied, will tend to assure load flank bearing at one end of the full depth threads on the pin and stab flank bearing at the opposite end during make-up, thereby improving leak resistance.
6. Poisson's effect in the threads during make-up results in an increase in pin crest clearances with make-up turns, thus reducing leak resistance. During make-up, Poisson's effect on the threads in minor, but does result in a reduction of crest or root contact pressures.
7. Unlike some 8-Round connections, additional make-up turns beyond nominal may not result in increased leak resistance because of the effects described in 6 and 3.
1. For improved leak resistance, it is recommended that API Specifications 5B on machining tolerances for lead be modified so that thread clearances are reduced or eliminated. Within the existing tolerance range, stay on the negative side of lead tolerance for the pin and on the positive side of lead tolerance for the coupling if at all possible during machining. Do not allow lead to vary in the opposite direction (positive on pin or negative on coupling). This is consistent with the original patent on the API Buttress thread form for casing.
2. Establish a set of leak criteria for utilization of the results presented here. Specifically, two leak criteria are needed to better understand leak:
2.1 Leak past a sealing surface as a function of contact pressure and related variables.
2.2 Leak past thread compound in a thread clearance as a function of clearance dimensions and changes in clearances.
3. Due to the clearances and displacements involved in BTC connectors, and the unknowns concerning thread compound, Buttress connectors intended for high pressure service should be tested prior to field use. Testing should include thermal cycling with axial load and internal pressure cycling with dimensions in worst case combination.
4. Calculate API pressure performance properties for each diameter, weight, and grade based on contact pressures, clearances, and changes in clearances presented here and leak criteria recommended in 2. Change API performance properties as needed.