Located in approximately 4,000 ft (1,219 m) of water in the Green Canyon block of the Gulf of Mexico, Chevron’s Tahiti prospect has well depths exceeding 28,000 ft (8,534 m). These are challenging conditions for high-pressure subsea completions. Because of the extreme conditions of the reservoir, the Tahiti #1 discovery well’s completion and well test required new equipment to successfully perforate, frac pack and flow test. Technologies developed for the Tahiti well test include a new perforating and frac-pack completion string.

The subsea completion string had to provide fishability within the 9 7/8-in. tieback (limiting the connection outside diameter (OD) to 7 in.), efficient fracturing hydraulics, mechanical and pressure integrity to absolute pressures up to 29,000 psi during screenout, sustained elevated external pressures of 24,500 psi absolute during fracture stimulation with a tubing leak operations and the capability to complete multiple trips (20) while maintaining sealability.

Similar wells have been completed with up to four OD tapered strings. Tapered strings limit the treating rate during fracture stimulation and often yield a less desirable fracture network, potentially reducing production capability. To maximize Tahiti’s production rates, a frac-pack completion was selected to create an extensive fracture network in the highly overburdened formation. Treating rates during the frac pack were deemed critical for the greater than 17,000 psi treating pressure. A large diameter completion tubular and a new connection that could meet all loads from top to bottom were required.

Design and qualification

The design and qualification program included predicting loads applied to the completion string and identifying safety factors. Finite element analysis (FEA) was used to simulate loads and confirm the design. The connection was physically tested to predicted loads and physical test data correlated back to FEA-generated predictions.

Determination of loads and stresses applied to the completion string was crucial. The completion string was modeled for a series of load cases including hook load, fracture stimulation, fracture stimulation with leak, screenout during stimulation and perforation of the well. Values for axial loads, pressure loads and temperature were calculated along the depth of the completion string for these load cases.

Defining tubular specifications was straightforward. The tube had to meet load requirements, accommodate a maximum connection OD of 7 in., a minimum connection inside diameter (ID) of 4 ¼ in. and have a large tube ID to maximize hydraulic efficiency during the frac pack operation. A 5 7/8-in. OD, 0.500 in. wall, S-135 grade pipe suited all load conditions and provided the largest ID possible for minimal fluid friction loss during the frac pack.

For the connections, Grant Prideco’s eXtreme torque XT connection offered an improved thread form and reduced taper compared to the API connection design. For completion operations, the company’s metal-to-metal seal version of XT, the XT-M connection, provided a pressure-rated rotary shouldered connection (RSC). This connection, with a radial metal-to-metal flank seal on the pin nose, is gas-tight pressure rated to 15,000 psi internal pressure and 10,000 psi external pressure. A 7-in. OD and 4 ¼-in. ID configuration is standard for XT-M57 on 5 7/8-in. pipe.

The connection was evaluated for absolute pressure, a novel FEA approach. When analyzing connections with FEA, it is typical to apply only differential pressure across the connection. If internal and external pressures are high, this approach can be misleading. The application of absolute internal and external pressure, even when equal in magnitude, generates tangential and radial stress within the connection. FEA results revealed that the connection’s extended box counterbore lacked stability under the anticipated external pressure and compressive loads at the bottom of the well. The fracturing-with-leak case (4,000 psi leak) resulted in high compression (215,160 lb plus compression from makeup) and external pressure (25,279 psi), which significantly exceeded the material strength at the box counterbore. Collapse was predicted. A modification of the existing connection was necessary.

To provide additional rigidity and structural stability to the box counterbore, the counterbore length was shortened from 2 ¼ in. to 5/8 in. A series of design iterations determined the proper pin length, box depth, and gap width to balance the stresses in the box counterbore and pin nose. The metal-to-metal seal design remained unchanged.

The modified connection (CT-M57) was analyzed for anticipated Tahiti load cases. Although there was significant improvement, the modified connection still showed stress values in the box counterbore exceeding the 120,000 psi material yield strength for the frac-with-leak-load case at the bottom of the well.

High-strength tool joint material

API tool joints have a minimum yield strength requirement of 120,000 psi. The Tahiti project would require a 135,000 psi minimum yield strength connection. A program to develop the industry’s first commercially available 135,000 psi tool joint material was initiated. Mechanical properties were identified.

The material’s fracture toughness for the upper end of the strength/hardness range was critical. Fracture toughness equivalent or better than current requirements for standard tool joint material would be required. This was accomplished with special 4130 modified steel with high alloying elements, low nonmetallic impurity elements and a quenching medium that assures martensite transformation.

Testing challenges

Simultaneously applying internal and external absolute pressures instead of the more common differential pressure created a challenge in physically testing the CT-M57 connection. To qualify the connection for Tahiti, the desired test would prove the ability to seal fluid pressure and the ability to withstand the elevated axial, hoop and radial stress levels the connection would be subjected to in the field. Six CT-M57 connections underwent a three cycle test program over a 2-month period, and all passed without incident.

To further validate the design process, strain gauge data obtained during physical testing was compared to earlier FEA strain predictions. FEA prediction correlated reasonably well with the test data, especially in predicting overall trends. Physical test data did not indicate overall strain levels higher than FEA predictions and no failures or leaks were observed during the test program. The connection was accepted for use on the Tahiti project.

Results

The Tahiti #1 well was completed and flow tested from May to August 2004. The 5 7/8-in., CT-M57 string was used to clean out the well, displace the drilling mud with completion fluid, perforate and fracture the reservoir. The large ID allowed efficient hydraulics, enabling elevated frac pack treating rates and successful screenout. The string performed without incident in all operations.

Evolution of rotary shouldered connections has continued with the development of a new, third generation RSC design, TurboTorque. CT-M design improvements have been incorporated in this new design.

Engineering and design typically result in tradeoffs. Second generation designs such as CT-M have increased torque and hydraulic performance but with a trade-off of increased make-up time. Because of their slower tapers and shallower stab depth, second generation connections typically require 2-3 times the number turns from stab to makeup compared to API connections. TurboTorque has eliminated this tradeoff by using a proprietary double-start thread form that reduces the revolutions from stab by at least 50%. This provides a considerable savings in rig time and money.