Benchmarking a new connection technology in the deep Gulf of Mexico

The use of a new generation of double-shoulder drill pipe connection yielded savings from both improved connection time and reduced damage.

Sheldon Langdon and John Connor, Chevron; R. Brett Chandler
and Michael J. Jellison, NOV Grant Prideco

Located in an ultra-deepwater region of the Gulf of Mexico, Chevron’s Saint Malo, Lewis, Turtle Lake and Northwood wells represent drilling projects that demand careful planning, precise execution and advanced technology. Because these and previous wells in the same area of the deepwater GOM were drilled with the same rig, equipment and crew, and had very similar well plans, they represented a unique opportunity to benchmark in the field a new double-shouldered, double-start drillstring connection.

A key objective for the development of the third-generation double-shoulder connection (DSC) was to improve connection makeup and breakout speeds relative to second-generation DSCs. The new connection incorporates a double-start thread that reduces by half the number of revolutions to assemble and disassemble the connection. Also, the 3G connection provides increased mechanical performance, larger ID and enhanced fatigue performance.

The first 3G DSC drillstring used by Chevron was 5⅞-in. 26.30-lb/ft S-135 in the Saint Malo well. The rig crew experienced an adjustment period during makeup of about the first 10 connections. Makeup speed of the new connection was realized, and analysis of well data demonstrated improved efficiency and cost savings. Chevron’s second use on the Lewis well resulted in similar time savings.

In response to the field performance feedback of these first two wells, a program was initiated to monitor and improve the value of the connection on future wells. The subsequent implementation of improved maintenance procedures and equipment modifications improved running times.

A key benefit observed throughout the project was reduced connection damage—over 80% reduction in repair rates and costs compared with similar wells using the 2G connections.

BACKGROUND

Typical deepwater GOM wells are located in water depths ranging 4,000–7,000 ft, and often reach total depths in excess of 30,000 ft. These wells are often directional with deep kickoffs and final well inclinations in the range of 30–40°.Bottomhole pressures in excess of 25,000 psi are encountered while nearing TD.

Spread rates for these deepwater GOM wells are approaching $1 million/day. Since these wells can often take over 100 days to complete, they can cost more than $100 million to drill. If the well is completed for production, the final total well cost can be nearly $200 million. In this environment, safely reducing non-productive time is vital to well construction cost reduction.

Due to the high number of casing strings to reach the target zone (Fig. 1), use of a single size of drill pipe, such as 6⅝ in., is generally not possible. Casing drift restrictions and the impact of the large-diameter tubulars on system hydraulics dictate the use of a tapered drillstring. A typical drillstring can often include 5-in., 5⅞-in. and multiple weights of 6⅝-in. drill pipe, in either drilling or completion mode.

Fig. 1. A typical deepwater GOM well with seven casing strings.

In drilling mode, Chevron typically uses about 7,500–10,000 ft of the 5⅞-in. drill pipe after the 13⅝-in. casing has been installed and cemented in the well, to drill below the 13⅝-in. casing to TD. The tapered string design and the location of the 5⅞-in. drill pipe in the lower portion of the drillstring allows for connections to be made up using the iron roughneck on the trip into the wellbore instead of using the top-drive system while drilling.

Because of this, and because the iron roughneck is not automated to capture connection makeup time, the only available measure of connection makeup efficiency is the time in slips during a trip into or out of the wellbore. Ultimately, the time in slips is a critical measure for the operations team. On a typical trip into the hole, over 60% of the trip time can be in the slips. The minimum number of trips per well with the 5⅞-in. drill pipe is normally three, but unplanned trips may also be required.

The first four wells in the deepwater drilling program discussed here used 5⅞-in. 2G DSC drill pipe in a tapered drillstring to drill the lower portion of the well. Subsequent wells switched to the new 3G connection, which was expected to promote faster drill pipe running and tripping speeds and improved overall drilling efficiency.

CONNECTION DEVELOPMENT

 First-generation DSCs, introduced in the early 1980s, were API rotary-shouldered connections (primarily numbered-connection and full-hole) with a second shoulder added inside the box member at the pin-nose interface. First-generation DSCs and API connections shared the same basic design features, such as thread form, taper, lead and pitch diameters, while the secondary shoulder provided a simple means to increase the connection torsional strength by about 40% over the corresponding API connection.

The increase in torsional yield strength allowed streamlined designs with increased ID and/or decreased OD for improved hydraulic performance of the drillstring. As more aggressive drilling programs were implemented, it became clear that a new generation of DSC with higher torque and more streamlined profiles was required. The second-generation DSC drill pipe, introduced in 1998, was designed with an enhanced thread form to reduce stress concentration, a flatter taper to increase shoulder area, and tighter tolerances.

The 2G connection provided about 25–30% more working torque capacity than 1G DSC did, or an improvement of about 65–70% compared to a standard API connection with the same OD and ID. The streamlined connection dimensions, enabling one pipe size larger to be run in the same hole size, dramatically improved hydraulic efficiency while maintaining equivalent fishing capability.

Since the introduction of 2G DSC, the industry’s continued trend toward deeper and longer-reach wells has dictated the need for drill pipe connections with enhanced mechanical and dimensional characteristics coupled with improved makeup and breakout speeds. As a result, a project was commissioned in 2005 to design, analyze and test third-generation DSC. One of the primary differences between 3G DSC and its predecessors is the addition of a double-start thread, which incorporates two thread leads spaced 180° apart to reduce by 50% the number of turns needed to assemble or disassemble the connection, Fig. 2. Changes in thread taper and thread pitch further reduce the number of revolutions required to make up or break out the connection. When comparing the 3G DSC to its 2G counterpart for use on 5⅞-in. drill pipe, the number of revolutions to make up or break out the connection is reduced from 13 to four. At equivalent rotation speeds, laboratory tests indicate an approximate 11-s time savings for either operation.

Fig. 2. All other things equal, the double-start thread form in 3G DSC reduces revolutions from stab to makeup by 50% compared with 2G DSC. Changes in thread taper and pitch further reduce total revolutions from 13 to four.

API tool joints are produced with specified minimum yield strength (SMYS) of 120,000 psi. Capitalizing on advancements in metallurgy and heat-treatment techniques for high-strength, high-toughness steel grades, the new 3G DSC employs tool joints of 130,000-psi SMYS to enable improved torsional capacity and more streamlined dimensions for hydraulic efficiency.

RUN-TIME ANALYSIS

To determine a baseline for analyzing 3G DSC speed performance, slip-to-slip times were monitored on four wells using 5⅞-in. 2G DSC. All wells were drilled using the same rig with the same top drive, pipe-racking system, iron roughneck and handling equipment. Review of the raw data indicated the need for data cleanup to eliminate short times corresponding to events such as repositioning of slips and long times corresponding to events such as circulating while in the slips.

The actual connection makeup time may only represent 15–30% of the total slip-to-slip time. However, when strictly considering increased value to the operation, connection time savings during makeup or breakout must positively impact slip-to-slip time in order to achieve operational time savings. This, along with the ease of obtaining accurate slip-to-slip time data through the mud-logging company, was the reason for use of this measurement to evaluate connection run-time performance.

Run-time analysis of 2G DSC wells. A total of 956 2G connections were made up or broken out online (i.e., while RIH/POOH) during the first four wells. Data from these wells (called Wells 1–4 in this article) was combined to provide a single data set to represent the baseline performance of 2G DSC. The total number of connections decreased to 703 following the data cleanup. Review of the 703 connections indicated the mean time for all TIH and POOH runs on the four wells to be 101.4 s, with a standard deviation of 22 s. Sixty-eight percent of all data fell between 79.4 s and 123.4 s, and 95% of all data fell between 57.4 s and 145.4 s.

Run-time analysis of Wells 5 and 6. Chevron first used 3G DSC on the Saint Malo well (Well 5), which was similar to the first four wells and was drilled with the same rig and equipment. Service personnel were dispatched to the rig to provide running and handling procedure training. The rig crew experienced a quick adjustment period making up the connections, and balance between use of the iron roughneck slow and fast spinner modes was found after about 10 connections.

On Well 5, 1,396 connections were selected for review and estimated to be a representative sample size for the entire population of connections made on the well. Of the 1,396 connections analyzed, 1,053 remained following data cleanup. Results were encouraging, but the expected 11-s time savings observed in the laboratory was not observed in the actual slip-to-slip times. The mean slip-to-slip time for the well was 99.0 s, a marginal improvement over the 101.4 s observed on the first four wells using 2G DSC. Though only a 2.4-s improvement per connection was observed, this resulted in about $25 savings per makeup or breakout of connection using the $900,000 spread rate. The significant number of connections made on this well resulted in substantial savings using the new connection.

Immediately following Well 5, Chevron again used 3G DSC on the Lewis well (Well 6). Like the previous wells, Well 6 used the same rig and equipment. It was drilled to 32,608 ft TVD with a reach of 2,689 ft. The 5⅞-in. drill pipe was on rent during Well 5 for a total of 319 days, while rented for only 64 days on Well 6.

A total of 725 3G DSC connections were made up/broken out online during Well 6, and all slip-to-slip time data was reviewed for the 725 connections. Following data cleanup, 565 connections remained. Results were nearly identical to those found on Well 5. The mean slip-to-slip time for the well was 100.8 s, a 0.6-s improvement over the 101.4 s observed with 2G DSC and 1.8 s longer than the 99.0 s observed on Well 5. Though some time savings were realized on both wells using the new connection, only marginal savings were observed in the slip-to-slip time data. More substantial cost savings were actually realized in connection repair costs following each well. A collaborative team was defined between the operator and connection manufacturer, and a project was initiated to increase the time savings value of the new connection while maintaining the repair performance observed to date.

Service field visit on Well 7. A field service specialist from the connection manufacturer was sent to the rig to investigate and work with the rig crews while drilling the Turtle Lake well (Well 7). During the rig visit, several issues were identified that adversely affected connection running speed. A major finding was that the automated iron roughneck required maintenance to provide the optimum performance it was designed to deliver. Slip-to-slip times observed during Well 7 were noticeably longer, due to the field specialist working with the rig crews during TIH and POOH.

More instructive was the change in slip-to-slip times after 1) completion of rig crew education and 2) completion of the iron roughneck maintenance to bring it into specification. Following rig crew education, slip-to-slip times improved by 5.9 s per connection. Following iron roughneck service/maintenance, slip-to-slip times improved by an additional 4.2 s. In total, a 10.1-s (or $105) per-connection improvement was realized after rig crew education and iron roughneck service/maintenance.

Run-time analysis of Well 8. The Northwood well (Well 8) was drilled immediately following the rig crew education and iron roughneck service/maintenance performed on Well 7. Well 8 was drilled with the same rig and equipment as the prior seven wells and was similar in design and depth. Data from 673 connections on Well 8 (after data cleanup) illustrate the importance of coupling rig crew education and equipment service/maintenance with new technologies. The mean slip-to-slip time for the well was 94.1 s, faster than all previous wells and 7.3 s faster than the 2G DSC time of 101.4 s. On average, about $76 was saved on each connection makeup or breakout.

Well 8 is now a benchmark for future well performance, and run-time averages of 94.1 s or better are expected. Efforts are in place toward realization of the 11-s time savings observed during laboratory testing of the 3G connection.

While it is clear that slip-to-slip time savings can be realized with 3G DSC, Chevron’s use of 5⅞-in. 3G DSC is unique to deep water. The connection is not used throughout the entire drillstring; only about 7,500–10,000 ft is used for the 33,000-ft-TD wells, less than 30% of the total string length. The connection is not used in all hole sections, and only after 13⅝-in. casing is set. As a result of these parameters, more 6⅝-in. API connections are actually made online than 5⅞-in. 3G DSC connections. Projects using longer full-length strings of 5⅞-in. 3G DSC throughout all hole sections, such as extended-reach projects, can likely realize significantly greater time savings due to the increased amount of 5⅞-in. 3G DSC in the string.

INSPECTION/REPAIR ANALYSIS

The greatest cost savings Chevron realized were from reduced repair costs with 3G DSC. In three of the four wells drilled with the new connection, reduced repair costs provide greater cost savings than slip-to-slip time savings. As of this writing, the pipe had not completed inspection on the fourth well (Well 8). Nonetheless, observed repair rates were substantially less than with 2G DSC.

The 3G design reduces thread stress during stab and reduces the amount of sliding distance during makeup, resulting in a dramatic positive impact on repair performance, based on results observed during Chevron’s use of the connection. For each pipe rental using 2G DSC, on average about 24% of the connections required recut and about 42% of the connections required refacing. This resulted in an average of over $74,000 in connection repair charges following each rental, based on one rental for three wells at Tahiti Field plus Wells 1–4. Adjusting for pipe footage and days on rent, Chevron paid about $0.0401/ft/day in connection damage charges on these wells.

Results from four rentals on Wells 5–7 using the new connection show that, on average, about 0.5% of connections require recut, with no single rental exceeding 2%, a 98% reduction in connection recut frequency. Improved running and handling practices and the connection design combined to reduce refacing rates by 82% to an average of about 7% of the connections requiring reface following rental. In total, connection repair costs per rental have decreased from about $74,000 per rental (Tahiti wells plus Wells 1–4) to about $5,300 per rental (Wells 5–7). Adjusting for pipe footage and days on rent, connection repair charges have been reduced 83% to an average of $0.0068/ft/day for Wells 5–7.

ECONOMIC ANALYSIS

Table 1 presents the overall economic impact of the new connection, assuming a $900,000/day spread rate, or $10.42 for every second saved.

TABLE 1. Economic impact of 3G

Well 5 observed about $201,000 in total savings with about $98,000 coming from time savings and about $103,000 from repair savings. Well 6 observed about $27,000 in total savings—about $5,000 from time savings and about $22,000 from repair savings. Well 7, where emphasis was made to work with the rig crews on running and handling, actually observed increased costs. An overall loss of about $20,000 was realized on the well, representing a $34,000 loss due to increased time running the pipe and a $14,000 gain on repair savings. Well 8 yielded about $79,000 in time savings alone, saving 7.3 s per connection over 1,039 connections. This well is the only well to reduce the cost per connection to less than $1,000. Though inspection and repairs are not complete on the pipe, if average repair rates from the first three wells continue, an estimated additional $20,000 may be realized in repair savings.

Over all four wells and not including expected repair savings on Well 8, Chevron has realized about $287,000 in savings using 3G DSC, excluding incremental rental rates for use of the connection.

GAS-TIGHT CONNECTION

Since the introduction of the 3G DSC, a project was completed to develop a version of the connection with the ability to seal high-pressure gas, for applications including high-pressure completions, drill stem testing, high-pressure workovers, underbalanced drilling and intervention risers. The key objective was to improve the pressure ratings of the gas-tight 2G DSC while maintaining the performance benefits of the 3G connection. The design requirements for the gas-tight 3G DSC were defined as 20,000 psi internal (5,000 psi more than the gas-tight 2G rating) and 10,000 psi external.

A radial metal-to-metal interference seal on the pin nose is similar to the type used on the gas-tight 2G DSC. This new connection incorporates ultra-high-strength tool joint material with a minimum specified yield strength of 135,000 psi to optimize the connection’s structural and pressure integrity for the most demanding drilling, completion, workover and riser applications.

With the primary seal being the 15° radial metal-to-metal seal on the pin nose, the external shoulder provides a secondary seal. The secondary shoulder internal to the connection acts as a torque stop and is not considered a pressure seal. This permits the connection to be racked back on the pin nose while tripping without concern about damaging a seal.

Interference is one of the controlling factors in the effectiveness of a radial metal-to-metal seal. Of primary importance was optimizing the seal interference to ensure adequate contact stresses at the sealing surfaces. Insufficient interference would adversely affect the connection sealability, while excessive interference could cause galling. Other design parameters, such as double-start threads, dual-radius thread form increasing connection fatigue performance, optimized taper and thread pitch, are identical to those found in the standard 3G DSC.

CONCLUSIONS

Chevron’s use of the third-generation dual-shoulder connection in the high-cost deepwater GOM environment has yielded about $287,000 in savings on the first four wells. About $147,000 was saved due to faster pipe running and tripping speeds using the new connection. An even more dramatic improvement was observed in repair rates and costs. Recut rates were reduced 98%, reface rates were reduced 82%, and overall damage costs were reduced 83%, contributing to about $140,000 in repair savings.

Key to realizing these savings was a focused effort to train rig crews on best practices to safely optimize running, handling and connection makeup times. In addition, a rigorous maintenance program was implemented to ensure pipe-handling equipment such as the iron roughneck was in proper working order.
 

ACKNOWLEDGMENTS

This article was prepared in part from IADC/SPE 128316 presented at the IADC/SPE Drilling Conference and Exhibition held in New Orleans, Feb. 2–4, 2010. The authors thank the management of Chevron and National Oilwell Varco for their support and encouragement in publishing this article.

THE AUTHORS
	Sheldon Langdon is a Drilling Engineer assigned to Chevron’s GOM Deepwater Exploration and Appraisal department. Mr. Langdon earned a master’s degree in engineering from the Memorial University of Newfoundland, Canada, in 1998. He has 11 years of deepwater drilling and completion experience—the initial five years with a major rig contractor and the past six with Chevron.