COMPLETION TECHNOLOGY: Part II - Reducing completion costs in deepwater subsea developments

PART II: This is Part II of a two-part series on reducing completion costs in deepwater. Part I appeared in July.

A key innovation pursued for the Macaroni, King, and Europa projects was to unload the wells to the flowline and host facilities, thereby eliminating any signif-icant flow of hydrocarbons to the rig after the tree is installed. Unloading deepwater subsea wells directly to host facilities presented a number of concerns across the project disciplines. Key challenges included:

Fluids and materials compatibility
Hydrate management
Requirement for a debris-free completion
Handling of the produced fluids at the host.

These issues were addressed, and unloading to the host was successfully implemented. The costs savings associated with this approach were significant, reducing rig time and the associated costs for the unloading equipment and personnel. The elimination of hydrocarbons coming to the surface also removed one of the major design requirements for a completion riser system.

Standardized tree system

Benefiting from experiences at Tahoe, Popeye, and Mensa, Shell decided that efforts to standardize their subsea tree system, including the guidebase (or tubing head spool), the subsea tree and its control system, would be worth the substantial investment. The immediately recognized goals were to: reduce capital costs, reduce engineering and manufacturing cycle times, and reduce installation times.

Initial studies developed many potential improvement areas of equipment design. Through a combination of standardization suggestions from selected suppliers and value engineering workshops with input from suppliers and internal groups, four major equipment categories were selected in which to further concentrate efforts:

Compact modular tree design with interchangeability
Concentric monobore tubing hanger system and completion riser
ROV-installable (remotely operated vehicle) and retrievable tree cap
Dependable tree workover control system.

The modular design and the interchangeability of the current subsea tree systems in the Gulf of Mexico have allowed for numerous substitutions of various pieces of the tree systems during the course of several simultaneous subsea projects, preventing costly schedule delays. Workovers on some of those wells have been easier to plan due to the fact that the completion riser/workover control system interfaces can be managed more efficiently.

Adding flying lead connections from the workover control system on the completion riser to the tree not only simplified the controls connection, but allowed the possibility of using a multiplexed (MUX) control umbilical of opportunity, instead of being limited to the particular one used during initial installation of the tree. This also simplified testing capability during integration testing at the manufacturing plant and pre-installation testing onboard the rig.

At the manufacturing plant, the time required for testing components has begun to show improvement and the time onboard the rig for testing prior to running a subsea tree has also been improved due to the standardization of the tree system components. The newly developed monobore completion riser and its workover control system are discussed below.

The ROV-installable tree cap innovation was a tremendous breakthrough, which not only saved installation time by eliminating the separate running of a tree cap, but also allowed for a more standard approach to providing the running tools for a subsea tree. The adaptation from the top profile of the subsea tree to a running tool connector at the bottom of a completion riser system could now be simpler and easily adaptable to any one of several available completion riser system designs.

Monobore completion riser

In cooperation with FMC and benefiting from earlier experience with completion riser designs, the monobore riser system that can be used in water depths up to 7,500 ft was developed. The system incorporates a state-of-the-art MUX electro-hydraulic control system and compact-designed components to allow it to fit almost any deepwater rig and/or subsea tree combination. The riser is rated at 10,000 psi working pressure for the main bore of the system, but being a monobore system means that although circulation is only possible through a 1-in. 5 ksi hose and circulation downhole against most expected wellhead shut-in pressures is not possible.

However, it does retain the necessary capability to circulate and clean out the riser above the subsea tree to remove any hydrocarbons present. Full emergency capabilities are included with its emergency-disconnect and lower workover riser packages as well as a state-of-the-art riser angle and direction monitoring system, which were designed to easily accept other riser/tree combinations when required. As water depths have increased, the importance of monitoring riser angle and any induced rotation in the riser has also increased making this a very valuable improvement.

The main connection of the monobore riser joints is a Wyman-Gordon connection, which has been thoroughly tested in other applications. The 7-in. outside diameter riser joints are connected using traditional torque-turn tools with a traditional riser spider operating in the rotary table. There were some initial problems with the connectors after the first run, but overall the system performed as designed.

Not only did the connection prove to be almost as simple to make up as a standard drillpipe connection, it proved to be almost as fast to connect. After the first several runs, the only difference between the running times, comparing drillpipe to the monobore riser, was the fact that the drillpipe joints were already in 90-ft. stands in the derrick, while the monobore riser joints (45 ft long) had to be picked up from the pipe deck.

Weather limitations and the depth below mudline of the Europa (North Sea) wells did not allow room in the derrick to "rack back" doubles of the monobore riser joints. In conclusion, the monobore system performed in world record style compared to other completion riser systems.

On Europa, we were able to land one tree and then move from tree to tree in the field for over a month, making over five connections to the subsea trees in the field, completing all of our required work without one downtime incident concerning the monobore riser. To date, no one in the world has a better record for a deepwater completion riser.

For Shell's upcoming Na Kika project (Gulf of Mexico) in water depths of over 7,000 ft, the only modifications required will be to build the extra required riser joints and make a few changes in some of the minor components to prepare for the greater water depth.

Drillpipe riser use

Although drillpipe has been used for years to land subsea trees in shallower water, its use was expanded to deepwater applications. For the Europa project, concerns over the first use of the monobore completion riser system predicated the preparation of a back-up riser system. A high degree of confidence was developed during the system integration testing (SIT) of the monobore system for the MUX control system components, but there were some remaining concerns about the use of the new Wyman-Gordon connections.

In addition, after removing the requirement for unloading subsea wells to the completion rig (thus removing most of the risk of bringing hydrocarbons to the surface after the subsea tree was installed), the functionality of a conventional completion riser was already being challenged on both the Europa and Macaroni (Gulf of Mexico) projects. As a result of the concerns over the monobore riser and the removal of the requirement of bringing hydrocarbons to the surface, a backup riser system for Europa was developed using the 5 1/2-in. drillpipe already in the derrick.

All riser analysis work showed that, although marginal, the drillpipe string would support the revised requirements at the expected water depth of 3,900 ft. For Macaroni, using drillpipe became the primary option since it would also allow the release of the Sonsub completion riser, which it had kept on a rental basis for the project. The Macaroni team took the drillpipe riser concept a step further and used a drillpipe connection slightly different than that of the drillstring on the rig to allow more internal clearance for the tubing hanger plugs. Also, the drillpipe string was internally coated to lower internal friction.

Thirdly, due to the modular tree design, the flying lead connections for the subsea tree controls allowed them to use an available control system on a rental basis. On the basis of revised operating procedures, the requirement for an emergency-disconnect package for moored operations was also challenged and eliminated by the Macaroni team. For both projects, the use of drillpipe as a completion riser saved considerable rig time and in the case of the Macaroni project, it saved several million dollars of riser system rental costs.

Technical limit philosophy

A concept of dissecting each step in the drilling process to determine if it was really necessary and/or if it could be done simultaneously with other ongoing operations is underway. This process now is referred to as the "technical limit" process, whereby each activity in drilling a well is broken down in small enough steps and assigned a specific amount of time.

Teams of personnel from varied backgrounds and from all departments, as well as outside service companies, were invited to participate in an informal atmosphere. Certain assumptions are made, such as all:

Equipment functions properly and does not fail in service
Equipment is available on time
Personnel are available as needed
Equipment represents the latest and most efficient design.

When the entire process is completed, you end up with the absolute "technical limit" as to time required to perform the activities. The goal was to establish a "perfect" operation estimate (the technical limit) and then apply any limitations that are unavoidable to predict a more accurate "engineering" estimate. Comparing these two estimates to the actual results obtained not only produces a well-defined performance measurement, it gives opportunity for new ideas from all the participants in the process to surface. Over the last several years, the process has been extremely successful in decreasing drilling times and has been repeated with comparable results in the well completion arena.

Winch-landed tree

One of the main goals of the technical limit process was to eliminate the costly round trips of the marine riser/BOP (blowout preventer) stack and the completion riser. Some other means of lowering subsea equipment, such as the tubing head spool (THS) and the subsea tree, had to be developed. The derrick or moonpool were kept occupied by both riser systems.

Since the accepted flowline pull-in operation requires a rig-mounted winch of considerable tension capacity and a wire length long enough to reach the seafloor, the idea of using this winch to lower subsea equipment was investigated. The key concerns identified with using a winch to install a subsea tree and/or tubing head spool were:

Weather sensitivity of operations to deploy the tree/THS
Proximity of the subsea tree to the suspended marine riser
Proximity of the winch cable to the ROV lifting cable/umbilical
Potential for damage to sealing surfaces and tubing hanger penetrations during subsea tree landing operations.

During the planning process, the anticipated heave at depth was determined as a function of sea-state. It was apparent that the anticipated heave would be minimal for wave heights up to 6 ft. By limiting operations to these sea conditions, the risk of damage to the subsea equipment during landing was considered minor. The unique shape of the Noble Jim Thompson MODU contributed to the low anticipated heave, and also provided a convenient location for the winch package approximately 120 ft from the center of the main derrick. This reduced concerns regarding collision between the marine riser and the tree package.

The winch package used for the Europa flowline pull-in operation was initially used for the installation of the THS on the Europa A-1 well. The operation demonstrated the feasibility of winch line installation of the equipment. The winch package and the wire were then upgraded to account for the heavier subsea tree (90,000 lb, versus 60,000 lb). Changeout to the larger winch was accomplished within 24 hours, while the rig continued other completion activities. The subsea tree was successfully landed, locked, and pressure tested from a port on the flowline hub.

On the King Project, the winch package also was successfully used to install the THS. In addition, the winch package was utilized to successfully install the 80-ft well jumper that connects the tubing head spool to the flowline sled. This operation was performed totally offline from the rig's other critical path activities.

Conclusions

Through the technical limit process, strategic long term planning and from the experiences of completing more deepwater subsea wells in the Gulf of Mexico, a solid base of knowledge and expertise to enhance the profitability of future deepwater subsea well completions has been built.

Simpler and more efficient tree handling equipment, more efficient MODU's, more efficient and more reliable completion risers/workover control systems, coupled with a "standard" tree system and a concentrated effort to reduce the time spent on each activity in the completion process has resulted in faster completion times in the Gulf of Mexico.

Editor's Note: This an edited and updated version of OTC 13122, delivered at the Offshore Technology Conference (April 30 - May 3, 2001).