As the exploration and development of offshore energy reserves moves into ever deeper waters, the use of Dynamically Positioned (DP) vessels to undertake drilling and/or completion work is being considered as an alternative to the more traditional methods. The use of DP vessels is perceived as introducing the additional risk associated with the potential loss of station keeping during drilling or completion operations. At least one major operator is now assessing this risk and the corresponding consequences as part of the decision making process.This paper illustrates a recent study that considered a variety of deepwater completion operations by examining the following key features of the project. • The basis for the risk analysis.• The collection of input data for the model, including the contribution of expert opinion provided by Completion Engineers and DP experts. • The methodology adopted to evaluate the risk costs associated with individual operations. • The interpretation of the model outputs.Finally, examples of points at which the model may be of use within the decision making process throughout the deepwater development cycle are provided.

When a producing well fails, the engineering team needs to diagnose the failure quickly, in order to determine the best course of remedial action. This is particularly important with an early-life failure when other wells are still being installed using the same design, raising legitimate concern that the failure may be repeated in those other wells. In practice, diagnosis of a failure when evidence is incomplete, and even inconsistent, can lead to conflicting or incorrect conclusions. A systematic method of root cause analysis forces the discussion to converge on a useful outcome. This paper illustrates how such a method can be applied, using a fault-tree structure, to analyze hypothetical failures in two subsea multizone intelligent completions. The objectives of the analysis is to -understand failures as fully as possible;determine what steps are needed to prevent failures from recurring on subsequent installations;determine what other tests may be useful to help understand the problem. The work starts by collecting and reviewing the evidence (written reports, and links to knowledgeable people), and defining the symptoms of failure. Then Fault Trees are constructed from all the potential failure modes that fit the symptoms of failure (in this example, 32 in total). These are then used in a structured workshop session to review all those potential failure modes. Those that do not fit the evidence are discarded, in this example leaving just 4 Possible Causes by a process of elimination. The detailed review of processes and controls required to understand the circumstances of failure is a useful way of identifying improvements to Engineering Processes (documentation control, specification and engineering management) in addition to design and operational issues. Extra tests are then identified to confirm each Possible Cause, providing further elimination. A structured comparison to find analogues in other similar projects can provide further insight. Finally, further characterisation tests and other actions should be identified to help avoid future failures. Specific recommended actions (in this example, 32 in total) are identified to address Possible Causes and so prevent failures from recurring on subsequent installations. A FMECA constructed during the well design phase can be consulted to determine any correlation between the failure modes identified therein and the observed failures in the two zones. The conclusions from the FMECA can also be examined to identify whether there are any outstanding recommendations or remedial actions. This study shows that a systematic application of root cause analysis, using fault trees as the underlying structure, can be very effective in focusing engineering effort on resolving the important issues and avoiding time-wasting distractions. 1. Background - Case Study This case study assumes a subsea development located in deepwater Gulf of Mexico using wells which are intelligent, stacked frac pack producers. Each intelligent well has two "on/off" Downhole Flow Control (DHFC) Valves. This case study assumes the failure of one of those DHFC Valves.

DownHole Flow Control (DHFC) is a technology that was originally developed to remotely control the production from multiple reservoirs in the same well. BP recognized the value of using this capability in injection wells to manage the injection of water into separate reservoirs. BP's Field of the Future program worked with service providers to develop the capability of "choking downhole flow-control for water injectors" that allows one well to serve the role of two injection wells thereby reducing the number of injection wells required for field development. At Intelligent Energy 2008, BP reported on development and qualification of DHFC technology for high rate water injection (SPE Paper No. 112143). Since then, BP has implemented the newly developed DHFC technology in two fields with over a dozen installations. BP has achieved its goals in one field and not achieved its goals in another field. This paper reviews the key learnings associated with both the successes and the failures. A summary of key learnings: The DHFC technology is very reliable, although difficulties have been experienced with some of the associated multizone completion technologies.Remotely operating the valves to vary the injection rate has proven to be a straightforward operation.Higher than expected injection pressure in excess of the maximum differential pressure rating of the open hole isolation packer in one field is preventing the DHFC system from achieving its objective. In summary, DHFC application for high rate water injectors is considered a success. Development engineering continues on open hole isolation packers to allow deployment of DHFC in fields with high injection pressure requirements.

Consistent with the concerns of many deepwater developments being contemplated throughout the world, this paper reviews the challenges, methodologies and detailed conclusions associated with the implementation of selective water injection for pressure support in ultra-deepwater, subsea developments operated by BP. The potential benefits associated with remote control of selective water injection in any of 2 – 6 zones within a single well, include reducing well count, reducing interventions and increasing recovery. This paper focuses on the operational requirements and considerations associated with the development of soundly engineered intelligent well systems, capable of a 15-year design life, with sustained injection of 45,000 BWPD per well, into any remotely selected zone. Principal areas of discussion include: mitigation of potential erosion at these rates through 3–1/2" tubing flow control valves within 7" liner completions; inter-zonal isolation competency with differential pressures up to 7,500 psi; and downhole hydraulic systems design for remote-control at hydrostatic pressures of 18,000 psi and differential pressures up to 12,000 psi. In addition to the downhole extremes in flow rates and pressures, we discuss the considerations within the subsea control systems of these wells; the assessment of various downhole hydraulic control architectures; and the effects of the operating temperature extremes. The principles, processes and ‘lessons learned’ in design of this system can serve as a benchmark for other high-value, deepwater, subsea injection applications in which the reservoir barrier conditions are not well understood, and hence considered candidates for intelligent well systems. Introduction Many deepwater oil projects around the world are planning to rely on water injection for pressure support and maximization of oil recovery. As depths and pressures increase, the requirements placed on the downhole systems increase, in some projects beyond currently available equipment. The development and qualification of water injection well completions technology required to meet these technically challenging requirements is the central focus of this paper. Many if not most significant deepwater geologies are comprised of multiple stacked reservoirs. In many of the reservoirs there are multiple sand layers separated by significant shales. With reservoir sections hundreds of feet thick, and waterflood the main depletion mechanism, control of injection water placement becomes essential to prevent early water breakthrough and to achieve effective oil sweep and recovery. This task is complicated by variation in permeability and thickness among the pay sections. Downhole flow control (DHFC) enables controlled water placement. The ability to toggle between zones is also seen as a valuable optimization feature of downhole flow control. When water injection wells are completed with subsea trees, interventions are very expensive. A simple reduction in number of interventions to change injection intervals generates significant value. The elimination of downtime while planning interventions, waiting on rigs, etc. accelerates production, adding further value. Stranded reserves resulting from uneconomic workovers can also be eliminated by the use of downhole flow control. In recognition of the significant value potential, Baker Oil Tools and BP's Exploration and Production Technology Group have been working on the development of DHFC technology to meet the requirements of some deepwater projects for which currently available equipment is simply not sufficient.

This paper addresses the techniques developed for design and installation of long interval, high angle gravel-packed completions. These techniques were used for single and multi-zone completions installed during eight rig-years of operations on Exxon's Lena guyed tower, located in the Mississippi Canyon 281 field. The paper also discusses the productivity and life of the different completion types. A major part of the discussion will include the successful application of prepacking and water-packing techniques in the long, highly deviated completion intervals (up to 445 feet at angles approaching 70 degrees) typical at Lena, While some operators consider this technology antiquated, field experience at Lena demonstrates that water-packing has resulted in high performance completions in challenging completion environments. INTRODUCTION Exxon's Lena guyed tower is located in the Gulf of Mexico approximately 50 miles southeast of Grand Louisiana, in 1000 feet of water. The platform was installed in 1982 at Mississippi Canyon Block 280 for the development of Blocks 280, 281, and 360. Development drilling from 1983 to 1989 resulted in 93 completions in 56 wellbores. The completion intervals are characterized by fine-grain, unconsolidated sands that are finely interbedded with shale. These Pliocene reservoirs, which have limited areal extent, dip at 20 to 40 degrees from a piercement salt dome. Radial faults, which dissect the reservoirs, add further complexity to the production strategy. Field characteristics and the resulting development strategies necessitated drilling aggressive S-turn wells to penetrate stacked targets and build-and-hold wells to reach objectives located up to 11,500 feet horizontally from the platform. Wellbore deviations averaged 45 degrees, with some approaching 80 degrees. Due to high incident angles at reservoir contacts and thick sand packages, gravel-packs were frequently installed across long completion intervals, some exceeding 400 feet. Lena well depths ranged from 7,400 to 15,300 feet measured depth and 7,000 to 12,000 feet true vertical depth. Wells were typically designed with 7 inch casing for single completions, 7-5/8 inch casing for dual completions, and 9-5/8 inch production casing with a 7 inch liner for wells requiring protective casing due to high torque and drag or differential sticking conditions. Each well was completed with the drilling rig immediately following setting of production casing. BASIS FOR GRAVEL PLACEMENT METHOD From the beginning of the program, the completion designs were based on Exxon's previous operating experience and gravel-pack modeling performed in the mid-1970s1. Subsequent research2 conducted with a more elaborate model reinforced the design basis and permitted a more detailed investigation of gravel placement. This research was conducted using a 7 inch by 2-3/8 inch gravel-pack completion model that was 25 feet long and contained up to 12 perforations per foot. The model was fabricated from clear plastic so that visual observation of the gravel-packing process could be made. The assembly could be rotated on its stand, which allowed simulation of wellbore deviation up to 110 degrees from vertical. Over 200 full-scale gravel-pack tests were performed with varying well deviations, pump rates, gravel concentrations and fluid viscosities.