Copyright 2005, Society of Petroleum Engineers This paper was prepared for presentation at Offshore Europe 2005 held in Aberdeen, Scotland, U.K., 69 September 2005. This paper was selected for presentation by an SPE Program Committee following review of information contained in a proposal submitted by the author(s). Contents of the paper, as presented, have not been reviewed by the Society of Petroleum Engineers and are subject to correction by the author(s). The material, as presented, does not necessarily reflect any position of the Society of Petroleum Engineers, its officers, or members. Papers presented at SPE meetings are subject to publication review by Editorial Committees of the Society of Petroleum Engineers. Electronic reproduction, distribution, or storage of any part of this paper for commercial purposes without the written consent of the Society of Petroleum Engineers is prohibited. Permission to reproduce in print is restricted to a proposal of not more than 300 words; illustrations may not be copied. The proposal must contain conspicuous acknowledgment of where and by whom the paper was presented. Write Librarian, SPE, P.O. Box 833836, Richardson, TX 75083-3836, U.S.A., fax 01-972-952-9435. Abstract This paper describes the worlds first true coiled tubing under balanced casing exit using a Nitrogen (N2) gas mist as the milling fluid, which allowed a well in Sharjahs Sajaa field to remain live throughout the casing exit, ensuring uninterrupted gas production. Previous attempts to mill a window using a N2 gas mist with conventional milling technology proved unsuccessful. This was largely due to the lack of cooling and lubricity of the milling fluid resulting in a rapid deterioration of the window mills. Subsequent windows were milled using water only which resulted in massive fluid losses to the formation, due to the low reservoir pressures within the field, and more importantly caused interruptions to existing gas production. The technology described in this paper allowed gas production to be maintained throughout the window milling operation and resulted in a dramatic reduction in fluid losses to the formation and a reduction in overall casing exit time. Following the casing exit, the build and lateral drilling assemblies were deployed through the window at the first attempt, allowing uninterrupted progression to the under balanced drilling phase. This technology has helped to extend the life of the Sajaa field and secure the future of the Sharjah Coiled Tubing Drilling (CTD) project. Introduction The Sajaa onshore gas field is located in the United Arab Emirate of Sharjah and is operated by BP Sharjah and their partners. It is a low pressure (~1000psi) high temperature (300 F) reservoir, producing gas condensate from the Thamama limestone formation ranging from 11,000 to 13,000 feet true vertical depth. Gas production from the field was in rapid decline.

Under balanced (UB) CTD was identified as the clear recovery solution and in April 2003 the Sharjah UB CTD campaign commenced. The objective being to sidetrack from the existing production casing and drill multi-lateral well bores into the reservoir to increase recovery rates. To date, 20 wells in the field have been sidetracked and over 230,000 feet of new well bore drilled, the majority being drilled under balanced. The gas produced from the Sajaa field is exported to the nearby Sajaa Gas Plant which generates electricity for the Emirate of Sharjah. For this reason it is critical that interruptions to the gas supply are minimized during the CTD operation. All drilling is performed using a N2 gas mist of approximately 1500 cubic feet per minute of N2 and 8 to 12 gallons per minute of water. Early attempts to mill the casing exit in a similar fashion using conventional window milling technology proved unsuccessful, largely due to the lack of cooling and lubricity from the milling fluid resulting in deterioration of the natural diamond window mill and an extremely low rate of penetration. This approach was abandoned and subsequent window milling operations were performed with water only resulting in a high volume of fluid loss to the formation due to the overbalanced condition created, interruptions to gas production and temporary production impairment. However, the risk to high rate gas producing wells was considered too great to continue in this manner and a gas milling solution was ultimately required to extend the scope of the CTD campaign. Phase 1 Qualification Testing Baker Oil Tools in Aberdeen was selected to provide the fishing, milling and whipstock services along with their sister division of Baker Hughes Inteq providing the directional drilling services. In February of 2003, several months prior to the start up of the CTD project, both service providers in conjunction with BP embarked on a program of bottom hole assembly (BHA) qualification and assurance testing. This testing was performed at the service companies joint R&D facility in Celle, Germany. The test program objectives included:

SPE 96282

World's First Coiled-Tubing Underbalanced Casing Exit Using Nitrogen Gas Mist as the Milling Fluid B. Webster and M. Pitman, Baker Oil Tools, and R. Pruitt, SPE, BP

2 SPE 96282

Pre-qualification of electrical drilling assembly; Elastomer testing at simulated BHP & BHT, Vibration testing, Compatibility testing and interface with milling assemblies, Pre-qualification of milling assemblies; Nipple milling performance with N2 gas mist, Window milling performance with N2 gas mist, Determine optimum operating parameters, Evaluate performance of Air Driven Motor (ADM) at simulated BHP, Establish motor stall indications. In order to conduct the milling trials under simulated BHP conditions a purpose built hyperbaric test fixture was constructed. The fixture also had to be large enough to house the entire BHA as it would be deployed in the field in order to allow for a full hook up of the electrical drilling assembly. The test spread also included a coiled tubing reel to accurately simulate surface pump pressures and a test separator package to control and manage the high pressure gas downstream of the test fixture. Weight on bit during the milling operations was provided by a hydraulic ram incorporated into the lower end of the test fixture. The BHA for all tests included the following standard components; Power and Communication Sub (P&C), this provides the power to the electrical components within the assembly and also communicates the data from the BHA to the surface data platform via the electric wireline, Drilling Performance Sub (DPS), this provides real time downhole measurements including bore pressure, annulus pressure, weight on bit and downhole vibration, Directional Gamma Sub (DGS), this sub contains the gamma ray sensor as well as providing real time MWD measurements such as azimuth, inclination and toolface, Air Driven workover motor (ADM). A natural diamond window mill was used for the window milling trials and a multi-step carbide insert mill was used for the nipple profile milling trial. Phase 1 test results. The directional drilling surface data acquisition system was utilized throughout the testing to provide real time data monitoring. This proved to be a hugely beneficial tool throughout the testing as it allowed for variables such as weight on bit to be accurately monitored, controlled and adjusted helping to establish optimum milling parameters. One of the objectives of these trials was to establish the most effective means of detecting motor differential pressure and stalls on surface. When milling in a liquid environment motor differential is indicated almost immediately at surface as an increase in pump pressure or circulating pressure. However, when milling with gas the surface indications are far less obvious because, due to the compressibility of the fluid, small pressure changes downhole are not seen at surface. This can be extremely time consuming and also result in accelerated motor wear or damage. The real time measurements provided by the Drilling Performance Sub proved to be invaluable in this respect, providing the only real indications of motor stalls. Motor differential pressure could be monitored via the bore and annular pressure sensors, however, the most obvious means of

detecting a motor stall proved to be through monitoring toolstring vibration. The design of the positive displacement motor is such that vibration is created while the motor is turning; however, when the motor is stalled the vibration will reduce dramatically. This sudden change in BHA vibration can be detected immediately at surface.

The overall performance of the 2 7/8 motors used during the testing was extremely positive. Very few motor stalls were experienced indicating that the motors were delivering sufficient power. The pump rates used during the testing ranged from 1400 to 2400 cubic feet of N2 and 10 to 20 gallons per minute of water. An interesting observation was made while monitoring motor speed. At lower flow rates the motor speed would fluctuate constantly due to the compressibility of the gas as the motor load increased, reducing the effective volumetric flow through the motor and also as a result of a slight slugging of water at low volume. However, this was less evident at higher flow rates of around 1800 cubic feet per minute of N2 and 15 gallons per minute of water.

The performance of the multi step carbide mill used for the nipple profile milling was extremely good. The motor performed well with no motor stalls encountered and the nipple profile was milled out in approximately 1 hours. The mill was in good condition with minimal damage and no gauge wear. Two separate tests were performed with the diamond window mill to simulate the pinch point at the top of the window and also the mid point in the window. The motor performed consistently well throughout the test with very few motor stalls and the window mill showed minimal signs of wear, however, the rate of penetration (ROP) was slower than expected. It was concluded that the lower than average ROP was a result of the slower rotary speed produced by the air driven motor. The average motor speed during these tests was around 250 to 300 rpm, whereas, the diamond window mill is typically operated at 400 to 500 rpm and performs more efficiently at higher rotary speeds.

Early Field Experience As with many projects of this type a great deal was learnt during the first 2 or 3 wells. It was only during this early phase of the project that many of the technical challenges of working in this extremely difficult environment were highlighted. The standard completion configuration for most of the Sajaa wells consists of a 5 packer-less production tubing string from surface to a depth of approximately 11,000ft, and a 7 production liner cemented across the reservoir. Some of the higher rate wells are completed with only 1 or 2 joints of 5 tubing at surface and 7 casing to TD. A large percentage of the completion strings also contain landing nipples, ranging from 3.187 internal diameter. All nipple profiles need to be enlarged to a minimum ID of 3.833 to allow for the creation of a 3.800 casing window. The casing exits are generally executed from within the 7 liner with the use of a high expansion Through Tubing Whipstock (TTW) system. All sidetracks are typically performed within the reservoir with the whipstock being set within existing perforations and milling operations performed

SPE 96282 3

under live well conditions. The majority of the original wells were drilled vertically through the reservoir. One of the most significant issues highlighted during the early phases of the project was downhole vibration. The severity of this became evident during the initial nipple milling operations. Although these operations were completed successfully, multiple milling runs were often required due to the level of damage caused to the mills and the high mortality rate of the carbide cutters. In addition to the increased mill damage the electronics within the direction drilling components of the BHA also suffered an increased failure rate, again resulting in multiple trips for BHA changes. Around the same time as the start of the Sharjah CTD project the service company had just completed an extensive test program for the development of a window milling system for Duplex 25% Chrome casing. This offered the opportunity to transfer some of the learning from this development project to address the problems being experienced in Sharjah. The Multi-Step nipple profile mills were subsequently redesigned and a new generation cutter with greater impact and wear resistance was incorporated. Further testing was performed at the service companies on site test facility in Aberdeen. The new nipple mills delivered an increased level of longevity and on the second well 2 nipple profiles were successfully milled out with a single mill with minimal damage sustained. This operation was performed underbalanced with a N2 gas mist at an average rate of 1500 cubic feet per minute of N2 and 15 gallons per minute of water. It was in the same well that the first underbalanced, gas mist window milling operation was attempted. The milling performance during this operation was less encouraging. Although the window was completed successfully, numerous milling runs were required to replace the window mill and motor as a result of very slow ROP, and also to change out the components of the drilling assembly due to electronic failures. Approximately 5ft of window was milled with N2 gas, in 3 separate milling runs, before switching to water only to complete the remainder of the window and rathole. On inspection of the diamond window mills it was concluded that the natural diamonds had suffered increased thermal degradation due to a lack of cooling, resulting in reduced milling efficiency. The downhole motor appeared to perform efficiently throughout and was selected primarily as it provided a slightly higher rotary speed than the motor used during the phase 1 trials. However, this increase in rotary speed also significantly increases the rate of heat generation at the mill. It was concluded that this particular style of natural diamond mill was not suited to operating in this environment

Subsequent casing exits were milled with a more conventional approach using water only. This approach was far from ideal as it presented the operator with a number of concerns. A huge volume of water (in excess of 5000 bbls) was introduced to the wellbore during milling operations, resulting in an interruption to existing gas production and impairment to future production.

Additionally, a conventional coiled tubing work reel was used for the milling operations which resulted in a significant increase in operating time swapping between the work reel and e-line drilling reel. This also presented an HSE concern due to the increased activities involved, i.e. additional crane operations, pipe stabbing, pressure testing, BHA handling, etc.

Technology Development The challenge was to deliver a reliable gas driven window milling solution which would allow the scope of the CTD project to be extended to the high rate producing wells in the field. There was already a reasonable level of industry experience with 2 phase or gas mist milling and drilling applications, however, a casing exit in this environment had never been successfully achieved. A team of service industry experts from within the service companies various divisions were brought together in Aberdeen to review the existing data and results from previous test programs and field experiences. A clear goal was established and a development program agreed. The development process was controlled and managed by the service companies firmly established Product Development Management (PDM) system. The PDM system is an integrated approach to get new products or technologies from concept to customer. PDM is a comprehensive management technique that is common across all divisions of the service company. The process ensures that goals and deliverables are clearly established and defined and resources are aligned in the most productive manner for technology and product development. This process provides the structure for cross-divisional developments such as these to be efficiently and successfully managed. Window Mill Design. The first step was to review existing and established designs of window mills, fishing service mills and drill bits to assess whether or not existing technology could be refined to suit this application. Several new mill designs were also produced. By the end of this process a total of 6 prototype window mill designs were produced and manufactured for testing. In order to succeed in this harsh environment and deliver consistently reliable performance the final window design had to meet some very specific design criteria:

High wear and impact resistance to minimize cutter damage and increase longevity.

Minimize BHA vibration to reduce impact damage.

Minimize torque requirement for use with a downhole motor.

Fundamental requirements for cooling and lubrication had to be a primary consideration.

Compatibility with BHA and whipstock system. Sufficient longevity and durability to complete the

window in a single trip. Acceptable rate of penetration.

All six prototype mills were tested at the service companies purpose built test facility in Aberdeen, UK and test rig facility in Louisiana, USA. Initial milling trials were performed with water

4 SPE 96282

to assess the overall milling performance and compatibility with the whipstock system. Following these tests two mill designs were selected for the next stage of testing. Full system tests were then performed at the test rig facility in Louisiana with the two final mill designs. The first was a Polycrystalline Diamond Compact (PDC) based Formation Mill (Figure 1) which was the result of extensive collaboration between the service company and its sister drill bit division. The Formation Mill consisted of a new generation of high performance thermally stable PDC cutters combined with Carbide inserts and was designed to efficiently cut both steel and formation. The second was a full Carbide dressed mill of a similar profile to the Formation Mill.

Both mills delivered a high level of performance during testing, demonstrating a satisfactory level of ROP and good durability, completing the window in a single trip with minimal wear or damage sustained. Due to logistical and HSE reasons these tests were also conducted with water. Conducting further gas milling trials would have resulted in a considerable delay to delivery of the solution; therefore, the operator agreed that the system would effectively be field trialed in an appropriate candidate well. The PDC Formation Mill was selected as the primary mill for the system as this was determined to be the less aggressive of the two mill solutions, producing less vibration than the Carbide mill and delivering a higher level of performance under these wellbore conditions due to the thermally stable PDC cutters. Downhole Motor Selection. Sufficient experience of milling and drilling with N2 gas had already been gained through prior testing and also throughout the initial phases of the project to determine that a satisfactory level of performance could be achieved from a positive displacement motor (PDM) in this underbalanced environment. The PDMs used during the drilling phase of the project had been refined to the point where they were delivering exceptional performance and reliability. Therefore, it was concluded that a PDM would be the most logical choice of motor to continue with for the window milling. The only question was which PDM to select to deliver the performance required for milling the window. As well as performance the motor also needed to have sufficient strength and durability to withstand the bending forces and vibration present during window milling. It is also critical that underbalanced flow modeling is performed to establish the effective volumetric flow through the motor at downhole conditions. The N2 gas flow rate needs to be optimized to ensure that all of the water pumped is effectively lifted from the well while also ensuring that the motors designed operating parameters are not exceeded. Figure 2 illustrates the motor performance modeling software for the PDM used. This modeling helps to determine the safe operating parameters for the motor in relation to the bottomhole conditions and flow characteristics; calculating the effective volumetric flow through the motor, the bit speed and available motor torque. The motor selected for this application was a high torque, low speed air driven motor with equidistant power section technology. Figure 3 shows a cross section of the equidistant motor stator housing. This motor was selected primarily for its

high torque output, compared to a conventional PDM workover motor, and its high strength construction. Additional information on this technology can be referenced in SPE paper 74825; Latest Technology Equidistant Power Section Increases Overall Performance of a Workover Motor. BHA Considerations. Previous window milling operations had been performed on both an electric line coiled tubing reel, using the electrical directional drilling assembly, and a conventional coiled tubing work reel. The use of a conventional work reel for milling operations introduces additional operating time, swapping between the e-line reel and work reel between whipstock setting, window milling and drilling operations. Eliminating the swapping of reels would result in a considerable time saving. The goal was to run the window milling assembly on the e-line reel but without the use of any of the electrical components of the directional drilling BHA. This meant that a suitable emergency release tool had to be identified as the electrical disconnect would not be available and a conventional ball operated disconnect could not be used due to the wire inside the coil. A straight pull shear type release tool was not considered due to the extreme vibration known to be present at the BHA during milling. A number of alternative release tool designs were considered and tested at the test facility in Aberdeen. The final solution was a hydro-mechanical release tool activated by a combination of flow rate and applied tension; both of which need to be applied simultaneously. After further testing and qualification the tool was successfully integrated into the window milling assembly. A rupture disc was later incorporated below the release tool to equalize pressure across the tool during activation. Field Performance In June of 2004, the casing exit system was deployed in the field for the first time in Sajaa well S19. This was a relatively low rate well and had been shut in prior to the CTD intervention. This well was selected specifically for this application in order to minimize the potential risk to production, as this was effectively a field trial of a prototype system. All operations were performed using the electric line coiled tubing reel. The whipstock was deployed using the directional drilling assembly, providing accurate depth control and toolface orientation. The whipstock was set using N2 gas only with the use of an adapted hydraulic setting tool. The top of the whipstock was set at a depth of 11,424ft MD with a face orientation of 150 degrees left of high side. Deviation at the KOP was approximately 4 degrees. Due to the high level of electrical tool failures experienced during milling operations none of the electrical components of the drilling BHA were used in the window milling assembly. The electric wireline was anchored and terminated within a blind sub at the top of the BHA. In order to minimize vibration and prolong mill life the window mill was run directly below the motor with no string reamers or flex joints in the assembly. As this was the first ever deployment of this system a cautious approach was adopted. The initial period of window milling was also complicated slightly by erratic weight indicator readings caused by the stripper rubber binding on the dry coil. It is also critical to ensure that all well conditions are as stable as possible

SPE 96282 5

prior to starting milling. A great deal of patience is required at this stage as it can take up to 2 hours for the rates and pressures to stabilize. Once all parameters had stabilized the milling commenced and the window was time milled at an average rate of 1 ft/hr with an average circulating rate of 2000 cubic feet per minute of N2 and 20 gallons per minute of water. The window and rathole were completed in a single run with no problems experienced and no motor stalls throughout. The window mill was pulled out of hole in extremely good condition with only a few damaged inserts and approximately 0.035 of gauge wear. The window was milled through 7 35 lb/ft AC-95 casing. The build and lateral assemblies were successfully deployed through the window at the first attempt and the drilling program completed as planned.

The total operating time for the casing exit phase was reduced by 60% against the previous average for this phase of the operation. This time saving was achieved primarily through utilizing the wireline CT reel and eliminating time consuming reel swaps. The total volume of water lost to the formation was 175 bbls, compared to over 5000 bbls on previous operations; a reduction of almost 97%. On completion of the casing exit operation produced gas was being exported to the gas plant from a well that was previously dead. Performance Improvement Since this initial application a further 3 wells have been successfully completed using this technology. However, performance issues have been experienced with the whipstock system. The development project focused purely on the BHA and window milling technology. No changes were made to the standard whipstock system. The use of a more aggressive style of window mill and a high powered motor, combined with excessive levels of vibration experienced in this gas environment was concluded to be the primary cause of the whipstock problems experienced. The Through Tubing Whipstock system has since been redesigned with considerable performance enhancement to improve the overall stability and anchoring capability, and provide a more stable platform for these extremely challenging window milling operations. The PDC window mill has also been redesigned to further reduce the overall aggressiveness of the mill and to minimize the vibration. Conclusions This unique milling system breakthrough is the result of extensive collaboration between Baker Oil Tools, Hughes Christensen, Baker Hughes INTEQ, BP and their partners in Sharjah, UAE. The cross-divisional cooperation and ability to transfer the knowledge and experience from within the service company organization proved to be the framework for the successful development and delivery of this solution. The teamwork demonstrated throughout this development project and throughout the Sharjah CTD project in general has undoubtedly played a significant role in realizing this goal. Another major factor in successfully delivering this solution was the performance and attitude of the field supervisors. Having a

dedicated team of field personnel assigned to the project since start-up has ensured that the supervisors responsible for execution of the operation have an intimate knowledge and understanding of the procedures and processes involved and, most importantly, of the unique problems experienced and the issues around working in this challenging environment.

The successful implementation of this technology has enabled the future development of the field by allowing the high rate production wells to be sidetracked with minimal interruption to gas production. Overall the Sharjah UB CTD project has proven to be a huge economical success for BP. The overall impact of UB CTD on field production has been substantial with the production decline having been reversed. The table in Figure 4 demonstrates the additional incremental gas production immediately after CTD. Another notable highlight of the project is the exceptional HS&E performance; with over 800 days worked on the project without a days away from work case. Figure 5 shows a typical Sajaa well site rig-up. Acknowledgements The service company would like to thank the government of Sharjah and BP for the opportunity to be a part of this hugely successful, groundbreaking project and for being given the scope and opportunity to develop this technology and present this paper. We would also like to commend BP for their continued support of CTD technologies and willingness to encourage the development and implementation of new technology within the scope of the Sharjah UB CTD project. Additional information on the Sharjah CTD project can be referenced in the following SPE papers: SPE/IADC 87146 Basis of Design for Coiled Tubing Underbalanced Through-Tubing Drilling in the Sajaa Field. P.V. Suryanarayana, SPE, Blade Energy Partners, Bruce Smith, SPE, BP Sharjah, ABM Hasan, SPE, Blade Energy Partners, Charlie Leslie, BP Sharjah, Richard Buchanan, BP Sharjah, and Randy Pruitt, BP Sharjah SPE 89644 Sajaa Underbalance Coiled Tubing Drilling Putting It All Together Randal Pruitt, SPE BP, Charlie Leslie, SPE BP, Bruce Smith, SPE WUU, Jeff Knight (Blade) SPE, Richard Buchanan, SPE BP

6 SPE 96282

Figure 1, PDC Formation Mill

SPE 96282 7

Figure 2, Underbalanced motor performance modeling

Figure 3, Cross section of conventional PDM stator and equidistant stator

8 SPE 96282

Figure 4, Gas production before and after CTD

Figure 5, Typical Sajaa rig site

Before & After Well Gas Rate Performance

0

10

20

30

40

50

60

Sajaa

3

Sajaa

15

Sajaa

7

Sajaa

16

Sajaa

4

Sajaa

32

Sajaa

17

Sajaa

42

Sajaa

5

Sajaa

1

Sajaa

14

Sajaa

21

Sajaa

15

Sajaa

26

Sajaa

19

Sajaa

07

Sajaa

28

Sajaa

03

Sajaa

06

Sajaa

40

Sajaa

41

Sajaa

16

Sajaa

17

Sajaa

12

Sajaa

14

Sajaa

18

Sajaa

4

Gas

Rat

e, M

MC

FD

Before CTD Initial Rate from CTD