Nigel Bonnett, John Dalton; Chevron
Peter Van Dijk, Baker Hughes Inteq
The Captain field in the North Sea UK sector (owned 85% Chevron as operator, 15% Korea Captain Co. Ltd) produces heavy oil from its main reservoir body at 2,850 ft TVD subsea by the means of horizontal wells and electrical submersible pumps (ESP). The field has reached maturity, and infill drilling is needed to target the remaining pockets of oil.
The Captain platform has 28 slots and its horizontal wells cover an area that can best be described as half a circle from north of the platform turning over east to south of the platform. Due to the relatively shallow reservoir the wells are not far apart which makes infill drilling challenging.
The UP17 project discussed here is the most challenging to date.
Design history
The targets for the UP17 project were immediately below the platform with a slight easterly offset. The alignment of the targets was from south to north, which was also the preferred orientation. Several months of planning resulted in a well plan that consisted of a build and turn to horizontal based on a 10° per 100 ft dogleg severity (DLS). This plan stayed inside the footprint of all wells and crossed 13 slots with all its associated anti-collision risks. Additionally, the ESP had to be redesigned in order to be able to set it at the required depth without exceeding the bending limits for the pump.
On the day before a Chevron internal peer review for the project, another option came to light.
With the same set of targets but with its orientation from north to south, the UP17 well could be drilled from another slot. The existing well (C28z) associated with that slot was shut-in due to a sand screen failure. All that was needed was to pull a section of gravel-packed screens, drill a 206° flat turn with an 8.5° per 100 ft DLS, avoid collision with well C12, a field major injector, and run stand-alone sand screens around the flat curve and into the horizontal reservoir section.
Initially a hole size of 9.0-in. was assumed for running the 5 1/2-in. sand screens. However, drag modeling sensitivities covering 9 1⁄4-in., 8 3⁄4-in., and 8 1⁄2-in. hole, showed that 8 1⁄2-in. hole size was optimal for screen deployment.
Plan view showing UP17 profile and existing Captain wells.
Essentially there were three critical phases in the construction of well UP17: slot recovery, drilling, and the lower completion (5 1⁄2-in. sand screens). The upper ESP completion was a standard Captain platform design.
Planning challenges
Slot recovery:The 9 5⁄8-in. casing shoe was set into the Upper Captain Sand reservoir at 92° inclination. Due to concerns with a 900 ft of 8-10° per 100 ft DLS at the base of the 9 5⁄8-in. casing, a whipstock exit was considered to be a contingency plan. The primary plan consisted of pulling a section of gravel-packed screens, leaving a minimum 100 ft of open hole. This was the first time that a gravel-packed section of screens were planned to be recovered and consequently, there was a lot of internal management focus on this. After cutting the screens, it was planned to use an Andergauge Agitator tool to vibrate the screens in order to loosen up the sand and break its friction. A 400-ft abandonment/kick-off cement plug was then required to be set in 92° hole.
Drilling:The trajectory of well UP17 required drilling a flat turn at horizontal through 206° of azimuth at a rate of 8.5° per 100 ft DLS. Since the lower sand screen completion required an 8 1⁄2-in. hole size, this immediately presented a challenge as the required rate of turn exceeded the technical specification of the 6 3⁄4-in. AutoTrak Rotary Closed Loop System (RCLS) drilling assembly. The solution was to utilize the 4 3⁄4-in. AutoTrak RCLS which has a design capability of 10° per 100 ft DLS and drill the section in 6 1⁄8-in. hole size which would require enlarging to 8 1⁄2-in. to accommodate the lower completion.
To precisely monitor the DLS and drilling vibration in real time, the CoPilot system was selected. This system measures and transmits to surface in real time bending moments experienced by the drilling assembly and, in doing so, can monitor the DLS being achieved as the well is drilled. Advanced sensors and data processing packaged in the CoPilot tool also provide a comprehensive view of real-time vibration and stresses in the downhole assembly to optimize overall drilling performance. It also assists in monitoring real-time equivalent circulating densities at the bit.
Selection of the bit size (6 1⁄8-in.) run on the 4 3⁄4-in. AutoTrak tool complete with Hole-Opener assembly was based on providing compatibility with the fallback pilot hole AutoTrak Extreme assembly, which would be used if the 4 3⁄4-in. AutoTrak x Hole-Opener assembly failed to deliver. A larger bit size (6 1⁄2-in. or 6 3⁄4-in.) would have helped both prime and fallback assemblies in generation of the DLS, however the largest-sized bearing house stabilizer on the positive displacement motor for the AutoTrak Extreme assembly was 6 1⁄8-in. This essentially drove the bit size to be run on the 4 3⁄4-in. drilling tools.
Enlargement of the 6 1⁄8-in. hole could be done in a separate run after drilling or simultaneously while drilling. Separate hole opening would require the additional time of picking up (and laying down) 3 1⁄2-in. diameter pipe. It was identified that simultaneous hole enlargement while drilling posed certain risks to successfully achieving the objective, which had to be addressed if this option was selected. The main concerns were: the 8 1⁄2-in. Hole-Opener compromising the steerability of the 4 3⁄4-in. AutoTrak tool; the split flow corrupting the logging-while-drilling (LWD) signal; and hole-cleaning challenges. After further analysis and risk assessments by Inteq it was decided that with the use of the newly introduced SARA II decoding system, the signal from the LWD system could be detected and decoded successfully.
Balancing the simultaneous drilling and hole opening assembly was critical. It was important that the hole-opener did not take a significant portion of the applied weight as this would compromise the ability of the 4 3⁄4-in. AutoTrak tool to deliver the required 8.5° per 100 ft DLS curvature. Flow balance was also needed to ensure appropriate flow rate in the 8 1⁄2-in. and 6 1⁄8-in. portions of the drilling assembly and annulus. The optimum split was derived to be 45% flow (300 gpm) through the bit and 55% through the Hole-Opener.
Geology:To reach the UP17 targets required drilling out of the Upper Captain Sand reservoir through a shale section and into the chalk where the entire 206° flat turn would be completed, prior to going though the shale and back into the Upper Captain Sand reservoir again, where the well would be lined up to the targets.
Typically, Captain well reservoir sections avoid drilling shale sections through use of geological pilot holes. This enables the selection of a BARACARB weighted 9.5 ppg Baradril-N drill in fluid to be used in the reservoir. The fact that two sections of inherently unstable shale were exposed in the UP17 well profile necessitated using a higher mud weight to control the shales. For this reason an 11.5 ppg potassium chloride/sodium bromide BARADRIL N fluid was designed and employed. The avoidance of barite as a weighting material, as well as minimizing drilled solids, would also be critical in maximizing productivity.
Apart from the magnitude of the DLS required to hit the UP17 reservoir targets, the other key drilling concern was the collision avoidance while passing the heels of eight horizontal wells. One in particular, well C12, a water injector, had a minimum separation of 40 ft and a clearance factor of 0.59 from the UP17 wellpath. As the well crossing was perpendicular, anti-collision LWD tools were of limited use. Managing this close well proximity was planned through use of different lithology-the UP17 wellpath would enter shale then the Upper Captain Sand approximately 350 ft prior to approaching the drilled depth representing minimum separation, with well C12 known to be in the chalk at this depth. The different lithology would provide confidence on well separation-not observing the prognosed geology would require a low-side sidetrack before the drilled depth representing minimum separation was reached.
Actual well path, showing 8.5º DLS flat turn through the chalk to target oil reserves between heels of existing wells and gas cap.
Lower completion:The major issue with running the screens was the Open Hole/Cased Hole ratio, with the 9 5⁄8-in. casing shoe set at 3,778 ft and the 8 1⁄2-in. section drilled to 9,728 ft. Drag modeling showed that friction-reducing devices were required on the blank pipe run in the open hole behind the 1,600 ft of 5 1⁄2-in. screens. A mixture of 17- and 26-ppf 5 1⁄2-in. blank pipe was required. The 26-ppf blank pipe was non-standard and was required to give extra stiffness at a depth where helical buckling was being predicted. The screens were planned to be run without 4-in. washpipe due to drag issues. The landing string was 5 1⁄2-in. HWDP but 8-in. DC were available as contingency.
Execution challenges
Slot recovery operations were relatively slick with two cuts being made and 300 ft of packer/gravel pack assemblies and 5 1⁄2-in. sand screens being recovered. The Baker SC-1R packer and gravel packing assemblies were recovered down to within 19 ft of the 9 5⁄8-in. casing shoe. Then a spear fishing assembly complete with an Andergauge Agitator tool was run to engage the cut stub of the 5 1⁄2-in. sand screens. An overpull of 250K lbs was taken on the screen fish without success. Then pumping began through the string at 400 gpm to function the Agitator, holding 40K lbs overpull. After a short period, the spear fishing assembly “shake” sucked out the sand screens from across 105 ft open hole.
Two attempts were required to set the 400-ft 17.0 ppg cement plug. The first attempt was unsuccessful in pulling the 3 1⁄2-in. cement stinger out of the cement plug after placing it. It subsequently transpired, from inspection of the running tool, that the Open Hole Bridge Plug run-on cement stinger had been inadvertently re-engaged while disconnecting cement hoses. The second plug was successfully set, using 16.0 ppg lead and tail spacer, to minimize the slumping effect in 92° inclination hole. When tagged with the drilling assembly, the cement plug was 14 ft high and sound.
Drilling operations proved more challenging. The assembly with the selected 6 1⁄8-in. polycrystalline diamond compact (PDC) bit steered and functioned perfectly in the Captain reservoir sequence, immediately out of the 9 5⁄8-in. casing shoe. However, once into the shale, bit balling prevented the system from steering as required. Initially this was suspected to be purely a function of the shale, but subsequent runs confirmed that the PDC bit balling was not allowing the 4 3⁄4-in. AutoTrak x 8 1⁄2-in. Hole Opener system to steer sufficiently in the chalk either. Runs with 6 1⁄8-in. insert bits proved that the 4 3⁄4-in. AutoTrak x 8 1⁄2-in. Hole Opener assembly was capable of delivering the required DLS, but only at low rate of penetration (ROP).
An alternative PDC bit was sourced. This delivered the required turn at the expected ROP but cleaning of the bit was still an issue as plugging/unplugging of nozzles was apparent during the run, even though bit nozzling was significantly tightened.
The ECD at the bit was monitored using downhole pressure from the LWD sensors with the objective to keep the ECD below 14.0 ppg in the reservoir. While drilling through the chalk sequence, total pump rate of 700 gpm was achieved. As the reservoir was re-entered, total flow rate was gradually stepped down to keep within the ECD limit, though at TD, it was still possible to pump at a respectable 600 gpm.
The two sections of shale underlying the chalk and totaling 197 ft were open for in excess of three weeks and crucially, posed no borehole stability problems during drilling or running of the lower completion.
While challenges were met along the way, including three washouts in the 5 1⁄2-in. drillpipe after the 206° flat turn was completed, this very demanding directional profile was successfully drilled and simultaneously enlarged as per plan.
The lower completion operations progressed relatively smoothly. Prior to beginning this phase of operations, two stands of 8-in. DCs were picked up to provide contingency to the5 1⁄2-in.HWDP landing string. At a depth of 4,616 ft the hook load went down to 65 Klbs (10 K over block weight). At this stage with the 26 ppf 5 1⁄2-in. blank pipe at the rotary, the string could be worked through an apparent ledge in the chalk. Thereafter running was fine until a depth of 9,100 ft. At this depth, the two contingency stands of 8-in. DC were needed to provide extra weight at surface to move the screens to a final setting depth of 9,559 ft. Variances between modeled and actual highlighted the difficulty in modeling drag when side loads become as large as were experienced due to the magnitude of the turn on this well.
The standard ESP upper completion was run without any incident and this completed the UP17 well, which has a 98% net to gross sand length in the reservoir.•