Extended-reach drilling develops Sacate field, offshore California

Exxon Co. USA, a predecessor to ExxonMobil, acquired the leases in 1968 and discovered the Hondo field shortly thereafter, along with the Pescado and Sacate fields in 1970.

The company installed the Hondo platform and drilling rig in 1977, began producing first oil in 1981, and has continued an extensive but intermittent program of redrills in the Hondo field since the initial drilling program was completed in 1984.

In 1993, Exxon installed the Heritage and Harmony platforms and drilling rigs to produce the Pescado field and the western half of Hondo, as part of the SYU expansion project.

Extended-reach drilling technology has allowed ExxonMobil to develop the Sacate reserves from the Heritage platform.

The SYU fields produce primarily from the Monterey formation, a highly fractured chert reservoir that contains low gravity, 12-20° API sour crude. Secondary production comes from the deeper Gaviota and Vaqueros sandstone reservoirs that contain lighter gravity sweet crude.

To date, the company has drilled 105 wells in the unit, with current production at about 60,000 bo/d and 80 MMcfd of gas from 75 wells.

Heritage rig upgrades

Work began on the Heritage drilling rig upgrades after approval of the Sacate ERD development program in late 1997.

The company approved five wells for the initial development plan, which it could double in a second phase depending on drilling, production, and geologic data obtained from the initial phase. Horizontal deviations could ultimately exceed 35,000 ft.

The Heritage rig is owned by the SYU, and engineers originally designed it with ERD capabilities in mind, although not to the full extent of a Sacate field development.

Currently, the rig has a 2,000-hp DC drawworks, nominal 500-ton capacity 30 ft x 30 ft derrick, 45,000 ft-lb DC top drive, and a 5,000 psi high-pressure mud system with two 2,200-hp mud pumps.

The company modified the rig to handle the initial Sacate wells, refurbishing and upgrading it to current equipment standards, as well as extending its ERD capability as follows:

• The company purchased a new ERD drillstring that can handle high torque loads (30,000-45,000 ft-lb) and allow for higher circulating rates (900-1,100 gpm) in the long 121/4-in. intermediate hole sections.

The company ordered the 51/2-in., 21.9 lb/ft, MG-105, H2S compatible drillstring with 63/4-in. OD x 4-in. ID wedge-thread connections having 109,000 ft-lb torsional yield.

Engineers had selected this connection type because:

1. It had been proven on similar ExxonMobil affiliate ERD wells in Australia.
2. The 63/4-in. OD tool joint provides the option to drill the 81/2-in. pay section without the need to pick up a smaller string of pipe.
3. The large 4-in. ID minimizes pressure drop.
4. The torque capabilities are less sensitive to OD wear (or a large ID) than shouldered connections.
5. The connection is reasonably immune to downhole makeup damage.

To minimize casing wear and drillstring torque, the company had a low-torque soft-facing product applied flush to the connection boxes.

• Engineers upgraded the top drive stem to 75/8-in. regular connections to better accommodate higher torques.
• The company debottlenecked the mud processing interconnecting piping to handle circulating rates as high as 1,400 gpm and replaced the shale shakers with more modern high-G shakers to handle the higher circulating rates with mineral oil-based mud.

The company also replaced the desilter and mud cleaner with a higher capacity unit and added a second centrifuge.

• ExxonMobil upgraded the closed-system containment to minimize the risk of mineral oil mud spillage.
• The company installed a cuttings grinding and injection system for cuttings and oily waste disposal.

The rig injects the oily cuttings and liquid waste below the surface casing depth on selected Pescado field wells, which is a lower-cost option than hauling cuttings for shore disposal or using synthetic oil mud.

Existing ocean discharge permits do not cover synthetic oil muds.

• The rig added a 2,400-bbl reserve mud storage system, consisting of four 600-bbl tanks, each with a circulating and transfer pump. This brings the total rig active reserve capacity to about 4,300 bbl, minimizing the need to haul mineral-oil mud to shore while the rig drills water-based sections.

To save space and rig-up time between wells, rig designers installed the new tanks on the platform's capping beams, connected to the rig skid base, to move with the rig to each new well slot.

• The rig upgrade added an active "soft-torque" torsional-vibration damper to the electrical rotary drive and enhanced the rig's shore-based electrical power system to accommodate new equipment and allow higher total power throughput.
• The upgrade modified the 30 ft x 30 ft conventional derrick racking board to maximize the setback capacity with 51/2-in. drill pipe to about 25,000 ft.

A third mud pump would have improved non-productive time, but the company did not include it in the upgrade due to space constraints and the more extensive electrical system upgrade, which would have been required.

Well history

Several wells in the SYU reached horizontal displacements of 10,000-13,500 ft during the pre-1998 drilling programs. None of the wells, however, had been drilled with today's ERD practices.

In late 1998, ExxonMobil set out to drill one of the corporation's most challenging wells, the Hondo H-42, which was a 75° S-turn ERD well with about 16,000 ft of horizontal displacement.

Engineers designed, drilled, and completed the well to stay within the limited capabilities of the Hondo drilling rig, which had a 700-bbl total active and reserve mud system, a 500,000-lb rated mast, a 24,000 ft-lb top drive, and two 850-hp mud pumps.

Minor upgrades to the rig included adding a 320-bbl base fluid storage tank, completing the closed-mud containment system, installing two new high-G shakers, strengthening the rig floor to increase racking capability to about 20,000 ft, and adding a cuttings grinding and injection system.

The rig's available circulating rates and torque limitations were the primary drivers for designing and drill ing a "designed for existing rig" ERD well.

Engineering analysis determined that the rig could not successfully drill and clean the long 75° intermediate hole section for a hole size larger than 105/8-in., when limited to a maximum circulating rate of about 600 gpm.

This limitation governed the well design for use of a downsized hole diameter and casing program.

The design called for a 143/4-in. surface hole with 113/4-in. surface casing, a 105/8-in. intermediate hole with 95/8-in. x 85/8-in. production casing, and a 75/8-in. production hole with 51/2-in. production liner.

The rig successfully drilled well H-42 and completed it to a total MD of 19,555 ft, with total vertical depth of 8,356 ft.

The H-42 well represented both a significant accomplishment and learning opportunity for the SYU drilling team, proving that:

• Almost any rig can drill ERD wells, when the wells are designed and engineered within the rig's limitations.
• Training and education in extended-reach drilling practices for the rig crews, service-company personnel, engineering and rig supervisory personnel is vital to the success of the wells.
• Cuttings injection is a workable and viable option to hauling oil-based mud cuttings and liquids at SYU.

Although the ERD well had only minor problems during drilling, ExxonMobil experienced a costly stuck pipe and sidetrack situation due to mechanical failure of the top drive while backreaming from the base of the 105/8-in. hole.

Lessons learned from the H-42 well played a pivotal role in success of the Sacate drilling program.

After completing well H-42, ExxonMobil stacked the Hondo rig. The rig crews, service-company personnel, and rig supervisors moved over to the Heritage platform to drill the first Sacate well, Sa-1.

After drilling well Sa-1, crews drilled the Harmony platform Ha-27 Hondo field stepout well and the subsequent Heritage platform Sacate Sa-2 well.

Sacate program

The Sacate program represents significant achievement for ExxonMobil in the continued development and utilization of ERD practices and equipment.

The Sa-1, an 84° S-turn well drilled to 21,720 ft MD with 6,096 ft TVD, set Exxon Corp.'s record at the time for the longest horizontal displacement at 18,686 ft, with the Sa-2 well setting ExxonMobil's recent displacement record in North America.

The initial Sacate well design and casing program was driven primarily by four major requirements, which were:

• To use a low toxicity, oil-based mud for stabilizing the highly water-sensitive Pico and Sisquoc shales that would be exposed for long periods in the high angle tangent section.
• To drop angle to vertical within the reservoir section, ensuring the well path penetrates several lobes of the Monterey formation (Fig. 2).
• To use water-based mud within the Monterey reservoir to enable the use of formation imaging logs for locating fractures, which along with the best zones, are perforated during well completion.
• To install 113/4-in. production casing to accommodate a dual electric submersible pump (ESP) system, run on a single 41/2-in. tubing string with an option to gaslift.

The requirement to drop angle as the wellbore enters the reservoir leads to well paths with sail angles of 84° and HD/TVD ratios greater than 4.0, which results in wells that experienced "negative weight."^{2 3}

Negative weight occurs when the sliding frictional force exceeds the available slackoff weight resulting in the rig having to rotate or push the drillstring into the well. The same applies to drill pipe-conveyed logging tools, casing and liners, tubing, and other tubulars.

The Sacate wells experienced negative weight with most of these operations.

The need for water-based muds in the top-hole sections and while drilling within the reservoir, but oil mud required through the shales, led the company to install the four 600-bbl mud tanks.

These tanks minimize the swap out times for drilling muds and completion fluids by allowing on loading, off-loading, and storage of the different systems outside the well's critical path.

To meet the completion requirements and remain within a normal hole size and casing program scenario, engineers selected the following casing program:

• 185/8-in. conductor liner set below existing 26-in. drive pipe to isolate an unstable sand bed.
• 133/8-in. surface casing set 1,000-1,500 ft MD past the end-of-build in the 171/2-in. surface hole.
• 113/4-in. x 95/8-in. production casing set at the top of the Monterey reservoir in the 121/4-in. hole.
• 41/2-in. to 51/2-in. production liner (as needed) set at the 81/2-in. hole TD.
• 41/2-in. production liner tieback from the top of the production liner to the base of the dual ESP's.
• Dual ESP's run on 41/2-in. tubing with two electrical cables and multiple control lines set in the 113/4-in. portion of the production casing.

Drilling challenges

The Sacate drilling team had to overcome numerous challenges to drill and complete the wells successfully.

The team set high priorities on drilling as smooth a wellbore as possible, ensuring ERD hole cleaning and tripping procedures were practiced routinely, using rotary assemblies when possible.

Other key issues to be overcome were:

1. Finding a way to install the 113/4 x 95/8-in. production casing without incurring complete lost returns or collapsing the casing.
2. Finding a way to maintain directional control in the 121/4-in. hole when slide drilling would not be an option due to negative weight.
3. Finding a way to get drill pipe-conveyed logging tools down in a negative-weight situation.
4. Finding a way to install a complex completion in a negative-weight environment.

17 1/2-in. surface hole

The Sacate team realized that maintaining a smooth well path would help minimize torque and drag and delay the onset of negative weight. The use of inclination-at-bit technology offered one method for drilling a smooth hole.⁴

Engineers had the idea that by obtaining real time at-bit inclination while drilling the build section of the surface hole, they could minimize hole tortuosity and take the guesswork out of the slide.

At the time the wells were drilled, however, only 63/4-in. tools were available, which led to drilling the 171/2-in. hole as a 97/8-in. pilot hole followed by a hole opener run using special built poly-crystalline diamond compact (PDC) cutter hole opener bits.

In addition to the at-bit technology, the company drilled short 10-15 ft slides to reduce the overall dogleg severity in the build section, rather than sliding 30-40 ft and rotating down the stand as was the typical practice.

Drilling engineers used the at-bit technology and modified sliding procedure for the first two Sacate wells.

Although the at-bit technology made the directional driller's job easier and perhaps more efficient, a post-well analysis of the dogleg severity and actual-vs.-planned build rates did not dem onstrate a "smoother" profile as compared to a non-at-bit drilled build section.

Future use of the technology, for this type of application, requires additional study. Two positive results, however, came out of the surface hole drilling operation.

First, the use of PDC hole openers proved highly successful, yielding longer life and almost double the penetration rate in water-based mud, compared to earlier attempts using conventional roller cone hole openers.

On Sa-1, the rig opened the pilot hole in two stages. For Sa-2, however, the rig opened the hole in only one stage, with use of the modified hole opener based on the successful Sa-1 experience.

Secondly, the Sa-1 well proved that the debottlenecking work on the mud processing pits was successful.

The rig achieved 1,450 gpm during the 171/2-in. hole-opener runs, which significantly improved penetration rates and hole-cleaning capability. Previous wells had been limited to about 800 gpm.

13 3/8-in. surface casing

The 133/8-in. casing running and cementing operations were trouble-free for both Sacate wells.

The team set surface casing as shallow as 5,900 ft MD on Sa-1 and as deep as 8,000 ft MD on Sa-2, based on completion design issues and 95/8-in. casing flotation needs.

The rig ran the 133/8-in. casing without flotation even though it was run to more than 3,000 ft into the 84° tangent section.

The addition of the four 600-bbl mud storage tanks simplified the operations and shortened the time required to swap muds.

The company chose Escaid 110, a low-toxicity mineral oil, as the base fluid because of its low environmental impact (less than 0.2% aromatics), proven success on numerous ExxonMobil wells, and low kinematic viscosity (1.7 cst at 38° C.).

Kinematic viscosity is directly related to plastic viscosity. A base fluid with a low kinematic viscosity will inherently have a low plastic viscosity, which in ERD wells means increased hydraulics and a lower equivalent circulating density (ECD).

121/4-in. hole section

Both the Sa-1 and Sa-2 wells had 121/4-in. intermediate hole intervals of about 15,000 ft with the Sa-2 being the deepest or longest, ending at about 22,800 ft MD.

In addition, the well paths for both wells were virtually identical with 84° sail angles that S-turn dropped into the Monterey reservoir.

Several factors were responsible for successfully drilling these intermediate hole sections; however the company's goal of maintaining a clean hole was the main reason for success.

For the drilling team, this meant keeping the circulating rates and drill pipe rotation speeds at a maximum, maintaining the correct mud properties and minimizing sliding operations.

This and the goal of maintaining a smooth well path drove the bottomhole assembly (BHA) designs toward rotary assemblies.

Well known in the industry, rotary assemblies provide superior hole cleaning, compared to steerable assemblies. Due to lack of rotation, slide drilling does not effectively lift cuttings into the mud flowstream.

The team, therefore, had planned to use rotary assemblies to maintain better hole-cleaning capabilities.

The eventual onset of negative weight is another reason for the use of rotary assemblies. When negative weight occurs, the rig can no longer slide the drillstring, making it virtually impossible to steer and slide drill.

The drawback to using rotary assemblies, however, is their limited directional control, especially azimuth control.

When the drilling team had planned the Sa-1 well, rotary steerable assemblies were relatively new and proven 121/4-in. tools were not available in the US.

The Sacate team knew that azimuth control would be essential to locate the well path within the target constraints.

Since slide drilling would not be possible and the availability of rotary steerable tools was minimal, the team chose the fallback option of "walking" PDC bits.

The team chose this solution based on earlier success with walking bits by an ExxonMobil affiliate in Australia.

The drillbit manufacturer designed second generation walking bits for the Sacate program, based on that experience and initial PDC bit runs on other SYU wells (Fig. 3).

The bit's design and drilling parameters of rotary speed (rpm) and weight on bit control the tendency for the bit to walk left or right.

Since PDC bit walk on SYU wells was not consistent, the team decided to use a neutral PDC bit run below a multi-position variable gauge stabilizer (VGS) on the initial bit run below the 133/8-in. surface casing on Sa-1.⁵

The multi-position stabilizer, with eight positions at 1/4-in. increments, from 101/2-in. to 121/4-in., provided full inclination control. Right-hand walk, however, did occur at 0.3-0.5°/100 ft.

After making a steerable motor correction run, above the negative weight point, the team ran the left-hand walking bit in the hole with the variable gauge stabilizer.

The left-walking bits were able to combat the right-walk tendency of the drillstring, while drilling more than 10,900 ft of hole prior to entering the drop section.

In the Sa-2 well, left-walking bits drilled more than 12,200 ft of hole. Due to the longer throw of the Sa-2 and a harder right-hand walk tendency in the bottom half of the interval, however, the well path had to make a course correction.

A rotary steerable tool, which had become available in the US, made the course correction.

Between the two Sacate wells and the Harmony Ha-27 well, the first left-walking bit had drilled 26,498 ft of hole.

The combination of the VGS and left-walking bits were also highly successful maintaining a low dogleg severity of about 0.4°/100 ft in the 121/4-in. hole tangent sections.

Even with a smooth wellbore, however, negative weight occurred below about 12,000 ft MD.

Left-walking PDC bits require low rotary speed (40-60 rpm) when left-walk tendency is desired, which is their only drawback.

Low rotary speed reduces hole-cleaning capability but does not compromise the overall penetration rate because the rig must increase weight-on-bit.

To compensate for loss in rotary speed, drilling operations required additional circulation and rotating time.

Based on the predominant right-hand walk tendency of the Sacate wells, the walking bit manufacturer has since redesigned the left-walking bit to require high rpm's and avoid compromising hole cleaning. The industry continues to mature walking bit designs.

Industry experience on ERD wells has led to the use of 65/8 x 51/2-in. drillstrings for improved hydraulics and hole cleaning.⁶

Most rigs, however, are not outfitted for 65/8-in. drill pipe plus the size and weight of 65/8-in. drill pipe makes it difficult to handle.

The Hondo H-42 well has proven that maintaining a clean hole is not just a function of circulation rate.

Effective hole cleaning is a mix of adequate circulation rate, plus high speed rotation (120-150 rpm), combined with good hole-monitoring practices, such as recording, plotting, and analyzing torque and drag parameters while drilling and tripping.⁷

Effective hole cleaning also requires adequate hole cleanup cycles of 3-5 bottoms-up, while monitoring cuttings returns on the shakers.

The use of these techniques has allowed ExxonMobil successfully to drill 121/4-in. holes using a 51/2-in. drillstring and circulating rates as low as 900 gpm, with room to spare.

Surveying accuracy

Survey accuracy is normally a critical issue on ERD wells, especially those that are horizontal in the reservoir section.

Measurement while drilling (MWD) survey tools provide reasonably accurate surveys, but the earth's magnetic field and tool sag in the drillstring affected reading accuracy.

Fortunately for the Sacate drilling program, the geological plan required hitting the reservoir targets at a mid-angle approach (40-60°) rather than near horizontal.

This requirement eliminated the horizontal landing issue and gave flexibility in vertical uncertainty of the reservoir.

Even so, the Sacate team knew that obtaining an accurate survey was necessary for penetrating the target box and for potential well control issues.

The normal practice for reducing MWD survey uncertainty is to run high-accuracy inertial rate gyros at each casing point and sometimes deep in the wellbore although shallow enough to make a course correction prior to penetrating the reservoir.

However, the Sacate team took a different approach for obtaining an accurate survey.

On ERD wells, survey accuracy is not critical at surface-casing depths, since there is plenty of distance below the surface casing to make course corrections.

As such, MWD surveys to this point are adequate and running a gyro inside the surface casing is of little value.

However, if the high-accuracy gyro is run through drill pipe about 500 ft or so in the open hole below the surface-casing shoe, in a fairly flat tangent section, then all succeeding MWD tools can use the gyro survey as an in-hole reference calibration tool.⁸

Upon tripping in or out of the hole with a BHA, the rig positions the MWD tool at a known gyro survey depth. The tool then takes four or five surveys at several toolface orientations, normally every 90°.

Technicians then average and compare the MWD azimuth surveys to the gyro azimuth survey, then apply the calculated azimuth adjustment (positive or negative) to future drilling surveys.

If crews change the BHA or replace the MWD tool, they must repeat the survey reference process.

One should note that sufficient nonmagnetic drill collars and stabilizers must be run both below and above the MWD tool to minimize magnetic interference.

Technicians must determine inclination adjustments by computer software programs that model BHA and MWD tool sag.⁹ They enter the entire survey into the model, which calculates the tool sag or bending of the collar.

Future drilling surveys must apply the calculated tool sag adjustments. Technicians must recalculate the model for the new parameters, if the well path angle changes significantly, such as entering a drop or build section, or the BHA is changed.

To confirm the accuracy of these survey adjustments, the team ran a high-accuracy inertial rate gyro at the 95/8-in. casing shoe at the top of reservoir.

The gyro survey indicated a vertical depth delta of only 14 ft and a horizontal displacement delta of only 47 ft when compared to the corrected MWD survey.

For the Sacate program, this surveying method provided the required accuracy and is now the standard practice on all Sacate wells.

Intermediate casing

The 113/4 x 95/8-in. intermediate casing job presented two significant challenges to the team. The first was to get the string to the bottom of the hole in a severely negative-weight environment.

The second was to get the string down without breaking down the 133/8-in. shoe or collapsing the air-filled 95/8-in. casing due to the high surge pressures created by the 113/4-in. casing.

Initial slackoff tension modeling indicated that the rig had to float-in the casing string to reach bottom and that the casing would require additional top drive "push" assistance during the process.

Further modeling showed that reaching bottom was marginal, at best, with the rig running the 95/8-in. portion of the casing string as a liner on the 51/2-in. drill pipe.

Estimated rotational torque values would probably exceed the top drive's capability and the drill pipe would have insufficient weight or stiffness to push the 95/8-in. casing.

In addition, surge modeling for running the 113/4-in. portion of the casing string inside the 133/8-in. casing showed that a large dynamic pressure surge could exceed the openhole integrity causing lost returns and could collapse the air-filled 95/8-in. casing.

Therefore, the team had to develop an alternative method to run the tapered casing string.

One suggested method called for running and cementing the 95/8-in. portion as a liner, on a 95/8-in. "running string" and then tieback the 113/4-in. portion to surface.

This option would maintain the needed push weight capability on the 95/8-in. casing and minimize the 113/4-in. casing surge problem.

It would create another challenge, however, of maintaining the full casing ID needed for the flotation equipment and at the same time be able to disconnect after cementing so that the 95/8-in. running string could be removed and the 113/4-in. tieback string installed.

The drilling team worked with the liner-hanger supplier to develop a full-bore polished bore receptacle disconnect and tieback assembly to run integral with the 95/8-in. casing.

After cementing, the rig would disconnect the 95/8-in. running string by either right-hand rotation (5,000 ft-lb shear) or straight-pull release (300,000-lb shear), then pull out of the hole.

Afterwards, the rig would run the tieback seal stinger and 113/4-in. casing open ended, eliminating any surge concerns, pump the cement job, and sting into the receptacle.

This casing string would not require floats because the cement hydrostatic differential is isolated upon stinging into the receptacle.

Engineers were initially concerned with the proposed disconnect shear methods because it eliminates casing rotation and limits the option of pulling the casing back out. The team accepted the risk, however, after reviewing the slackoff modeling results.

The 95/8-in. casing would not require rotation to get down and the worst-case running tension load on the disconnect would never exceed 150,000 lb. The possibility of attempting to pull out of hole after running more than 15,000 ft of casing was remote.

The rig backreamed the 121/4-in. hole and 133/8-in. surface casing, up into the surface-hole build section, prior to running the 95/8-in. casing.

If operators use good hole cleaning and monitoring practices, backreaming becomes unnecessary. Floating-in the casing, however, eliminates the option of rotating and washing down. Ensuring the hole is as clean as possible becomes critical.

Backreaming out reduces the high angle cuttings bed to a minimum. The rig used a minimal BHA that included a reduced OD bit, reduced OD stabilizer, and one drill collar, to reduce the risk of packing off while backreaming.

The backreaming objective was to get cuttings into the mud flowstream for circulation out of the well, rather than redrilling the hole backwards.

Using a 95/8-in. running string, the company successfully ran the 95/8-in. casing to bottom using a mud-over-air flotation sleeve and collar assembly.¹⁰

The flotation sleeve assembly failed to shear properly, however, and had to be pushed to bottom with drill pipe after filling the casing with mud.

After cementing the casing in place, the rig rotation-shear released the running string and pulled out of the hole.

The rig ran the 113/4-in. open-ended tieback, stabbed into the receptacle, and cemented it in place through a stage-cementing collar. The disconnect-tieback system worked perfectly. The company has since run it successfully on a similar well.

Post-failure analysis of the flotation sleeve indicated that the type of aluminum used in the sleeve had too wide of a yield range. Flotation sleeves run successfully on subsequent wells used a different type of aluminum.

In addition to flotation and disconnect systems, the use of casing roller centralizers reduced the slackoff drag friction.¹¹ Slackoff modeling had indicated that getting the last 2,000-3,000 ft of casing to bottom may be difficult, even with pushing and flotation.

Post-job analysis indicated that casing rollers, run primarily on the 95/8-in. running string with some in the open hole, reduced friction by two-thirds and allowed the casing to run to bottom easily.

Fig. 4 shows the 95/8-in. as run, slackoff hookload chart. The chart shows five predicted slackoff plots, calculated for varying friction factors and without roller centralizers.

The actual slackoff plot shows a running friction factor of 0.35-0.4 until about 15,000 ft, the depth at which the rollers provide benefit and the friction factor drops to about 0.1.

The crew recovered the rollers from the running string, which were reused on the next 95/8-in. flotation job in well Ha-27, after minor reconditioning.

ExxonMobil successfully used the same flotation system with roller centralizers for the Sa-2 well, which was 2,500 ft further out than well Sa-1.

The additional horizontal displacement of Sa-2 required that crews run the rollers in two batches to minimize top drive push requirements, which was limited to 30,000 lb force. This placed about one-third of the rollers in open hole.

As on well Sa-1, the rollers reduced casing friction by about two-thirds.

The drilling program for Sa-2 did not call for the disconnect-tieback system because the well did not require 113/4-in. production casing.

8 1/2-in. production hole

The 81/2-in. hole through the Monterey reservoir presented two main challenges: first was managing torque and drag while drilling with water-based mud and second, maintaining equivalent circulating densities below the formation integrity limit.

Engineers controlled torque and drag in Well H-42 with both nonrotating drill pipe protectors (NRDPPs) and a water-based mud lubricity additive.¹² Torque modeling on Sa-1, however, indicated that NRDPPs would provide only marginal benefits.

Side-force loading through a dogleg interval from pipe tension, such as the drill pipe applying contact force in the build section, generates most of the drilling torque.

With negative-weight conditions in the Sacate wells, created by the shallow kickoff and long, near horizontal well paths, the long tangent section generates most of the drilling torque.

With minimal or no tension, the build section generates little torque. To be effective, the rig would have to install NRDPPs on the entire drillstring, which would be excessive for a 20,000-ft well.

This left the options of a tapered drillstring or lubricity agents. The company obtained, inspected, and shipped a tapered 5 x 41/2-in. drillstring to the dock as backup to replace the bottom two-thirds of the 51/2-in. drillstring, if needed for Sa-1.

The water-based mud lubricity additive reduced torque and drag effectively in the H-42 well, but it formed a putty-like emulsion in combination with hematite in the mud and became apparent after displacing to completion fluid.

Thought to be a contamination problem with cement, the same problem occurred on an earlier well. The H-42 well proved, however, that cement was not the problem.

The putty-like substance adhered to the drill pipe and casing and prevented the production packer from getting down.

A surfactant, which broke the emulsion, successfully cleaned the putty-like substance from the well. Subsequent laboratory testing was able to duplicate the emulsion.

Continued testing of other lubricity agents had identified a combination of two additives that would provide lubricity without forming emulsions. One provided lubricity in cased holes, and the other provided openhole lubricity.

ExxonMobil had special-ordered the 51/2-in. drillstring with 63/4-in. OD connections to minimize ECD and allow drilling of the 81/2-in. reservoir section, without changing to a tapered drillstring.

Maintaining a full string of 51/2-in. drill pipe, however, would require lubricity additives to manage torque.

This became apparent in the 81/2-in. hole section of Sa-1 well from the start.

While attempting to drill out the float equipment with water-based mud, drilling torque reached 40,000 ft-lb, which is the top drive stall-out limit.

Crews treated the mud with 3 vol % of the lubricity additives, which resulted in torque dropping by one-half, allowing the rig to drill out the float equipment.

Possibly in combination with the cement, however, the additives flocculated the mud, causing lost returns. The drilling mud was subsequently conditioned back to the desired properties.

Mud testing at the rig indicated that the openhole lubricity agent may have been the problem and use of the additive was discontinued.

The cased-hole lubricity additive proved highly beneficial for the remainder of the drilling operations and also for the drill pipe-conveyed logging operations.

Engineers included a pressure-while-drilling (PWS) sub in the logging-while-drilling (LWD) tool string to help manage ECD and minimize lost returns while drilling the Monterey reservoir.¹³

The naturally fractured Monterey reservoir has a history of lost returns, sometimes even with seawater. The team was concerned of the potential for lost circulation, even though the Monterey reservoir in the Sacate field area was originally pressured.

Engineers performed an ECD test, after they had drilled into the top of the reservoir.

The tests reached circulation rates up to 450 gpm and rotary speeds up to 120 rpm, both separately and in combination, to determine the maximum ECD the formation would experience and find out if it would break down, as it had when drilling out the shoe.

ECDs ranged from 1.5 to 2.25 ppg more than static mud density when drilling with 8.5 ppg water-based mud. The PWS sub provided real-time ECD data for adjusting circulating rates and rotary speeds.

The real-time ECD data from the PWS sub allowed crews successfully to drill the hole with 51/2-in. drill pipe, even when tripping operations required increased mud weight by almost 1 ppg to prevent swabbing in the well.

The 81/2-in. hole on Sa-2 had similar drilling issues as Sa-1, but it included a change in the mud program.

A re-evaluation of the development drilling program after completing Sa-1 led to some potential wells that would require the mineral oil-based mud to reach and drill the formation objectives.

The drilling team felt it was imperative to determine whether the Monterey could be drilled with mineral oil- based mud without experiencing severe lost returns in this fractured reservoir.

As on Sa-1, the PWS tool was a very valuable asset while drilling the Monterey with the mineral oil-based mud. Drilling losses occurred but were controlled with the use of lost circulation material and minimized ECD.

Engineers observed an interesting phenomenon with respect to ECD and mud type on Sa-2.

The pressure-while-drilling measured ECDs with the Escaid 110 mineral oil- based mud were 0.5-0.75 ppg less than the measured ECDs on Sa-1 using water-based mud, even though the rheology properties were similar.

The company had to obtain the drill pipe conveyed formation imaging log in the Monterey in Sa-2, even though drilling operations had used mineral oil-based mud.

After investigating the options, engineers determined that displacing the openhole interval with water-based mud should be adequate to obtain acceptable log quality.

The displacement operation caused only minor contamination of the oil and water-based muds and although the log quality was not high, it was acceptable for choosing the perforation interval.

Production liner

Obtaining a quality cement job on the production liner is always a high priority; however, the team also wanted to minimize the mechanical risk of having to perform cleanout operations inside the small production liners at high angles.

The team accomplished this by running an inner cementing string inside the liner, connected to the liner running tool.

The 27/8-in. inner string provided a conduit for the cement to pass through, shielding the production liner from the cement.

This system requires the use of cement darts designed to pass through both the drill pipe running string and the 27/8-in. inner string and land in a receiving float collar.

After cementing, the rig pulls the running tool out of the hanger assembly and the wellbore is circulated clean.

This procedure ensures that there is no cement inside the wellbore and displacement to completion fluid can be initiated as soon as the cement behind the liner overlap has set and been pressure tested.

To help ensure a successful cement job, positive body standoff low-torque centralizers center the liner in the hole and allow rotation during cementing.

A long liner overlap of 3,000 ft, provides annular capacity for excess cement.

This liner running and cementing technique was successful in both Sa-1 (with a 41/2-in. liner) and Sa-2 (with a 51/2-in. liner) and eliminated the typical post cement drill out runs on liner jobs.

The company has not attempted the option of running an integral liner top packer to reduce the wait on cement time.

Drag modeling indicates that due to buckling and inefficient weight transfer the running string cannot generate the weight necessary to set the liner top packer. Engineers are continuing development work in this area.

Completion

The company re-evaluated the need to install ESPs after obtaining openhole logs and cuttings analysis. Exploration well data, which indicated very low API gravity (in the low teens), led to the original plan of installing ESPs.

The Sa-1 logs and cuttings data, however, indicated that API gravity would be in the high teens and reservoir gas would aid in lifting the formation fluid.

The company decided to omit the ESPs and complete the well as a single conventional gaslift well, after taking into account the complexity of the dual ESP completion.

A service company made several unsuccessful attempts to run a through drill pipe cased-hole gamma ray correlation log with a 21/8-in. wireline tractor tool and by pumping the tool down, after the drilling mud had been displaced to brine.¹⁴

Neither the tractor nor the pumping operations could overcome the frictional drag forces on the wireline cable. Also, the tool had to pass through both 51/2-in. drill pipe and 27/8-in. tubing, limiting the tractor tool size and pump down swab cups.

The company decided to perforate with use of drill pipe conveyed guns, but without a cased-hole correlation log, since the planned perforation interval was continuous and had 10 ft of depth tolerance.

Engineers determined the drill-pipe perforating interval by comparing the openhole LWD logs with the openhole drill pipe-conveyed logs.

The team minimized stretch error by using the top of the liner as a landing point for the perforating assembly and perforated the well successfully.

High wellbore angle and the negative weight issue required that the company run the 41/2-in. production tubing and completion equipment in stages. The first stage consisted of a 41/2-in. production liner and tieback seal assembly, with a rotatable liner packer.

The rig ran the assembly through the long 84° tangent section on 51/2-in. drillstring.

The team maintained the option of rotating the drillstring to combat negative weight, but rotation was not necessary due to reduced friction from the use of lubricity additive and strategic placement of HWDP in the running string.

The second stage consisted of a short interval of 41/2-in. production tubing and the production packer. The team used the liner packer and production packer to isolate the 113/4-in. tieback cement stage collar and 95/8-in. tieback receptacle.

Since the field uses high pressure, sour gas for gaslift, the company decided that the collar and receptacle should not be exposed to the sour gas even though they were designed for full pressure sour service.

The third stage consisted of the remainder of the 41/2-in. production tubing, the bottomhole pressure gauge, chemical injection mandrel, safety valve, and associated control lines, and the gaslift mandrel.

Crews acidized the well after installing the christmas tree. The well tested at slightly less than 9,500 bo/d.

ExxonMobil completed the Sa-2 well in a similar manner.

The company again used drill pipe- conveyed perforating due to high wellbore angle; however, crews obtained a cased-hole gamma ray correlation log using LWD technology inside cased hole.

The gamma ray log correlated accurately to the openhole logs, which were tied-in to radioactive tracer subs in the 51/2-in. production liner.

The Sa-2 tested at more than 7,500 bo/d.

Sacate well Sa-3

ExxonMobil is applying the experiences from the Sacate program to other high-angle ERD wells, both locally and in other areas of the world.

The company continued the Sacate drilling program with the recent completion of the Sa-3 well, which presented a different technical challenge.

Even though it had a similar 84° well path and only slightly less horizontal displacement of 17,104 ft, the wellbore penetrated both the Monterey chert and the deeper Gaviota and Vaqueros sandstones.

The well's ERD S-turn profile created significant torque and drag loads due to drilling more than 3,000 ft of vertical hole below the drop back to vertical interval in the Monterey formation. Fig. 5 shows directional profiles of the Sacate wells.

The team also successfully installed a complex dual 27/8 x 51/2-in. completion.

Acknowledgment

The authors thank ExxonMobil management for the opportunity to publish this article. The authors and ExxonMobil acknowledge the contributions made by, but not limited to, K&M Technology Group LLC, Baker Hughes Inc., Schlumberger Ltd., Smith International Inc.-GeoDiamond, Leam Drilling Systems, M-I LLC, BJ Services Co., Helm erich & Payne IDC, and all other companies and personnel involved in the Sacate development program.

References

1. Extended reach well database received from Jonny L. Gent of BP and personal communications with C. R. Dawson of ExxonMobil.
2. Hood, III, J.L., Mueller, M.D., Mims, M.G., "The Uses of Buoyancy in Completing High-Drag Horizontal Wellbores," SPE 23027, presented at SPE Asia-Pacific Conference, Perth, Australia, Nov. 4-7, 1991.
3. Gadberry, R.L., Switzer, S.S., "Long Reach High Angle Drilling in the Gulf of Mexico: A Case History of Eugene Island Block 370 Well B-11," IADC/SPE 23866, presented at IADC/SPE Conference in New Orleans, Feb. 18-21, 1992.
4. Skillingstad, T., "At-bit Inclination Measurements Improves Directional Drilling Efficiency and Control," IADC/SPE 59194, presented at IADC/SPE Conference in New Orleans, Feb. 23-25, 2000.
5. Odell II, A.C., Payne, M.L., Cocking, D.A., "Application of a Highly Variable Gauge Stabilizer at Wytch Farm to Extend the ERD Envelope," SPE 30462, presented at SPE Conference in Dallas, Oct. 22-25, 1995.
6. Vighetto, R., Naegel, M., Pradie, E., "Total Drills Extended-reach Record in Tierra del Fuego," OGJ, May 17, 1999, p. 51.
7. Vighetto, R., Naegel, M., Pradi, E., "Teamwork, Downhole Technology Expedites Tierra del Fuego Operations," OGJ, June 7, 1999, p. 60.
8. Cooking, D.A., Bezant, P.N., Tooms, P.J., "Pushing the ERD Envelope at Wytch Farm," SPE/IADC 37618, presented at SPE/IADC Drilling Conference in Amsterdam, Mar. 4-6, 1997.
9. Williamson, H.S., "Accuracy Prediction for Directional MWD," SPE 56702, presented at SPE Conference, Houston, Oct. 3-6, 1999.
10. Mason, C.J., Allen, F.M., Ramirez, A.A., Wolfson, L., and Tapper, R., "Casing Running Milestones for Extended-Reach Wells," SPE/IADC 52842, presented at SPE Drilling Conference, Amsterdam, Mar. 9-11, 1999.

11. Mason, C.J., Williams, L.G., Murray, G.N., "Reinventing the Wheel - Reducing Friction in High-Angle Wells," SPE 63270, presented at SPE Conference, Dallas, Oct. 1-4, 2000.
12. Moore, N.B., Mock, P.W., Krueger, R.E., "Reduction of Drill String Torque and Casing Wear in Extended Reach Wells Using Non-rotatingrill Pipe Protectors," SPE 35666, presented at SPE Regional Meeting, Anchorage, May 22-24, 1996.
13. Green, M.D., Thomesen, C.R., Wolfson, L., Bern, P.A., "An Integrated Solution of Extended-Reach Drilling Problems in the Niakuk Field, Alaska: Part II- Hydraulics, Cuttings Transport and PWD," SPE 56564, presented at SPE Conference, Houston, Oct. 3-6, 1999.
14. McInally, G., Hallundbaek, J., "The Application of New Wireline Well Tractor Technology to Horizontal Well Logging and Intervention: A review of Field Experience in the North Sea," SPE 38757, presented at SPE Conference, San Antonio, Oct. 5-8, 1997.

The authors

William (Bill) C. Elks Jr. is a senior staff drilling engineer for ExxonMobil, Houston, where he is assigned to development drilling technical-operations support group as the company's primary advisor on ERD. He was the lead drilling engineer for the Hondo H-42, Sacate Sa-1, and Harmony Ha-27 ERD wells and other projects around the world. He joined Exxon in 1978 and has a BS in chemical engineering from the University of South Carolina.

Roy A. Masonheimer ([email protected]) is currently with K&M Technology Group, The Woodlands, Tex., a consulting firm specializing in ERD and complex directional wells. He retired from ExxonMobil in 2000 after 32 years, primarily as a drilling engineer. His background includes HPHT, arctic, platform, and subsea environments, as well as rig design and construction. He has a BS and MS in petroleum engineering from the University of California, Berkeley.