Horizontal, underbalanced wells yield high rates in Colombia

In the Middle Magdalena Valley, a U.S. independent drilled the country’s first two horizontal wells using underbalanced techniques

Scott C. Cooper, Senior Vice President of Engineering, Harken International, Ltd.; and Robert L. "Bob" Cuthbertson, PE, UBD Engineering Manager, Inter-Tech Drilling Solutions, Ltd.

arken de Colombia, Ltd., recently completed the Catalina 1 and Olivo 1 in Colombia’s Middle Magdalena Valley, two wells which aptly demonstrate the significant benefits of a properly planned / executed underbalanced (UBD) project. Technical discussions here describe experience gained at the first well and how that was implemented in the second well, providing rapid advancement up the learning curve.

This was evidenced by reaching the goals of horizontal, underbalance and extended reach classification (MD to TVD depth ratio of 2.1). Every step in the process was carefully coordinated, from initial geological review and candidate selections, through pre-engineering study, to final planning, drilling, testing and completion.

As these were to be the first such horizontal underbalanced wells in the country, a high degree of planning and training was scheduled. Results show that was justified: Catalina 1 yielded a 3-day annular flowing test from the Rosa Blanca formation through restricted surface facilities of 7,073 bopd (36°API) and 11.5 MMscfd gas, a 9,680 boepd equivalent. The well was tested for several hours at rates >10,000 bopd and 16.1 MMscfd, but was cut back when gas carry-over forced shut-off of one test separator. Flowing annular pressures were 600 to 700 psi.

Calculated productivity index (PI) of the Catalina well is 12.5 bbl/day/psi, equating to a calculated openhole flowing potential of 21,376 boepd. The well was subsequently cleaned-up to produce 38.5°API crude.

The Olivo 1 tested up to 10,800 bopd (16.5°API) with negligible water and gas through inefficient artificial lift and surface facilities. The well has a PI as high as 46 bbl/day/psi, yielding a calculated deliverability of 79,000 bopd, assuming absolute drawdown to zero flowing BHP.

Introduction

The Basal Limestone / Rosa Blanca and La Luna formations are very extensive, vertically fractured Cretaceous age limestone reservoirs. In Buturama field, the Rosa Blanca has produced 500,000+ bbl oil from three vertical wells drilled in the early 1950s. In nearby Totumal field, the La Luna has produced in excess of 800,000 bbl oil from three vertical wells.

The appearance of high initial test rates and good cumulative recovery from those conventional wells led to a comparison with other vertically fractured formations. Considering these other formations’ success when using horizontal and underbalanced drilling practices, management made a decision to utilize these techniques to exploit both the Rosa Blanca and La Luna. To assist with wellbore construction, Harken chose Inter-Tech Drilling Solutions, Ltd. to provide experienced personnel to work in Harken’s office during planning and at the wellsite during UBD implementation.

Well design. Fig. 1 depicts final well construction of Catalina 1 and Olivo 1. Both wells’ casing programs include 20-in. surface pipe, 13-3/8-in. intermediate casing and a 9-5/8-in. liner through the curved build-up section, with a perforated 9-5/8-in. tie-back casing utilized for concentric casing string gas injection. Production tubing is 4-1/2-in.

For its continuous-supply and safety considerations, membrane nitrogen was selected as the gas used to create underbalanced conditions. A 1,500-scfm generator system at 2,000-psi discharge pressure was selected for this application.

Inter-Tech’s rotating blowout preventer (RBOP) was chosen as the wellhead diverter for its reliabililty, a redundant quick change packer (QCP), and inner packer rubber sealing elements. Both manual and air-activated emergency shutdown (ESD) were placed between the wellhead and the special 3-choke (two 2-1/16-in. adjustable, one 2-1/16-in. manual) 5M manifold. A sample catcher and chemical injection pumps were placed between manifold and mud-gas separator. An open-top two-tank skimmer system separated oil, water and cuttings.

Twin 8-in. flare stacks with automatic igniters handled produced casinghead gas. A detonation arrester in the flare line prevented backfires and explosions. The operator — expecting high production rates and test volumes — installed four 10,000-barrel stock tanks to handle produced oil, Fig. 2. The flow diagram for UBD operations is illustrated in Fig. 3.

Planning and training. Projects progressed through both horizontal drilling technology and underbalance drilling workshops, to which, invitations included: company employees, contractors, Colombian national oil company (Ecopetrol) employees, the Mines and Energy Ministry, as well as faculty and students from the National University in Medellin and the University of America in Bogota. Further training / planning included: special study regarding safety issues during UBD, training with an RBOP / choke flow loop, macro and finite pre-spud meetings, and down-to-final implementation of drilling / completion operations.

All aspects of the well were evaluated in detail and designed with contingency, including safety / security, UBD flow modeling, torque / drag prediction, synthetic log modeling, and directional planning for geosteering with MWD. The latter included an annular BHP sensor, and resistivity / directional surveys at the bit, plus satellite transfer of real-time drilling data.

Additionally, wells were planned to be completed live in an underbalance condition with isolation packers and an electric submersible pump (ESP). Pre-engineering study and offset well log analysis selected geological markers for correlation during the wellbore construction / targeting. The final well plan was presented in Santafe de Bogota in pre-spud meetings, and formal presentation to Ecopetrol. In addition, key Colombian personnel with both the operator and drilling contractor were given more training in Texas, including visits to an ongoing, horizontal-underbalance drilling project in the Austin Chalk formation.

Pilot hole evaluations. The Rosa Blanca, and the overlying La Luna limestone formations were conventionally cored and logged in the vertical pilot hole of the Catalina well. Results of both intervals were favorable, showing bleeding oil and gas, vugular oil staining, open natural fractures and even rubblized sections.

Modular Formation Dynamics Tester (MDT) and Formation Micro-Imaging log (FMI) tools were included in the logging suite. In both Rosa Blanca and La Luna sections, natural fractures were judged to be more intense and higher quality than expected. Additional zones appear to be productive also, thereby opening the possibility of future multi-lateral completions / re-entrys.

Catalina Project Overview

In large part due to planning / training before the job start, the Catalina well proved the concept of underbalance drilling with clear water fluids in this type of reservoir rock. An extended well test through three separators showed the well capable of high oil / gas production rates. Multiple fractures were drilled and commingled in the drilling of only 1,230 ft of lateral section.

Prior to drilling the horizontal section, logs from the vertical pilot hole were transferred into the seismic workstation for better calibration of formation depths. The modern log data was merged with reprocessed seismic to further confirm the geological structure and aid with geosteering expectations.

Testing for hole length. As drilling progressed, the well became difficult to control. Higher-than-expected flowrates and gas carry-over overwhelmed the designed surface equipment. Safety concerns required shut-down several times. The decision was made to stop drilling at 2,000 ft vertical section (VS) and conduct an openhole test. If the test showed unsustainable productivity, drilling would continue. However, the well produced at exceptional rates, and it was therefore considered to be at total depth for a first well attempt.

The test was made by flowing up the casing / drill pipe annulus, through the BOP stack and RBOP with drill pipe in the well and the bit positioned at the bottom of the casing, should a kill string be required. Because the reservoir was producing high gas / oil volumes, concerns were that the test may not be truly indicative of the reservoir potential due to possible turbulence and restrictions at the bit, and through the small-diameter surface flowlines. Subsequent full-bore tubing flow tests have yielded in excess of 8,300 bopd and 14 MMscfd at a flowing tubing pressure of 500 psi on a 3-in. choke.

Equipment improvements. As with every new project of this magnitude, theories were proven, lessons were learned and improvements were planned for future wells. The importance of flare design, including use of a detonation arrester, is strongly recommended for such high-capacity systems. Chemical testing / injection is critical to minimize foaming and oil / water emulsions that lead to entrained gas problems. High-volume separation vessels, pumps and flowlines are necessary to accommodate the circulating volume of drilling water for proper MWD signal performance, plus influxing oil / gas as well as the returning nitrogen, while still maintaining required underbalance pressure.

The new string of 5-in., S-135 drill pipe used for this well also created excessive wear on the rotating diverter’s rubber packer elements due to the three concentric identification grooves on the tool joints of this API product. The sharpness of these grooves proved destructive to the nitrile rubber packer elements during stripping operations. The grooves were found packed with rubber slivers, which eventually ended up in the circulating system; these slivers were also likely pumped into the reservoir’s fracture system during the numerous well control bull-heading operations. These sharp grooves were beveled by hand grinding prior to drilling the Olivo well. Only two QCP rubbers were consumed during drilling 5,542 ft of that well’s UBD section, compared to eight used on Catalina’s 1,230-ft section.

Verbal communication is especially important, both on location as well as to the various offices. Qualified bilingual workers are helpful if not necessary when implementing new technologies requiring detailed attention. Multiple telephone / fax lines in a staffed communications trailer helped in sending and receiving information. A comprehensive wellsite data acquisition system provided by Inter-Tech helped gather / distribute operational data pertaining to water / gas injection volumes / pressures, oil / gas production, water return, downhole / surface annular temperature / pressures and other data.

Catalina Drilling Specifics

Prior to starting the Catalina well, the drilling contractor installed a top drive system, a third mud pump and a new string of 5-in., 19.5 ppf. S-135 drill pipe with NC-50 connections and Arnco 200 XT hard banding. The well was spudded using a 17-1/2-in. bit to drill a pilot surface hole to 1,963 ft. This interval was then reamed to 26-in. hole for running 20-in. surface casing. Cementing this string to surface was done through the 5-in. drill pipe stabbed into the float equipment.

Pilot hole. An 8-1/2-in. bit was used to drill out of the 20-in. surface casing shoe. This pilot hole was designed to cut oriented conventional cores in the La Luna and allow running of a complete suite of wireline logs, including: phasor induction resistivity, borehole compensated sonic, gamma ray, compensated neutron, litho density, formation micro imaging and seismic velocity survey. Build-up pressures and formation samples were also taken via wireline-conveyed MDT tools. This 2,847-ft interval to 4,810 ft was then reamed to 17-1/2 in., and 13-3/8-in. intermediate casing was run and cemented back to 1,850 ft.

The hole opening exercise was attempted with a single pass using a 17-1/2-in. PDC bit. Difficult operating conditions caused a premature failure of the PDC. A 17-1/2-in. hole opener with a 12-1/4-in. pilot bit provided unacceptable penetration rates. A single 17-1/2-in. roller cone bit drilled over the 8-1/2-in. hole proved to be the best hole opener method.

A 12-1/4-in. bit was used to drill out cement and the float shoe; it was then pulled. Another 8-1/2-in. pilot hole was drilled to evaluate, core and log the Rosa Blanca. This vertical test hole was plugged back with openhole, balanced cement plugs. A 12-1/4-in. bit then dressed off the cement to the kick-off point at 4,885 ft.

Kick-off, curve drilling, 9-5/8-in. liner. Directional tools were run to kick off the well and build the curve section. Quick turnaround on the pilot hole core and log analysis was accomplished, and the curve interval was adjusted accordingly to land in the proposed target.

A multiple-propagation-resistivity MWD tool was run with a positive displacement motor; drilling jars were included in the tool string. The curve was eventually drilled to 61° inclination at 5,979 ft MD through an overpressured shale section. Drilling fluid used was the same 8.9-ppg density as that used successfully in the pilot hole.

Drilling fluid was pumped at 750 gpm with the 6-3/4-in. MWD being the rate-limiting factor. Annular velocity around drill pipe was low at 150 fpm in the 12-1/4-in. hole. Multiple combination low / high viscosity sweeps were pumped, with torque and drag remaining within the expected range.

During a bit trip, the hole began caving in and cuttings were swabbed back into the vertical casing where they bridged off. After a casing clean-out trip with a rotary assembly, it became difficult to pass the directional BHA beyond the KOP. Drilling fluid density was raised to 10.2 ppg. Once drilling again commenced, an MWD failure after drilling only 62 ft required another trip that indicated further deterioration of the Simiti shale. The next trip was difficult and, unfortunately, the replacement MWD failed to function on bottom. Another trip was made at 6,021 ft, and a hole opener run was made. Barite supplies were low, preventing drilling fluid density increases.

The next MWD was a 7-3/4-in. to allow higher circulation rates at lower pressures; however, hole conditions were too severe to reach bottom. Drilling fluid was pumped at 870 gpm with 6-in. pump liner pressure then the limiting factor. Annular velocity around drill pipe was 175 fpm. Another attempt at a hole opener run was unsuccessful and it was decided to abandon that build curve interval.

Barite was replenished, fluid density was increased to 12.2 ppg, and the curve was re-drilled to 5,893 ft MD and 60° inclination into the hard Tablazo limestone. A 9-5/8-in. liner was run and cemented; however, the liner top seal test held only 750 psi. A test packer was run and verified integrity of the 13-3/8-in. casing. Retest of the liner top proved successful. A 9-5/8-in. tie-back string containing a seal assembly followed by a perforated joint was run and stung into the polished bore receptacle above the liner hanger.

Underbalanced drilling. Over the next few days, the UBD equipment was rigged up, drilling fluids were changed over to clear potassium chloride (KCl) brine, and nitrogen injection was started to provide underbalance conditions.

An 8-1/2-in. bit was used to drill out and finish the balance of the curve section plus the horizontal lateral. MWD logging provided data on: geosteering, geological control, subsurface annular pressure, and downhole weight on bit and torque. Underbalanced conditions were maintained by continuously injecting membrane nitrogen into the 13-3/8-in. by 9-5/8-in. concentric casing annulus. Drilling fluid pumped down the standpipe was 8.4 ppg (3%) KCl water. This concept allowed conventional pulse MWD equipment to be used while still providing underbalance hydrostatic pressures while drilling ahead.

The Basal Limestone section consists of both the Tablazo and Rosa Blanca limestones. The primary target of the wellbore was the vertically fractured Rosa Blanca zones identified in vertical pilot hole cores and logs. However, the first kick was taken within the Tablazo and was not identified on the FMI log as a fractured interval. This section was not cored but is now interpreted as being a horizontal rubble zone similar to those identified in other core portions.

Handling kicks. The Tablazo kick was strong; casing pressures were over 1,200 psi, and the well was difficult to control. Heavy brine slugs were bull-headed down the annulus for temporary well control and all trips were made stripping through the RBOP. The surface separation equipment requires that proper fluid levels be maintained; however, they were difficult to stabilize due to the gas and foaming oil. Gas carry-over at the skimmer tanks was severe and on several occasions required shutting in the well due to safety concern. A safety coordinator on location constantly monitored conditions and had authority to shut down the operation.

Because of the increased gas carry-over, equipment inspections were made, at which time it was discovered that there were several broken baffle plates inside the mud gas separator. Repairs were made, equipment was modified, additional equipment was brought in and drilling continued without incident. The four 10,000-bbl oil storage tanks and twin 8-in. flare stacks were necessary, as production while drilling often exceeded 10,000 bopd, with two 60-ft flares burning.

As drilling progressed into the Rosa Blanca, several additional kicks were taken after drilling more of the rubble zones. At one time, cuttings samples were 100% free calcite crystals, indicating wide, naturally open fractures. Fracture faces and free calcite were identified, with good hydrocarbon shows throughout the short horizontal lateral.

Clear water KCl drilling fluid at 8.4-ppg density was pumped down the standpipe with nitrogen pumped down the 13-3/8-in. by 9-5/8-in. casing annulus. BHP in the Catalina well was normal for this depth at 2,650 psi, but the well demonstrated its ability to kick and come in many times while tripping. During trips, a top head of 9.5-ppg KCl water was used to lower surface pressures and avoid pipe light situations. However, during trips, the well was not able to seek its own pressure balance fluid level due to the variety of fluids mixed in the wellbore and gravity segregation. The bull-headed heavy slug would control the well for only a couple of hours, after which annular pressures again began increasing. All trips were made by stripping though the RBOP.

Situations also arose in which kicks were taken with drill pipe out of the hole. To start the pipe back into the well, a 9.5-ppg KCl fluid slug was bull-headed and the pressure was subdued long enough to open the blind rams and run the first joint to where the RBOP could seal on the motor housing.

Lost circulation concerns. Due to the open fractures and rubble zones, pore pressure was the same as fracture pressure. Lost circulation was common, resulting in excess of 24,000-bbl drill water losses to the formation. Additionally, prior to stripping out the drill pipe, a slug of 9.5-ppg brine was bull-headed, again resulting in losses.

Because the well flowed strong and surface equipment had a difficult time keeping up with gas separation, the designed 300-psi underbalance could not be achieved. Annular BHPs were usually about 2,600 psi, with equivalent circulating density about 2,900 psi. This resulted in stuck pipe when pressures rose above 2,650 psi, and lost circulation at anything higher. At one point, the drilling fluid contractor ran out of KCl, and operations were stopped for two days.

Fig. 4 is a diagram of the planned vs. actual well plot; and Fig. 5 is the time distribution chart for Catalina and Olivo wells.

Underbalanced completion. Testing the Catalina well using three, 3-phase separators took place while flowing up the 9-5/8-in. by 5-in. drill pipe annulus, through the BOP stack, with 5-in. drill pipe and 8-1/2-in. bit positioned at the 9-5/8-in. liner casing shoe for use as a kill string, if necessary during testing. Stock tank gauge measurements were used in addition to separator flowmeters. Three-inch lines between choke manifold and test separators proved to be the choke point during the test. Once testing was completed, the well was controlled using 9.5-ppg KCl water, and the drill pipe was stripped out.

A 9-5/8-in. by 4-1/2-in. permanent production packer with a wireline retrievable plug in place was then stripped in on drill pipe and set inside the 9-5/8-in. liner beneath the point of nitrogen injection. This isolated the productive 8-1/2-in. openhole interval, which was left uncased as an openhole completion. A retrievable bridge plug was set just above the permanent packer as a secondary isolation point prior to nippling down the BOP stack and nippling up the tree.

The perforated joint in the 9-5/8-in. tieback casing was left in the well for possible future use to extend the lateral or deepen the well. The bridge plug was then pulled and the 4-1/2-in. tubing string was run in the well. The wireline plug was left in place for protection during the drilling of the next well.

Olivo's Success In La Luna

The rig was skidded 30 ft over to drill the Olivo 1 well targeting the La Luna formation (1,300-ft shallower than the Rosa Blanca) that was also identified through pilot hole cores, modular dynamic pressure testing and logs as having open fractures and rubblized pay. Drilling of the Olivo well required minimal mobilization charges, no location or road cost, no cores, no electric logging and no storage tank costs.

Again, by utilizing technologies associated with underbalance drilling and clear water fluids in these naturally fractured limestone reservoir rocks, great success has been the result, in an area where only three of 21 offset wells were producers in the La Luna. Open fractures and free calcite were identified in the drill cuttings samples throughout the 5,542 ft of openhole lateral section.

Unlike the Catalina well, which drilled only 1,230 ft of horizontal section, Olivo 1 drilled 6,029 ft of displacement from vertical enroute to the MD TD of 10,215 ft. Lateral length was limited due to the inability to slide drill at this extended reach; however, rotary drilling was able to continue at ROPs up to 100 ft/hr. As lateral drilling progressed, oil was present in the returning drill fluid flow stream; however, different than the Catalina well, the Olivo was not a full-time strong well-control event.

The wellbore was geosteered within a primary and secondary target window thickness of about 30 ft. At a measured depth of about 9,953 ft, the ability to slide was reduced to less than 10 ft/hr. The drilling team anticipated this and had positioned the wellbore near the bottom of the target, thus allowing for the natural rising tendency of the rotary mode to continue drilling within the pay interval for another 262 ft.

A well test was organized with the intent of having an option of including a multilateral sidetrack in a lower, previously identified payzone. The well produced at exceptional rates, and therefore was considered to be at total depth.

The initial well test was made via an electric submersible pump (ESP) run on drill pipe through the RBOP and BOP stack with the RBOP inner packer sealing against the ESP’s electric cable. The annular preventer was tested as sufficient back-up but, due to the chance of damaging the cable, it was not energized.

Using equipment immediately available, the pump test yielded only 5,664 bopd. As an additional test, the ESP was pulled and open-ended drill pipe was run to the casing shoe at 60° inclination. Nitrogen was injected down the drill pipe and production was returned up the annulus. However, the cooling effect of the nitrogen increased the viscosity of the oil too much, resulting in a reduced well test.

A second ESP test was conducted while also injecting diesel down the annulus as a viscosity-cutting diluent. This test resulted in a net crude production rate of 10,800 bopd. The PI was calculated to be as high as 46 bbl/day/psi, resulting in a wellbore deliverability estimated to be 58,000 bopd, assuming the ESP is set at the top of the curve, above the KOP, in the vertical section.

Drilling to the casing point. The learning curve was rapid as lessons learned at Catalina were immediately applied to Olivo. The 26-in. hole section was drilled in a single run with a full-size bit. The 17-1/2-in. hole section was not going to be cored and logged so it also was drilled with a full-gauge bit. No pilot hole was required since the Olivo is only 30 ft away from the Catalina pilot, so no logs were taken. Prior to drilling out the shoe of the 13-3/8-in. casing, a gyro was run to confirm true inclination / azimuth.

The well was drilled 97 ft to the KOP and the build curve BHA was run. In the build curve interval, the multiple propagation resistivity tool was judged to be excessive data and unnecessary beyond data recorded by simpler tools. By eliminating the MPR, costs were saved from both the operating rate and the additional lost-in-hole insurance coverage required for this expensive tool.

The curve interval was drilled with a combination of rotary and slide techniques to achieve the desired 6°/100-ft build rates. Several mud pump failures reduced drilling fluid flowrates to 850 gpm; however, multiple sweeps of polymer and gilsonite succeeded in providing clean hole conditions. An 8-in. PDM and a navigation MWD with a gamma ray provided the bottomhole assembly with higher possible flowrates, compared to the Catalina well, and yet also provided enough logging while drilling data to allow accurate determination of the marker beds.

The well was drilled with no problems to the real-time identified 9-5/8-in. casing point at 60° inclination. The casing point was selected based on borehole stability calculations and provided a shoe in the top of the target limestone reservoir.

Drilling the horizontal section. For drilling the 8-1/2-in. hole section, the BHA equipment included resistivity at the bit and downhole measurements of annular pressure, weight on bit and torque. Data obtained from wireline logs in Catalina were used to model the Olivo horizontal target area. From MWD data recorded in the 12-1/4-in. curve interval, bed dips were determined to be 3.5° to 4°, with the wellbore drilling upstructure. At these revised dips, the target landing area would rise 50 to 65 ft TVD above expected. The beds appeared to be thinning dramatically as the top of the structure was approached.

The BHA for drilling the curve between 60° and 90° inclination was adjusted to a 1.8° bend so that the build rates would be 10°/100 ft. The high deflection would result in the inability to rotate the MWD tools while located in this section. The BHA built up to 14.5°/100 ft and was used until 4,861 ft MD and 85.5° inclination.

The BHA was tripped and adjusted to a 0.9° bend and a PDC bit was picked up. The BHA could not be rotated and was slide drilled 64 ft to bury the delicate parts of the BHA in low-deflection angle hole. Rotary drilling commenced at 87.5° inclination and the primary target was entered 255 ft later at 88.5° inclination.

Over the next 360 ft of measured depth, the target was traversed top to bottom and identifying markers noted. The wellbore was turned up to 95.5° inclination and steered back through both primary and secondary targets. This maneuver allowed confirmation of the geologic lay of the structure as well as formulation of plans for traversing the section and re-entering the target zones on the opposite downdip plunge of the anticlinal feature.

Numerous gamma ray markers along with seven distinct zones of varying resistivity were used to discern structural correlation. The real-time MWD logs — combined with detailed geologic description of the drilled cuttings, tracking the ROP as well as monitoring oil / gas production at surface (underbalanced drilling techniques) — allowed the wellbore to be drilled within the two-target interval (30 ft TVD). As the crest of structure was drilled, the bedding plane dips turned to 4.5° downdip. Wellbore inclination was adjusted to as low as 82° to chase the target down the structure.

At 9,200 ft MD, the wellbore reached its lowest structural position. Near this depth, the ability to steer was impaired due to the ROP being reduced to less than 10 ft/hr in the slide mode. This low ROP was unacceptable due to increasing lost circulation, difficulty in achieving significant underbalance and increased incidence of stuck pipe. However, in the rotary mode, the BHA had a slight rising tendency (0.4°/100 ft) while still drilling at 100 ft/hr.

When the slide mode limiting factors started to become insurmountable, the wellbore was positioned near the bottom of the lower target. This allowed drilling to continue in the rotary mode, traversing back up into the target without requiring course corrections. Total depth was declared after the well reached its 6,000-ft vertical section (VS) mark. TD is recorded at 10,215 ft MD, 4,854 ft TVD and 6,029 ft VS. Fig. 6 is a diagram of the planned vs. actual well plot. The time distribution chart for Olivo 1 is shown in Fig. 5.

High-volume completion. Olivo was tested initially with a submersible pump run on the drill pipe through the BOP stack. The RBOP’s quick change packer was pulled and the softer, more-ductile rubber element of the inner packer was used to seal the drill pipe annulus. Using the inner packer allowed a seal with the ESP electric cable installed. This initial test confirmed substantial flowrates with almost no reduction in annular fluid level.

The final completion included 4-1/2-in. tubing, four gas-lift valves, a chemical injection capillary tube and injection sub, a dual packer and the ESP. Unlike the Catalina well, Olivo was completed conventionally; however, the well is capable of flowing naturally and therefore required installation of a retrievable bridge plug below the tieback prior to nippling down the BOP stack and nippling up the tree.

The Future

Proper well planning and training undoubtedly enabled successful completion of Colombia’s first horizontal underbalanced and extended-reach wells. A key factor in this was total corporate management support of the operator in all aspects of this drilling project. Proper training proved essential to execution of these accident-free jobs.

As with any new drilling project, many different services are combined to achieve the desired goals. Teamwork and logistics coordination become major objectives. Coordinating these "people skills" can easily become more difficult than the technical engineering details.

On a broader view, success of this effort in the Rosa Blanca, with just 1,230 ft of exposed interval in zone, plus the extended reach achievement realized in the La Luna formation, opens the E&P frontier in vertically fractured limestone reservoirs throughout Colombia to horizontal underbalanced drilling.

These first wells, with impressive production rates, should lead to many subsequent wells. The concept that works so well in similar formations around the world has now been shown to be effective in Colombia, and can possibly benefit other South American countries that have vertically fractured limestone formations.

Acknowledgment

The authors would like to thank Harken de Colombia’s President, Guillermo Sanchez, his operation manager, Alvaro Purerta, and his assistant operations manager, Hector Armando Espinosa. Additional credit and thanks go to Harken International’s President, Steve Voss, who also edited this article, the engineering manager, Joe Alack, and Inter-Tech Drilling Solutions’ VP of operations, Joe Kinder, for their strategy and management support for this project.

The author

Scott Cooper, senior vice president of engineering at Harken International, Ltd., Houston, received a BS degree in petroleum engineering from Texas A&M in 1983. His background includes working as an operator, for a service company and as a consultant. He previously held senior positions with Baker Hughes INTEQ (Solutions), Joshi Technologies International, as well as two independent international operators. Mr. Cooper is a member of the SPE.

Bob Cuthbertson, UDB Engineering Manager of Inter-Tech Drilling Solutions, Ltd. graduated with honors from the University of Texs with a BS degree in petroleum engineering. He began his career with Exxon USA, and he has more than 25 years industry experience. Mr. Cuthbertson is a registered professional engineer in Texas, and a 25-year member of SPE.

Copyright © 1999 World Oil
Copyright © 1999 Gulf Publishing Company