SLIM HOLE DRILLING-CONCLUSION WELL CORED TO 9,800 FT IN PARAGUAY

Kenneth B. Gunn
Texaco Inc.
Latin America/West Africa Division
Coral Gables, Fla.

The mining industry's slim hole drilling rigs have proven applicable to primary oil exploration. These "lean, mean drilling machines" are smaller than conventional drilling rigs and can be transported with relative ease to remote locations.

(Editor's note: The first of these two articles on slim hole drilling described mining industry core drilling equipment and techniques and appeared in OGJ, May 6, p. 85.)

A typical rig drills an entire well by coring, with the cores retrieved by wire line without tripping the pipe. The core drilling system is specially suited to drilling hard rock formations.

During the first half of 1990, Texaco Inc.'s Latin America/West Africa division drilled a deep core hole, the Mallorquin Well No. 1, as a stratigraphic test.

The project evaluated the geological aspects of the Parana basin and determined the applicability of slim hole, core drilling techniques as an exploration tool.

The Parana basin is found in the eastern third of Paraguay, part of northeastern Argentina, and part of southern Brazil. Much of the basin is overlaid by basalt flows up to 5,000-ft thick, and there are numerous igneous intrusions and dikes within the sedimentary section.

This combination makes seismic quality poor and interpretation extremely difficult. The formations are relatively old, with Triassic red beds occurring only a few feet below the surface or immediately below the basalt. Beneath the Triassic are Permian marine deposits, Permo-Carboniferous tillites, and then Devonian, Silurian, and Ordovician deposits to the basement. The section outcrops 100 miles west of the Mallorquin Well No. 1 site.

The Parana basin has been only randomly explored. To date, success has been limited to a minor gas find near Sao Paulo, Brazil.

CORE HOLE DRILLING

Core hole drilling differs significantly from conventional oil field drilling. The following are some of the more important characteristics of a typical operation:

Essentially all core drilling rigs employ some form of top or side drive. The units can pull up or down on the drillstring in the same manner as oil field snubbing units.
Most core drilling rigs are hydraulic, with the top drive, mud pumps, and hoisting equipment operated through power oil systems.
Cores are retrieved by pulling the inner core barrel through the drillstring by wire line, necessitating double-drum drilling units. Core barrels and retrieval tools have lock dogs and fishing necks similar to oil field wire line tools.
While a core is pulled, the danger of swabbing a well is reduced by pumping down the drillstring, against and around the core barrel.
Drill rods, the equivalent of oil field drill pipe, are 20-ft (6-m) long and internally upset. Drill rods are only slightly smaller in diameter than the hole itself. Annuli are typically 1/4 in.; 1/2 in. is considered large.
Because there is no room for buckling, the drill rods can have thinner walls than oil field drill pipe. The internal upset retains joint strength while the pipe body allows considerable fluid bypass around the core barrel during tripping.
Drill rods are normally rotated at 300-600 rpm, yet some rigs are capable of rotating rods at speeds up to 2,000 rpm. Cost per foot is more dependent on rotation speed than weight-on-bit (WOB).
A typical drilling combination is 500 rpm and 3,000 lb WOB. On a pound per square inch basis, the 3,000 lb is equivalent to running 25,000 lb on an 8-1/2 in. bit in a conventional well.
WOB is obtained by drilling with the bottom portion of the drillstring in compression, thereby eliminating the need for drill collars. Buckling and joint failure associated with drill pipe compression are not problems because of the support gained from the borehole wall.
Diamond-set core bits, similar to oil field core heads, are standard, but diamond-impregnated cast bits are gaining acceptance for hard formations. The drill bit kerf is very narrow, commonly 1/2-1 in.
This makes cutting more efficient because less formation is actually drilled up, and speed variation between the inside and outside of the cutting face is reduced.
Mud circulation rates are low, normally near 50 gpm in a 5-1/2 in. hole and 12-15 gpm in a 3-in. hole. Low rates are necessary due to high friction losses in the small annuli. Hole cleaning is not a problem because relatively few, and only very fine, cuttings are generated.
Even though low pump rates generally keep the mud in laminar flow in the annulus, equivalent circulating density (ECD) increases 1-3 ppg as a result of friction. Further increases up to 2 ppg may result from the high speed rotation. 1
Very few drilled or added solids can be tolerated in the mud. The high speed rotation of the drill rod creates a centrifuge effect, forming mud rings on the inside of the drillstring.
Furthermore, even a thin filter cake can cause high torque.
Because annular volume is about one tenth that of conventional systems, well bore influx during a kick must be minimized. If the well kicks while coring, a dynamic kill method is employed. The driller speeds up the mud pump, and the resultant increase in ECD effectively kills the well. The kick is then circulated out, additional mud weight requirements are calculated, and a control scheme is implemented.

SLIM HOLE RIG

The Mallorquin Well No.1 was drilled by Longyear Co. rig PM 603, a 400-hp, skid-mounted core rig rated at 10,000 ft. The rig is completely hydraulic with electricity used only for lighting and instrumentation.

The rig uses three independent power oil units. The smallest of these drives the centrifuge and part of the mud system. Another unit powers the remainder of the mud equipment and the mud pumps. The third and largest system is located on the drilling rig itself and can drive any combination of five separate pumps, each assigned to a major function on the rig. These five principal components are rotary drive, draw works, drive head raising rams, sand line winch, and the tongs and slips.

Torque is transmitted from the drive head to the drillstring through an hydraulic chuck closed around the kelly. The head contains a manual, six-range transmission with low-load rotational speed capability up to 900 rpm. Clamps are placed at 10-ft intervals on recesses in the kelly and ride on top of the drive head to support the drillstring. It is necessary to remove the clamp and rechuck every 10 ft even though the kelly is 40-ft long. Although this seems cumbersome and time-consuming, it does not result in significant time loss.

Drill rod is normally added to the string when rotation is stopped to pull a core.

The rig has two small triplex mud pumps. One pump was fitted with liners rated at 173 gpm and 1,200 psi and the other at 108 gpm and 2,000 psi. The pumps were used in tandem during the surface hole drilling to increase circulation rate.

The mud system consisted of four 60-bbl tanks mounted two per skid. The first tank was divided into three sections of 10 bbl each and a 30 bbl sand trap. The other three tanks were undivided and had hydraulic agitators for solids suspension.

Mud cleaning equipment included two 24-in. standard shale shakers, a 100-gpm centrifuge, and a bank of two 6-in., two 4-in., and four 2-in. hydrocyclone desilters. Mud was mixed and sheared in a Sidewinder vortex mixer.

Well control equipment included a Shaffer LWS double-ram BOP, a Hydril GK annular BOP, a McGuire high-speed rotating head, and a Cameron hydraulic valve. The choke manifold consisted of a manual choke, a Thornhill-Craver hydraulic choke, and an open line. The choke and kill lines were tied into the middle outlets on the double ram unit (Fig. 1).

Three sizes of drill rod and core barrel were supplied: CHD 134, CHD 101, and CHD 76. Table 1 lists common specifications for the drill rods, bits, and core barrels.

The rig contractor provided all bits and downhole equipment, fuel and lubricants, coring equipment, and mud engineering services. Texaco provided drilling mud, casing, cement and cementing services, electric and mud loggers, and core handling services and facilities.

Longyear rig PM 603 was mobilized out of Salt Lake City, Utah. It was trucked to Houston and consolidated with five containers of Texaco's supplies and mud. The combined shipment was sent by ocean transport to Montevideo, Uruguay, then transferred to a smaller vessel for the upriver journey to Paraguay. It was off-loaded at the port of Villeta on Feb. 26, 1990, and trucked 270 km to the well site.

One 25-ton and one 50-ton crane were leased locally for unloading and rig up. The rig was ready to spud 8 days after arrival at the dock.

OIL FIELD SERVICES

Cementing services were provided by Halliburton Services. The pump truck, cementing heads, and tools remained on location throughout the drilling operation, and personnel were brought in when needed.

Western Atlas International Inc. received the electric logging contract because of its excellent suite of small diameter tools. The logging truck and equipment remained on site, and the crew was brought in for each logging operation. Vibrator units and vertical seismic profile equipment were mobilized from Argentina for the final logging run.

Mud logging services were contracted to Exlog. The mud logging unit was rudimentary, consisting only of a gas detector, pit volume indicators, and a kelly height graph. Two geologists, working 12-hr shifts, manned the unit and provided core descriptions on a mud log format.

Core handling was contracted to Core Laboratories. The equipment consisted of a drill press, slabbing saw, gamma ray recorder, and cameras. Two technicians, also working 12-hr shifts, ran the equipment and were responsible for marking, orienting, slabbing, photographing, and boxing the cores. Wooden core boxes were fabricated locally, each holding 10 ft of CHD 134 or 20 ft of CHD 1 01 or CHD 76 slabbed core.

Halliburton, Western Atlas, and Exlog units were driven from their bases in Bolivia and Argentina.

DRILL SITE

The exact location of the well was not very critical for this exploration well. "A flat spot near the road" was located to minimize construction costs.

The well location measured 65 m x 90 m, slightly less than 1-1/2 acres.

Local labor was used to dig and cement the cellar and to lay rough-cut, 3 x 10 in. lumber for the rig mat. A 10 ft x 20 ft shed was built to house core slabbing equipment.

Conductor, rat, and mouse holes were drilled and cased using a local water well driller before bringing in the core rig. The rat and mouse holes were drilled to 40 ft and lined with 7-in. casing. The conductor hole was drilled to 60 ft, and 9-5/8 in. casing was run and cemented.

SURFACE HOLE

The surface hole was drilled conventionally to 1,210 ft. The bottom hole assembly consisted of an 8-1/2 in. tri-cone bit, 6-1/2 in. drill collars, and blade-type stabilizers run on CHD 134 drill rod (5 in. OD). A 6-in. rotating head, flanged to the 9-5/8 in. conductor pipe, served as a diverter.

This section was drilled with bentonite and water. Mud returns were forced through the flow line to the shale shakers by closing the rotating head. In general this arrangement was not satisfactory and will need to be revised on future operations. Additionally, the low circulation rate made it necessary to restrict penetration to avoid overloading the hole.

A full suite of wire line logs, consisting of induction, acoustic, gamma ray, caliper, neutron, and density was run. After logging, 7-in. casing was set and cemented to surface. The conductor pipe was cut off at the bottom of the cellar and a Seaboard-Arval RT casing head was installed on the 7-in. casing.

FIRST INTERMEDIATE SECTION

The surface casing float collar and shoe were cored out with a 5-1/2 in. core bit and 20-ft core barrel on CHD 134 drill rod. The well was then cored continuously to 4,924 ft, believed to be a world record for a CHD 134 wire line retrievable coring system.

Core recovery over this section, 1,210-4,924 ft, was 99.8%; just under 30 days were required from drill out to logging.

After the pressure limits test, all mud was dumped, and a non-dispersed, low-solids system was mixed. Partially hydrolyzed polyacrylamide (PHPA) and polyanionic cellulose (PAC) polymers supplemented with starch were used to obtain viscosity and fluid loss control. Lubricants were added to reduce friction and should probably have been used more liberally. Very few other additives were required.

The inner core barrel, with 3.344-in. core, was tripped by wire line every 20 ft. As the full barrel reached the surface, it was laid down and a second barrel dropped down the well. Coring resumed when standpipe pressure indicated the replacement core barrel was on bottom.

While the empty barrel was pumped down and coring restarted, the retrieved core was handled on the surface. It was pumped out of the barrel into a trough where it was cleaned, marked, and geologically described. The core was next taken to the handling shed for gamma ray logging.

It was then slabbed vertically and boxed--one half for the Paraguayan government and the other half for Texaco. The cores were video taped and still photographed to provide a portable record.

Throughout this section, torque was higher than anticipated. As a result, maximum attained rotary speed was only 400 rpm, and much of the interval was drilled at only 200-250 rpm. After eliminating other possible causes, it was decided that a contributing factor to the high torque may have been the relatively large ID of the 7-in. casing.

The 11/16 in. in excess of bit OD and the abrupt change in diameter at the shoe were suggested by the contractor as a possible cause of the excessive friction in the system. Relatively high plastic viscosity (PV) and yield point (YP) may also have contributed to the torque problem.

Two full suites of logs were run in this section of hole; the first near 3,300 ft and the second at casing point. The caliper logs indicated the hole to be in gauge throughout. The CHD 134 drill rod was made up with conventional float equipment and run as casing.

The bottom of the string was cemented with 4 bbl of slurry, preceded by a 2-bbl cement wash. The volume was considered the minimum necessary to obtain a consistent mix. Because of the small annulus, the 4 bbl of cement covered a significant length of casing. Nevertheless, it was pumped even though it was planned to recover the casing at abandonment. Later, a batch mix procedure was developed, permitting smaller cement volumes to be mixed satisfactorily.

Overall drilling time for this section went on schedule. However, significant improvement should be realized on future wells when the high torque problem is resolved.

SECOND INTERMEDIATE SECTION

After nippling up on the 5 in. casing (CHD 134 drill rod), a CHD 101 string was run to drill the second intermediate section, 4,924-8,718 ft. The CHD 101 bottom hole assembly had a 4-1/16 in. core bit and 3.875-in. core barrel to cut 2.50-in. cores.

While drilling through the float equipment, a valve washed out in the standpipe, permitting mud to circulate at surface without going down the hole. This was not detected very quickly, and heat generated by the bit caused the drillstring to fail 5 ft off bottom. The resulting fish was retrieved 5 days later by coring over it; a 6-in. piece of melted steel with chunks of cement was recovered intact.

Progress over this section was slower than projected. Formation rocks were considerably harder and much finer grained than expected. Surface-set diamond bits yielded high initial penetration rates but dulled quickly, becoming uneconomical.

Impregnated bits, made of diamond chips in a cast matrix were then tried. As the chips on the surface of the bit dull, the matrix is designed to wear off, dropping the dull chips and exposing new ones. In the Mallorquin No. 1 well, however, the very fine formation particles tended to polish rather than strip the matrix. Drilling progress was slow with considerable time spent each time the matrix required stripping, which was accomplished by varying the WOB. The polishing tendency may be reduced by strict adherence to the bit rpm and weight guidelines.

A soft-matrix bit recently developed for mining applications may prove practical for this type of oil field situation.

In this section, low rotational speed caused by high torque, possibly a result of high mud flow properties, contributed to slower-than-expected progress. More horsepower at the rotary drive head and stouter equipment would also have been advantageous.

Throughout this interval, core was retrieved in 40-ft lengths. When the core barrel reached surface, it was pulled clear of the drillstring and lowered into a 40-ft, U-shaped carrier in the mouse hole. The carrier, barrel, and core were then pulled together. A guided trolley was fastened to the bottom, and the combination was lowered to the catwalk.

The full core barrel was exchanged for an empty one, the handling procedure was reversed, and the empty core barrel was dropped down the drillstring. The surface handling operation took less than 5 min. The total core trip time from 8,500 ft averaged less than 2 hr.

Midway through this section, the weight-supporting bearing in the top of the rotating head failed. This failure recurred several times, resulting in excessive lost time. Longyear and Texaco personnel isolated the problem--the bearing itself was undersized. Thus, it was necessary to cease drilling with the CHD 101 pipe and switch to a smaller and lighter CHD 76 system. A full suite of logs was run, and the calipers again indicated the hole to be completely in gauge.

After logging, an additional 53 ft was cored prior to setting the CHD 101 drill rod as casing. Because the string was already on bottom, the pipe was cemented open-ended rather than tripped out for a float collar and shoe. Three attempts were made to tack cement the pipe in place, but each was under-displaced and therefore unsuccessful. The uncemented string did not present a problem, and drilling proceeded without incident.

This was the longest single section of the well in terms of interval and drilling time. It also provided the most severe test for the equipment. The section took 47 days from the casing seat test to logs (average of almost 80 ft per day).

Although performance was somewhat disappointing, the Texaco personnel gained considerable knowledge of the equipment and procedures for subsequent operations. The contractor also identified specific areas where the rig was not "oil field tough" and may now upgrade key components to improve performance. Most of the slow progress was attributed to the abnormally hard, difficult drilling conditions.

BOTTOM SECTION

The bottom section of the well, 8,718-9,811 ft, was cored with CHD 76 drill rod. The bit diameter was 3-1/32 in., and the recovered cores were 1.713 in. OD.

Typically, 40-ft long cores were cut, except for one section where jamming became a problem. Because the shorter (20 ft) core barrel is more rigid, it was less subject to jamming where the severely splintered formation was encountered. During this interval the shorter barrels were used.

The same surface lay down system was used for the CHD 76 cores as for the CHD 101 cores. No problems were encountered drilling out the landing ring, stabilizers, and the inner gauge of the core bit left on the bottom of the CHD 101 "casing" string. The drill out operation was completed in 1 hr.

Drilling proceeded routinely to 9,81 1 ft. The decision to stop at this depth was based upon time constraints rather than geology or rig capability. A well bore diagram showing the hole configuration at final total depth (TD) is shown in Fig. 2. This depth is a record for both Texaco and Longyear, but not for the industry overall.

The bottom section of the hole averaged 70 ft/day. This was somewhat less than expected and mainly attributed to the extremely hard formations. The smaller and lighter cores were consistently tripped from 9,500 ft in 1-1/2 hr.

Rotation speeds of 450500 rpm were used throughout this section with 4,000-6,000 lb indicated WOB. On a pound per square inch basis, this weight is equivalent to 50,000-70,000 lb on an 8-1/2 in. bit.

An open hole gamma ray and a vertical seismic profile were run at TD. The CHD 101 string was pulled intact and used as a work string to set abandonment plugs up the hole. The CHD 134 string was backed off at 1,864 ft using a string shot, and the topmost drill rod was recovered and laid down. The rig was released on July 7, 123 days and 5 hr after spud.

RESULTS

Average penetration rate was slightly over 79 ft/day. This was 79% of projection and 60% of target. Fig. 3 shows the actual and target drilling time curves for the Mallorquin No. 1.
A total of 8,601 ft of core was cut; overall core recovery was 99.4%.
Cost of the well, including core analysis and office overhead, was $3.6 million (8.5% over budget). The small overruns were a result of the harder-than-expected drilling,
The estimate for conventional drilling was $4.5 million; however, Texaco believes conventional drilling may have been unsuccessful in reaching TD.
A world depth record is believed to have been set for a CHD 134 wire line retrieved core system--4,924 ft. Texaco and Longyear records were set for any size wire line retrieved cores--9,811 ft.
The rig and all Texaco equipment shipped to Paraguay at 721 pay tons. This compares to approximately 5,000 pay tons for a conventional rig.
The well was in gauge (a gun barrel) over all the cored section.
Slabbing the core on site provided an excellent face for observation and photographing.
During the fishing job a cement plug was set on the top of the fish before coring it out. This was found to be an extremely effective way of cleaning up junk in the hole and may also find applications in a conventional drilling operation.
In spite of pulling only 60-ft stands, the rig tripped pipe at roughly 1,000 ft/hr.
By cutting off and recovering the top of the CHD 101 casing string then using a CHD 101/76 mixed drillstring, depths near 14,000 ft may be attainable.
A low-density, low-solids mud appears to be a requirement for continuous coring. Small amounts of bentonite may be added, but care must be taken to ensure that the filter cake is kept thin and tough. PV and YP should be kept as low as possible to minimize torque.
Further experimentation and study are needed to determine the optimum circulation rates and fluid properties. Table 2 shows the range of mud properties throughout the cored sections of the well.
Two primary considerations of a core hole mud program are: Flow and friction properties are interrelated and critical, and the total volume of the system is so small that large per barrel additions usually do not add significantly to the well cost.
Excluding the surface hole, the centrifuge did an excellent job of cleaning the mud with solids less than 2% over essentially all the cored section.
The surface mud system can be reduced to two tanks instead of four, and mud cleaning can be achieved by use of a centrifuge alone. This is possible as relatively few cuttings are generated and all are quite fine. (No coarse cuttings are developed by the diamond inset or impregnated bits.)
The well control program was not tested during an actual situation, but the BOPs worked satisfactorily in programmed tests. The well control plan included numerous safety meetings and routine drills to ensure proper responses in the event of a kick.
The two areas of risk identified were danger of a blowout and the possibility of sticking the pipe. Considerable training of and communication with the contractor's drilling personnel reduced these risks.
With planned design changes, Longyear PM 603 should be capable of coring to depths of 12,500 ft with CHD 101 drill rod and 11,700 ft with CHD 76 drill rod. Danger of twist offs obviously increases as the depth capacities of the drill rods are reached, but no indication of such problems was seen on Mallorquin No. 1.

ACKNOWLEDGMENT

The author wishes to thank Texaco LA/WA for permission to publish this article, and Lester A. Brockmann and M. Pat Miller of the Texaco drilling department for assistance in its preparation.

REFERENCE

Bode, D.J., Noffke, R.B., and Nickens, H.V., "Well control methods and practices in small-diameter well bores," Society of Petroleum Engineers Annual Technical Conference and Exhibition, San Antonio, Tex., Oct. 8-11, 1989.