Wade Dickinson, Eric Dickinson
Petrolphysics Inc.
San Francisco
Herman Dykstra
Consultant
San Francisco
John M. Nees
Consultant
San Francisco
Ultrashort radius horizontal radials have enhanced oil production from a cycle thermal project conducted by the U.S. Department of Energy (DOE).
The project demonstrated that this innovative drilling and completion system is applicable in thermal operations in heavy oil reservoirs.
The radials improved the production stability in both the well with radials and offset wells. Production rate increased by two to four times. Gravity override was minimized or eliminated.
Furthermore, the radials provided access to new reserves previously precluded from the vertical well due to reservoir heterogeneity or low reservoir energy. Their penetration beyond near well bore damage enhanced the oil production rate.
The essentially flat production within the limitations of the heat input to the reservoir indicated stability.
HARDWARE
Ultrashort radius radials have a turn radius from vertical to horizontal of about 1 ft. This compares to about 30 ft for short radius drilling, and about 300 ft for medium radius. Multiple layers of radials are also possible with this system.
Typical hardware for this system is:
- An ultrashort radius whipstock erected in a 24-in. underreamed zone. Suitable underrearning is provided by mechanical or water jet tools (Fig. la).
- Coiled tubing, 1-1/4 in. drillstring. The long continuous joint is pushed into the earth by hydraulic force.
- Medium pressure, conical water-jet drilling at 5,00010,000 psi (Fig. 1b).
- Closed-loop control of the drilling trajectory to provide control while drilling (Fig. 1c). Rate of penetration is controlled by an independent hydraulic motion controller.
- Accurate positional surveying of horizontal drillstring trajectory by means of flexible wire line tools (Fig. 1d).
- Electrochemical downhole cutoff and perforation of the 1 1/4-in. tube.
- Horizontal, bidirectional gravel packing that is believed to be 100% -fill based on full-scale testing (Fig. 2).
- A slotting liner, flexible sand barrier with re-entrant helical slots for sand control (Fig. 3).
- Gravity reservoir drainage with vertical pumping for low-pressure reservoirs.
SITE SELECTION
An operator was sought that had both extensive experience in heavy oil operations and a competent geological, engineering, and operational support staff. A major oil company was selected because of its proven expertise in heavy oil operations in the San Joaquin Valley of California as well as the high caliber of the personnel that it was prepared to dedicate to the DOE project.
Midway-Sunset field was selected because of the long successful history of thermal operations there. The Midway-Sunset field is currently the second largest producing field in California. Over 15 major oil companies and independent producers conduct thermal operations in this field.
Many of these Midway-Sunset producers also have thermal operations in other San Joaquin Valley locations. The candidate well was on a lease that had existing cyclic producers for comparative analysis.
FIELD LOCATION
The Midway-Sunset field is in the southwestern corner of the San Joaquin Valley of Kern County. The field length is about 25 miles and has an average width of 3 miles.
The first recorded oil well in the Midway-Sunset field was drilled prior to the turn of the century (circa 1890). Discovered at a depth of approximately 100 ft, initial oil production was limited to a low gravity crude from the Tulare sands.
Currently, the field has approximately 10,000 producing wells with an average daily production rate of 16 bbl/well. During 1988, production from the Midway-Sunset field totaled 58 million bbl of oil. As of Jan. 1, 1989, the field had produced nearly 1.9 billion bbl of oil.
An estimated 400 million bbl still remain.
Oil gravity varies significantly throughout the field from a low of 8 API to a high of 35 API.
The Midway-Sunset field is a regional homocline which parallels the eastern flank of the Temblor uplift and trends northwest-southeast. The field has been structurally modified by a series of folds and converging unconformities.
Field production/reserve statistics as well as operatorship are often defined in terms of these major substructures.
Production from the Midway-Sunset field is primarily from the Miocene, although there is some production from the Pliocene and even from the shallower Pleistocene formations.
The upper Miocene sediments are approximately 8,000 ft thick and consist of a series of sandstone and shale formations. The middle and lower Miocene formations are also marked by a sand-shale sequence which reaches a gross thickness in excess 10,000 ft.
The upper Miocene is the major producing interval in the field. The Reef Ridge and the Antelope shale are its principal oil bearing formations. The DOE well produces from the upper Miocene Potter sands.
BREMER LEASE
The DOE-sponsored well, Bremer RI-53 was drilled on the Bremer Fee lease in the Midway-Sunset field. Operated by the a major oil company, the Bremer Fee is situated at the field's northernmost boundary.
The Bremer Fee lease provided offset wells for steam injection and oil production comparisons.
The Bremer Fee was also determined to be an environmentally sound location.
Initial production from the Bremer Fee began in 1968. Since that date, approximately 21 million bbl of oil have been produced.
Original oil in place (OOIP) is estimated at 92 million bbl. Twenty-three percent of the OOIP has been recovered to date.
Oil gravity for the lease averages 11.4 API. Due to its low gravity, thermal stimulation is required for economic production.
Average daily production per well from the Bremer Fee is less than 20 bo/d.
GEOLOGY
With the exception of limited production tests in the Marvic zone, all Bremer Fee oil production is from the Potter formation. The Potter, the most prolific sandstone within the upper Miocene Reef Ridge formation, is found at a drilled depth of 1,000-1,500 ft.
Known collectively as the Potter sands, the Potter formation is divided into four distinct zones: Potter A, B, C, and D.
The Bremer RI-53 was completed in the Potter C formation. As can be seen on a type log, the Potter C is a fairly clean quartz sandstone. Approximately 10% shale is present in this zone.
Discontinuous shale throughout the C interval. These shale zones may negatively impact steam distribution and oil productivity.
The C sand in the Midway-Sunset field has a porosity range of 9-45% with an effective average porosity of 20-25%.
Permeabilities vary sporadically with depth from a few milidarcies to 3 darcies.
The average porosity and permeability for the C sand at the Bremer RI-53 location are estimated at 25% and 1 darcy, respectively.
Structurally, the Potter dips at an angle of up to 35 in a northwest to southeast trending direction. The Potter C ranges in thickness from 450 net ft of pay to the east to 0 ft of pay at the updip pinchout to the west.
Because oil production from the Potter is dictated by gravity drainage, the relative structural position of Bremer RI 53 was a key factor in the site selection process.
THERMAL OPERATIONS
Major thermal operations in the Bremer Fee initially occurred from 1976 to 1977 when 60 producing wells and five continuous steam injectors were drilled.
The next major drilling activity took place in 1980 when an additional 46 producers were drilled on the lease. These wells were primarily cyclic producers.
Four years later, an additional 53 wells were drilled, resulting in current well count of nearly 220 producers.
Bremer Fee production peaked at 7,647 bo/d in August 1985. Oil production at year-end 1989 averaged approximately 5,000 bo/d.
The existing steam drive is located in the northeast portion of the field. This steam drive is isolated from the southwestern portion of the Bremer Fee where the wells are all cyclic producers.
Significant potential exists to enhance oil recovery in the southwest portion of the Bremer Fee in the area of the DOE well.
As of Jan. 1, 1989, recovery in the southwestern sector was only 18% of the OOIP. This compares unfavorably with a recovery of approximately 35% in the downdip steam drive portion of the reservoir.
Recent reservoir studies indicate that recovery from the lowermost Potter C zone may be particularly low. Temperature profiles show the basal portion of the Potter C to be particularly cold.
The temperature stratification within the Potter sand is due to a combination of long producing intervals and the gravity override of the injected steam. As a result, the oil in the Potter C is very viscous, and the recovery has been relatively poor.
One of the primary objectives of the DOE project was to stimulate production from the Potter C sand by injecting steam into horizontal radials drilled into the cold Potter C interval.
BREMER RI-53 WELL
The Bremer RI-53 well was drilled in April 1989 to a total depth of 1,200 ft. Eight radials were drilled in the Potter C sand of which four radials were completed with gravel pack.
Bremer RI-53 was unique to the Bremer Fee since it was only completed in a 30-ft section of the Potter C sand.
Typically, injectors and cyclic producers are completed throughout the entire 400-ft Potter C interval.
Two particular reservoir characteristics of the Potter C at the Bremer RI-53 location made it a good candidate for horizontal radials.
First, the Potter C zone at the RI-53 location appeared to be cold. Therefore, steam needed to be injected directly into the C to enhance production from this zone.
Second, due to its excellent vertical permeability, steam override generally occurs very near the vertical well bore in the Potter formation. Radials could be utilized to direct steam away from the vertical well thereby providing more heat to the surrounding Potter C formation.
FIELD OPERATIONS
Prior to radial placement the well was drilled to 1,250 ft. Then 7-in. 23-lb/ft casing was run to 870 ft. A guide shoe was placed at the bottom of the casing.
The guide shoe is a piece of casing that is slightly belled to allow the whipstock to reenter the casing easily when operations are completed.
After running the casing, the well was cemented using a Class G cement. The well was plugged back to 1,060 ft, leaving an open interval from 1,060 to 870 ft.
A perspective of the well radial placement is shown in Fig. 4.
Once the vertical well had been drilled, cased, and cemented, and the cement was given sufficient time to set, the well was underreamed.
These operations proceeded without event and were typical of drilling operations that occur in the Midway-Sunset field.
The Petrolphysics drilling technology employed at the Bremer RI-53 well requires the underrearning of a 24-in. diameter zone. The minimum height of this underreamed zone is 10 ft.
For this particular well, the company opted to underream a 30-ft high zone from 870 to 900 ft. The 30-ft zone allowed positioning of the whipstock elevation in the selected position and to provide sufficient clearance to allow the whipstock to be removed in the event that this soft unconsolidated formation were to collapse in the underreamed zone.
For Bremer RI-53, the underrearning was accomplished in four steps or changes in successive diameters: 7-13 in., 13-15 in., 15-20 in., and 20-24 in.
Each change in diameter required that the underrearning tool be tripped out of the hole and new arms installed to allow the next diameter to be cut.
Additionally, the underrearning tool body was changed after the 15-in. diameter pass was made.
Underrearning Bremer RI-53 took approximately 2 days.
The caliper log of the underreamed portion of Bremer RI-53 showed at least a 24-in. diameter throughout.
RADIAL PLACEMENT
After the vertical well was prepared, Petrolphysics moved its equipment onto the location. This included the wire line truck, high pressure frac pump, coiled tubing unit, and equipment trailer.
Once all the equipment was set up, the downhole assembly and tubing were run into the well. The 260-ft downhole assembly included the whipstock and motion controller tubing. A length of 995 ft, equipment and tubing, was needed to place radials at approximately 985 ft below the surface.
The remainder of the string length 725 ft was comprised of 4-1/2-in., 24.6-lb/ft Atlas Bradford TC4S tubing.
The radial placement operations at Bremer RI-53 required that the whipstock and high-pressure tubing string be run into the well twice.
The first time was prior to placing calibration Radials 1-3. The second time was prior to placing calibration Radial 4. Once the whipstock and high-pressure tubing were run into the hole to a depth of 985 ft, the whipstock was oriented with the gyroscope. The gyroscope was used to set the azimuth of each radial. This orientation was done prior to each radial being assembled and placed in the formation.
Eight radials were placed. The first four were short calibration radials and were not completed. The second set of four radials was completed for injection and production usage.
Fig. 4, a perspective view of radial placement, gives the orientations of the four radials that were completed in Bremer RI-53.
After the azimuth of each radial was established by the gyroscope, the radial tubes were assembled. This assembly included the joining of the conical jet nozzle, vertical control system, and the motion controllers to the 1-1/4-in. tubing.
The radial tubes were then ready to be run to the bottom of the high-pressure tubing string via the single conductor wire line.
Upon reaching the bottom of the high-pressure tubing string, the ultrashort radius radial system was pressured up via the frac pump. Upon reaching 10,000 psi operating pressure, the radials progressed through the whipstock and into the formation.
The water flow out of the nozzle at this point was approximately 160 gpm.
As the radial tubes passed through the whipstock and into the formation, the radial's rate of penetration (as well as pressure), flow, and other parameters were monitored, displayed, and recorded via the data acquisition system.
Fig. 5 shows the data acquisition system and a typical graphic display. Pressure, flow, and distance-vs.-time are shown.
For the Bremer well, the radial placement average velocities were calculated from the distance-vs.-time graph as 5.6 fpm for Radial 1, 3.5 fpm for Radial 2, 5.2 fpm for Radial 3, and 0.8 fpm for Radial 4.
When the radials reached approximately 7 ft beyond the exit of the whipstock, the control system was energized and remained on until the maximum radial distance was reached.
After the radials had been placed in the formation, the control system cable was removed from inside the radial tube. This allowed access to the radial for positional logging and completion.
A typical positional logging survey from the work at Bremer RI-53 is shown in Fig. 6.
Based on the DOE work, in normal commercial operations, four radials may be placed and completed in seven to eight 12 hr days made up of 1 day to underream and caliper, 1/2 day to set up Petrolphysics equipment, 6 days (1-1/2 days/radial) to place and complete the radials, and 1/2 day to remove Petrolphysics equipment.
CYCLIC STEAM OPERATIONS
Following the completion of Bremer RI-53 and prior to steam injection, the well was placed on primary production for 3 days. During this period, the well had an average daily production of 1 bo/d, for a total of 5 bbl gross fluid. The steam injection phase was conducted by the major oil company. Information is provided for the first cycle only.
STEAM INJECTION
Steam Cycle No. 1 was initiated on May 31, 1990, and continued for a total of 12 days.
During this 12-day period, 17,085 bbl of cola water, equivalent to 5,390 MMBTU, were injected into the well. The quantity and quality of the injected steam was consistent with the volume and quality of steam injected into offset wells on the Bremer lease.
No difficulties were encountered during the injection phase. No abnormal injection response nor reservoir receptivity to the injected steam was observed. Table 1 lists the observations.
Prior to injecting steam into Bremer RI-53, an observation well, Bremer 0-5, was drilled approximately 100 ft to the northwest. A temperature profile from the Bremer 0-5 location was taken before and after the steam injection phase (Fig. 7).
The temperature log substantiates the capability of radials to direct steam in specific azimuths and within specific elevations within a formation. The log also documents that steam has been effectively transported throughout the entire length of the radial, i.e., horizontal distance of at least 100 ft can be substantiated by Well 0-5 response.
On the log there is a substantial temperature increase from 800 to 873 ft. This temperature spike corresponds with the elevation of Radial No. 3 that terminated approximately 50 ft to the south of the observation well. Therefore, the steam has been effectively contained within the target interval.
The temperature log also indicates that the background temperature in the well, both above and below the radial effected zone, is at least 90 F. cooler.
Oil in these colder areas has not yet been mobilized because typical steam injection procedures have been ineffective in heating these zones.
Normally, steam is injected into the entire 400 + ft Potter interval. While the upper portion of the Potter is heated by this process, as indicated on the temperature profile from 350 to 700 ft, the basal portion of the Potter remains unaffected.
As a result, oil reserves attributable to the lower Potter sands remain cold and relatively immobile.
Therefore, one of the principal benefits of an injector or a cyclic producer completed with radials appears to be the ability to selectively provide heat to specific zones thereby accelerating production or perhaps even mobilizing additional oil in place.
Another key objective of the DOE project was to use radials to eliminate the deleterious effects of steam override near the vertical well bore.
The steam appears to stay within the selected interval, and there would be localized gravity override within that limited zone. Nevertheless, there have been no indications of steam override over the total Potter interval in Bremer RI-53 at this point.
Subsequent cycles will be necessary to determine if steam override in the total Potter interval is completely eliminated.
PRODUCTION PHASE
Production from Bremer RI-53 was monitored, and it was measured on various dates throughout Production Cycle No. 1. During the first week of production following the steam injection, production from Bremer RI-53 was quite low, averaging less than 5 bo/d. By comparison, a typical cyclic producer has a Relatively high initial production rate.
However, during the subsequent 3-week period, Bremer RI-53's production began to ramp up dramatically and peaked at approximately 60 bo/d. This production level was higher than typical producing wells drilled at a similar time and spacing.
This comparatively high level of oil production was also accompanied by a correspondingly low level of water production. A 30.5% water cut was measured.
Oil production remained at a relatively high plateau during approximately 3 months of observation from mid-July through the first of October. This consistent level of oil production was considered to be anomalously high, relative to the typical producer in the field.
Notwithstanding variations in reservoir characteristics and oil quality throughout the Bremer Fee, the operator considered early production results from RI-53 to be very encouraging.
Bremer RI-53, which was originally slated to be converted into a steam injector after first production, will remain as a cyclic producer for further testing. Future injection and production cycles Will now be available for additional evaluation. It is of interest that a pattern of relatively constant or unusually slow production decline had been observed in other heavy oil wells.
This desirable condition may be caused by several properties, including penetration of the near well bore damage, penetration of reservoir heterogeneity, and a lesser convergence or different flow condition associated with the multiple radials as compared to a conventional vertical well.
Multiple radials probably have two flow characteristics which are different from a vertical well.
First, with long multidirectional, high permeability radial flow channels, the fluid velocity is lower and in turn the pressure gradients within the reservoir are lower. This explains both reduced water/oil ratio or a lesser tendency to water cone.
Second, with multidirectional radials, deleterious reservoir heterogeneity and anisotropy effects should be less pronounced, and some oil that was immobile may become producible. Fig. 8 displays the initial production results from RI-53. Gross fluids, oil production, and temperature were recorded at various times. These data have been charted for the period from June 27 through Oct. 1, 1990.
In general, this first cycle production from RI-53 exceeded production rates from the typical producers having a much more extensive open hole completion.
ACKNOWLEDGMENT
The authors want to thank the U.S. DOE and Union Oil Co. of California for their support of this work.
BIBLIOGRAPHY
Dickinson, W., "Ultrashort Radius Radial System; A New Drilling and Completion Method for Rapid Development of Oil Sands Reservoirs," Fourth Unitar/UNDP Conference on Heavy Crude and Tar Sands, Edmonton, Alta., Aug. 7-12, 1988.
Dickinson, W., Knoll, R.G., Nordlund, R., and Dickinson, R.W., "Flexible Sand Barrier (FSB): A Novel Sand Control System," SPE Paper No. 18787, SPE California Regional Meeting, Bakersfield, Calif., Apr. 5-7, 1989.
Dickinson, W., Nelson, V., Smolarchuk, P.A., and Nordlund, R. "Conical Water jet Cleanout of Plugged Injector Wells," Paper No. 89-40-81, CIM 40th Annual Technical Meeting, Banff Springs, Alta., May 28-31, 1989. Dickinson, W., Anderson, R.R., and Dickinson, R.W., "The Ultrashort-Radius Radial System," SPE Drilling Engineering, September 1989, pp. 247-254.
Dykstra, H., and Dickinson, W., "Oil Recovery by Gravity Drainage Into Horizontal Wells Compared With Recovery From Vertical Wells," SPE Paper No. 19827, SPE 64th Annual Meeting, San Antonio, Oct. 8-11, 1989.
Dickinson, W., Pesavento, M.J., and Dickinson, R.W., "Data Acquisition, Analysis, and Control While Drilling with Horizontal Water jet Drilling Systems," CIM/SPE Paper No. 90-127, CIM/SPE Technical Meeting, Calgary, Alta., June 10-13, 1990.
Dickinson, W., Dickinson, R.W., Nees, J., Dickinson, E., and Dykstra, H.,"Field Production Results With The Ultrashort Radius Radial System in Unconsolidated Sandstone Formations," Proceedings of the 5th Unitar International Conference on Heavy Crude and Tar Sands, Caracas, Vol. 11, Aug. 4-9, 1991, pp. 307-26.
Toma, P., Reitman, V., and Dickinson, W., "Long and Ultrashort Turning Radius of Horizontal Wells: Predictions of Future Production Based on Today's Experience," World Petroleum Congress, Buenos Aires, Oct. 20-25, 1991.
Copyright 1992 Oil & Gas Journal. All Rights Reserved.
